Weakly Mal'tsev Objects

• Weakly Mal'tsev objects are defined in categories with local products by requiring at most one morphism in kite diagrams, capturing a unique extension property analogous to cancellation.
• They occupy a distinct position in the Mal'tsev hierarchy, being strictly weaker than Mal'tsev objects yet enforcing key local uniqueness conditions.
• Examples such as cancellative commutative magmas and distributive lattices illustrate how weakly Mal'tsev objects embody practical cancellation phenomena in algebraic contexts.

Searching arXiv for recent and foundational papers on weakly Mal'tsev objects and adjacent Mal'tsev notions. A weakly Mal'tsev object is an objectwise localization of the weakly Mal'tsev condition from categorical algebra. In the formulation introduced for categories with local products, an object $D$ is weakly Mal'tsev when every kite diagram over $D$ admits at most one compatible morphism from the associated local product; equivalently, the notion isolates a local uniqueness principle that behaves like cancellation in concrete algebraic examples (Martins-Ferreira, 18 Aug 2025). The concept belongs to a broader Mal'tsev hierarchy that includes naturally Mal'tsev categories, Mal'tsev categories, and several objectwise conditions such as unital, strongly unital, subtractive, Mal'tsev, and protomodular objects; however, the explicit notion of weakly Mal'tsev object is distinct from these nearby notions and was not developed in the earlier objectwise literature (Montoli et al., 2016).

1. Formal definition

Let $C$ be a category with local products. A local product is a pullback square of split epimorphisms

$\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$

with $p_1e_1=1_A$ and $p_2e_2=1_C$, where $E$ is the pullback of a split epimorphism $g$ along another split epimorphism $f$ (Martins-Ferreira, 18 Aug 2025).

In this setting, an object $D$ is called a weakly Mal'tsev object if, for every kite diagram over $D$0,

$D$1

with

$D$2

there exists at most one morphism

$D$3

such that

$D$4

When such a morphism exists, the diagram is said to be admissible, and $D$5 is the admissibility morphism (Martins-Ferreira, 18 Aug 2025).

Two points are intrinsic to the definition. First, the ambient category is only required to have local products. Second, the condition is a uniqueness condition: existence of $D$6 is not required in general. This distinguishes weakly Mal'tsev objects from stronger Mal'tsev-type conditions formulated by existence-and-uniqueness principles.

The same source denotes by

$D$7

This notation emphasizes that weak weak Mal'tsev behavior can be studied objectwise even when the ambient category is not globally weakly Mal'tsev (Martins-Ferreira, 18 Aug 2025).

2. Position in the Mal'tsev hierarchy

The weakly Mal'tsev object notion is designed to localize the corresponding categorical property. In particular, a category is weakly Mal'tsev if and only if every one of its objects is a weakly Mal'tsev object (Martins-Ferreira, 18 Aug 2025). Thus the objectwise notion is not merely analogous to the categorical one; it recovers it exactly by quantifying over all objects.

At category level, the weakly Mal'tsev condition is classically formulated using pullbacks of split epimorphisms. In a finitely complete category, weakly Mal'tsev means that the canonical maps

$D$8

are jointly epimorphic in every pullback of split epimorphisms along split epimorphisms (Egner et al., 2024). In the earlier structural account of Martins-Ferreira and Van der Linden, a category is weakly Mal'tsev when it has split pullbacks and every induced pair of morphisms into the pullback $D$9 is jointly epimorphic; with equalisers, this is equivalent to every strong relation being difunctional (Martins-Ferreira et al., 2012).

This places weakly Mal'tsev objects within a three-level hierarchy. The 2025 span-based synthesis organizes the hierarchy as follows: naturally Mal'tsev categories correspond to the strongest span class $C$0, Mal'tsev categories to $C$1, and weakly Mal'tsev categories to $C$2, the class of spans $C$3 such that kernel pairs of $C$4 and $C$5 exist and $C$6 is jointly strongly monic (Martins-Ferreira, 18 Aug 2025). In parallel, the 2012 study characterizes naturally Mal'tsev categories by uniqueness of internal groupoid structures on reflexive graphs, Mal'tsev categories by the fact that every reflexive relation is an equivalence relation, and weakly Mal'tsev categories by the analogous condition restricted to strong relations (Martins-Ferreira et al., 2012).

A crucial distinction is that weakly Mal'tsev is genuinely weaker than Mal'tsev. In weakly Mal'tsev categories with kernel pairs and equalisers, the forgetful functor

$C$7

is an isomorphism, so every internal category is determined by its underlying multiplicative graph; however,

$C$8

is an isomorphism only if every internal preorder is an equivalence relation (Martins-Ferreira et al., 2012). Thus weakly Mal'tsev behavior controls uniqueness of composition data more weakly than the full Mal'tsev property controls invertibility.

3. Kites, local products, and the uniqueness principle

The technical core of the weakly Mal'tsev object definition is the kite. Given a kite

$C$9

with

$\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$0

the associated local product has pullback object $\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$1 and induced maps

$\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$2

The weakly Mal'tsev object condition says that there is at most one morphism $\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$3 extending $\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$4 and $\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$5 along $\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$6 and $\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$7 (Martins-Ferreira, 18 Aug 2025).

This can be read as an objectwise admissibility criterion. The object $\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$8 does not force all kites to be admissible, but whenever a compatible filler exists, it is unique. A plausible implication is that weakly Mal'tsev objects capture the uniqueness half of a Mal'tsev-like extension property without imposing the stronger existence requirements typical of naturally Mal'tsev contexts.

The same paper introduces the Kite Condition, a stronger global condition formulated for diagrams

$\vcenter{\xymatrix{ A \ar@<-.5ex>[r]_-{e_1} & E \ar@<-.5ex>[l]_-{p_1} \ar@<.5ex>[r]^-{p_2} & C \ar@<.5ex>[l]^-{e_2} }}$9

subject to

$p_1e_1=1_A$0

$p_1e_1=1_A$1

$p_1e_1=1_A$2

$p_1e_1=1_A$3

together with kernel-pair assumptions for $p_1e_1=1_A$4 and $p_1e_1=1_A$5. Under these hypotheses there exists a unique $p_1e_1=1_A$6 with

$p_1e_1=1_A$7

For $p_1e_1=1_A$8, this yields the weakly Mal'tsev case in the paper’s unification theorem (Martins-Ferreira, 18 Aug 2025).

The emphasis on kites and local products connects the objectwise uniqueness condition to the older characterizations of weakly Mal'tsev categories by strong relations and universal filler properties. In the 2012 analysis, the weakly Mal'tsev condition is equivalent to a universal property for strong relations and to the statement that every reflexive strong relation is an equivalence relation, a preorder, or transitive (Martins-Ferreira et al., 2012). The objectwise theory retains the same formal geometry but relocates it from ambient-category structure to a condition on a fixed codomain object.

4. Algebraic characterizations and examples

The objectwise definition has concrete algebraic consequences, expressed as uniqueness-of-solution conditions.

In the category of commutative magmas, the following are equivalent:

1. $p_1e_1=1_A$9 is weakly Mal'tsev.
2. $p_2e_2=1_C$0 is cancellative:

$p_2e_2=1_C$1

1. For any $p_2e_2=1_C$2, the equation

$p_2e_2=1_C$3

has at most one solution $p_2e_2=1_C$4.

Hence

$p_2e_2=1_C$5

This identifies weakly Mal'tsev objects with cancellative commutative magmas; the same source notes that associativity is irrelevant to weak Mal'tsev-ness in this example (Martins-Ferreira, 18 Aug 2025).

In the category of commutative dimagmas $p_2e_2=1_C$6, weakly Mal'tsev objects are characterized by joint cancellation: $p_2e_2=1_C$7 equivalently, the system

$p_2e_2=1_C$8

has at most one solution $p_2e_2=1_C$9 (Martins-Ferreira, 18 Aug 2025).

In unary monoids $E$0, the condition becomes the uniqueness of solutions to

$E$1

The paper presents these examples as evidence that the abstract kite condition encodes a broad cancellation phenomenon (Martins-Ferreira, 18 Aug 2025).

Other explicitly identified examples are equally striking. The same paper states

$E$2

so a lattice is a weakly Mal'tsev object if and only if it is distributive, and

$E$3

so weakly Mal'tsev objects in commutative semigroups are precisely the cancellative ones (Martins-Ferreira, 18 Aug 2025). It also lists broader examples of weakly Mal'tsev categories, including distributive lattices, commutative magmas with cancellation, preordered groups, and the dual of the category of topological spaces.

These examples align with the syntactic theory of weakly Mal'tsev varieties. A finitary one-sorted variety is weakly Mal'tsev if and only if it admits a finite family of terms

$E$4

satisfying a chain of identities rather than a single Mal'tsev term (Egner et al., 2024). The distributive lattice case is one of the main applications: distributive lattices are weakly Mal'tsev even though they are not Mal'tsev in general (Egner et al., 2024). This provides a variety-level explanation for why distributivity reappears in the objectwise classification $E$5.

5. Comparison with Mal'tsev objects and adjacent local notions

Weakly Mal'tsev objects should not be conflated with Mal'tsev objects. In the objectwise theory of Montoli, Rodelo, and Van der Linden, developed in the finitely complete setting, an object $E$6 is a Mal'tsev object if the fibre category

$E$7

is unital (Montoli et al., 2016). Equivalently, for every pullback of split epimorphisms over $E$8, the canonical maps

$E$9

are jointly strongly epimorphic. This is a stronger condition than weak weak-Mal'tsev uniqueness of admissibility morphisms.

The same paper builds a local hierarchy: $g$0 with

$g$1

in a pointed regular category (Montoli et al., 2016). These notions are adjacent to weakly Mal'tsev objects by analogy, but they are not the same notion. The authors explicitly state that the paper does not define or discuss weakly Mal'tsev objects.

The 2023 comparison of object-level Mal'tsev notions sharpens this distinction further. It compares W-Mal'tsev objects in the sense of Weighill—defined via difunctionality of all induced Set-relations on hom-sets out of the object—with Mal'tsev objects in the sense of Montoli–Rodelo–Van der Linden, defined via the fibration of points (Clementino et al., 2023). Under regularity and binary coproducts,

$g$2

so Mal'tsev objects are stronger than W-Mal'tsev objects (Clementino et al., 2023). That paper also states explicitly that it does not develop a separate theory of weakly Mal'tsev objects; weakly Mal'tsev categories appear there only as ambient context.

Concrete categories show how these distinctions behave. In $g$3, Mal'tsev objects and protomodular objects are exactly groups: $g$4 (Montoli et al., 2016). The 2023 comparison recalls that in $g$5,

$g$6

as well (Clementino et al., 2023). By contrast, weakly Mal'tsev behavior at category level is formally weaker than Mal'tsev, and the objectwise weakly Mal'tsev notion is designed to capture a still different localization of that weaker property.

A plausible implication is that the term “weakly Mal'tsev object” now occupies a precise conceptual slot between global weakly Mal'tsev ambient structure and the earlier local theories based on points or difunctionality. It is not a renaming of those earlier notions, but a new local form of the weakly Mal'tsev axiom.

6. Scope, examples of strictness, and surrounding contexts

The weakly Mal'tsev condition is strictly weaker than the Mal'tsev condition both categorically and in varieties. The 2012 paper constructs weakly Mal'tsev quasivarieties that are still Goursat but not Mal'tsev, including a modification of Mitschke’s example with implication algebras satisfying

$g$7

where the resulting sub-quasivariety $g$8 is weakly Mal'tsev and Goursat but not Mal'tsev because $g$9 for specific kernel relations (Martins-Ferreira et al., 2012). The 2024 syntactic work likewise emphasizes that weakly Mal'tsev is a genuine weakening: a weakly Mal'tsev variety is characterized by a large finite family of terms, not by a single ternary Mal'tsev term (Egner et al., 2024).

The 2025 paper also stresses that weakly Mal'tsev objects remain meaningful under weak ambient assumptions: only local products are required for the definition, and in the broader structural theorem only suitable kernel-pair assumptions on the class $f$0 are needed (Martins-Ferreira, 18 Aug 2025). This mirrors the earlier observation that the main characterizations of naturally Mal'tsev, Mal'tsev, and weakly Mal'tsev categories can be proved without assuming binary products, working instead with kernel pairs and split pullbacks (Martins-Ferreira et al., 2012).

There are also significant ambient contexts in which related objectwise notions become highly nontrivial. In $f$1, which is recalled to be a weakly Mal'tsev category, W-Mal'tsev objects are exactly the symmetric $f$2-categories, whereas Mal'tsev objects can collapse to the empty $f$3-category when $f$4 is non-cartesian (Clementino et al., 2023). This does not provide a characterization of weakly Mal'tsev objects in that sense, but it shows that local Mal'tsev-type notions may diverge sharply once one moves away from classical algebraic categories.

Finally, the topological literature around Mal'tsev operations supplies useful contrast rather than a direct source of weakly Mal'tsev objects. In the study of free topological Mal'tsev algebras, a Mal'tsev topological algebra is defined by a continuous ternary operation $f$5 satisfying

$f$6

and varieties with such a term are exactly the congruence-permutable ones (Sipacheva et al., 2024). That framework concerns the classical strong Mal'tsev identity, not the weakly Mal'tsev object notion. This suggests that weakly Mal'tsev objects should be viewed not as objects carrying a ternary Mal'tsev operation, but as objects satisfying a local uniqueness property for admissibility across kites and local products.

In summary, a weakly Mal'tsev object is an objectwise incarnation of the weak weak-Mal'tsev axiom: every compatible local-product extension into the object is unique when it exists (Martins-Ferreira, 18 Aug 2025). Its concrete manifestations are cancellation phenomena, its categorical background is the theory of strong relations and split pullbacks, and its significance lies in separating a genuinely weaker local uniqueness principle from the stronger frameworks of Mal'tsev operations, Mal'tsev objects, and protomodularity.