Fibrational approach to Grandis exactness for 2-categories
Caviglia, Janelidze, Mesiti
In an abelian category, the (bi)fibration of subobjects is isomorphic to the (bi)fibration of quotients. This property captures substantial information about the exactness structure of a category. Indeed, as it was shown by the second author and T.~Weighill, categories equipped with a proper factorization system such that the opfibration of subobjects relative to the factorization system is isomorphic to the fibration of relative quotients are precisely the Grandis exact categories. In this paper we characterize those (1,1)-proper factorization systems on a 2-category in the sense of M.~Dupont and E.~Vitale, for which the weak 2-opfibration of relative 2-subobjects is biequivalent to the weak 2-fibration of relative 2-quotients. This results in a new notion of 2-dimensional exactness, which we then compare with similar notions in the context of categories enriched in pointed groupoids arising in the work of M.~Dupont and H.~Nakaoka.
academic
Fibrational approach to Grandis exactness for 2-categories
In abelian categories, the (bi)fibration of subobjects is isomorphic to the (bi)fibration of quotient objects. This property captures essential information about the exact structure of a category. As proven by the second author with T. Weighill, categories equipped with appropriate factorization systems whose subobject opfibration relative to that system is isomorphic to the quotient object fibration are precisely Grandis exact categories. This paper characterizes those (1,1)-appropriate factorization systems on 2-categories (in the sense of M. Dupont and E. Vitale) such that the weak 2-opfibration of relative 2-subobjects is biequivalent to the weak 2-fibration of relative 2-quotient objects. This leads to a new concept of 2-dimensional exactness and is compared with analogous concepts developed by M. Dupont and H. Nakaoka in the context of groupoid-enriched categories.
This paper aims to generalize the classical theory of Grandis exactness from 1-categories to 2-categories, establishing a theory of 2-dimensional exactness through fibrational methods.
2-dimensional homological algebra: Homological algebra in 2-dimensional categories is an emerging and important research area, particularly significant for studying homological properties of symmetric categorical groups.
Applications of fibration theory: Fibration theory provides profound geometric intuition for understanding exact structures of categories; generalizing this to the 2-dimensional case has important theoretical value.
Unifying existing theories: Existing concepts of 2-dimensional exactness (such as work by Dupont and Nakaoka) lack a unified theoretical framework; this paper provides such a unified perspective.
To establish a theory of 2-dimensional Grandis exactness through fibrational methods, providing a solid theoretical foundation for 2-dimensional homological algebra and unifying various existing concepts of 2-dimensional exactness.
Introduction of 2-ideals: Proposes rigorous definitions of 2-ideals in 2-categories (ideals of null morphisms and null 2-cells), providing complete characterizations via the profunctor approach.
Establishment of 2-dimensional kernel and cokernel theory: Defines 2-kernels and 2-cokernels relative to 2-ideals and proves their fundamental properties.
Proof of main equivalence theorem: Proves that a 2-category is Grandis 2-exact if and only if there exists a (1,1)-appropriate factorization system such that the weak 2-fibration of quotient objects and the weak 2-opfibration of subobjects are biequivalent on the 2-category.
Introduction of weak exactness concepts: Proposes concepts of weak Grandis 2-exactness and weak Puppe 2-exactness, proving that these concepts generalize existing theories by Dupont and Nakaoka.
Establishing connections with existing theories: Proves that this theory includes all 1-dimensional abelian categories as special cases and establishes explicit connections with existing 2-dimensional exactness theories.
Definition 2.2: A 2-ideal N in a 2-category L is a pair (N,ν), where:
N: L^op × L → Cat is a normal pseudofunctor (2-dimensional profunctor)
ν is an injective and faithful pseudonatural transformation:
L^op × L → Cat
N ↘ ↗ Hom(-,-)
ν
Theorem 2.4 provides an equivalent characterization of 2-ideals: through classes of null morphisms and null 2-cells, satisfying appropriate closure conditions.
Definition 2.10: The N-2-kernel of a morphism f: A → B is a morphism k: K → A, equipped with an isomorphic 2-cell α: f∘k ≅ n (n a null morphism), satisfying:
1-dimensional universal property: For any z: Z → A such that f∘z is isomorphic to a null morphism, there exists a unique u: Z → K and isomorphic 2-cell γ: z ≅ k∘u
2-dimensional universal property: For morphisms u,v: Z → K and 2-cell λ: k∘u ⇒ k∘v, if the corresponding 2-cell is null, then there exists a unique μ: u ⇒ v such that k⋆μ = λ
Profunctor method: Using the profunctor approach to define 2-ideals is more natural and powerful than direct generalization of 1-dimensional definitions.
Weak fibration concept: Introduces the concept of weak 2-fibration, requiring only local isofibration rather than complete fibration, which is more appropriate in the 2-dimensional case.
Relative orthogonality: In the weak exactness case, introduces the concept of relatively orthogonal factorization systems, an important generalization of classical orthogonal factorization systems.
Biequivalence rather than isomorphism: Uses biequivalence rather than strict isomorphism in the 2-dimensional case, reflecting the essential characteristics of 2-categories.
Three-step factorization: In the general case, morphisms have three-step factorization (Proposition 3.8), but this simplifies to two steps in the Grandis 2-exact case.
2-dimensional generalization of the first isomorphism theorem: Corollary 3.7 can be viewed as a 2-categorical generalization of the first isomorphism theorem in algebra.
Hierarchical structure of theories: Establishes clear hierarchical relationships between various 2-dimensional exactness concepts.
The paper cites 38 important references, primarily including:
I. Baković - Foundational work on fibrations in 2-categories
M. Dupont - Pioneering work on 2-dimensional abelian categories
H. Nakaoka - 2-dimensional cohomology theory and relative exactness
Z. Janelidze & T. Weighill - Fibrational methods for 1-dimensional Grandis exactness
R. Street - Foundational work on 2-category theory
These references form the important foundation for the theoretical development of this paper, reflecting the authors' deep understanding and comprehensive grasp of related fields.
Overall Assessment: This is a high-quality pure mathematics theoretical paper making important contributions at the intersection of 2-dimensional category theory and homological algebra. While technically complex, the theoretical value is significant, laying important foundations for the development of 2-dimensional homological algebra. The paper's unifying perspective and successful generalization of fibrational methods have important methodological significance.