Combinatorial categorical equivalences of Dold-Kan type (Q2348111)

The famous Dold-Puppe-Kan theorem gives an equivalence of categories between simplicial \(R\)-modules and nonnegative chain complexes of \(R\)-modules. It is often applied to simplicial objects arising from (co)triples in homological algebra to produce a variety of standard resolutions and other chain complexes for the purposes of computing homology. The present paper by Lack and Street puts the Dold-Puppe-Kan theorem into a new context via the following interesting generalisation: given a category \(\mathcal{P}\) and a subcategory \(\mathcal{M}\) satisfying some properties (examples will be given later), the authors' construction produces a new category \(\mathcal{D}\) enriched over pointed sets, and an equivalence of the form \[ [\mathcal{P},\mathcal{X}]\simeq [\mathcal{D},\mathcal{X}]_{\mathrm{pt}}. \] Here \(\mathcal{X}\) is any additive category with finite direct sums and splitting for idempotents (e.g. an abelian category), \([-,-]\) denotes the functor category, and \([-,-]_{\mathrm{pt}}\) denotes the functor category of functors that preserve the zero morphisms. The hypotheses on the category \(\mathcal{P}\) are somewhat extensive and subtle, so we will forgo listing them here, but yet they still encompass a wide variety of examples. One such is the usual simplicial category, and in this case the category \(\mathcal{D}\) is the expected ``chain complex category'' so that the functors from \(\mathcal{D}\) into an abelian category preserving the zero morphisms forms the category of chain complexes, so the Dold-Kan theorem is indeed a special case. Should one find the production of categories \(\mathcal{P}\) a little mysterious given the hypothesis on \(\mathcal{P}\) required for the equivalence, the authors prove a construction theorem that produces a suitable \(\mathcal{P}\) from the data of any category \(\mathcal{A}\) with a factorisation system \((\mathcal{E},\mathcal{M})\) (as introduced in [\textit{P. J. Freyd} and \textit{G. M. Kelly}, J. Pure Appl. Algebra 2, 169--191 (1972; Zbl 0257.18005)]) such that: (a) \(\mathcal{M}\) is closed under pullbacks along any morphism, (b) \(\mathcal{M}\) contains only monomorphisms, and (c) every object of the category \(\mathcal{A}\) has finitely many subobjects in \(\mathcal{M}\). Lack and Street then prove that the category \(\mathcal{P}\) whose objects are the objects of \(\mathcal{A}\) and whose morphisms are the isomorphism classes of partial maps with respect to \(\mathcal{M}\). Under this factorisation system construction, the paper also includes a second equivalence of categories of the form \[ [\mathcal{P},\mathcal{X}]\simeq[\mathcal{E},\mathcal{X}] \] where \(\mathcal{E}\) is the other class in the factorisation system (note the lack of pointed functors). A special case is given when \(\mathcal{E}\) is the category of finite sets and bijections, and \(\mathcal{P}\) is the category of finite sets and injective partial functions, which gives the equivalence of categories in the paper of \textit{T. Church} et al. [Duke Math. J. 164, No. 9, 1833--1910 (2015; Zbl 1339.55004)]. The authors state this example in fact as the original motivation for their paper and they also give many other interesting examples that should appeal to a wide variety of algebraists. The paper contains eight numbered sections and two appendices. What has just been described together with proofs and additional examples comprises the first seven sections. The authors also prove a modified result along the lines of \textit{D. Bourn}'s [Contemp. Math. 431, 105--124 (2007; Zbl 1143.18013)] with an additional assumption when \(\mathcal{X}\) is only semiabelian (such as the category of all groups and group homomorphisms), and in this case the prove the appropriate functor is crudely monadic. Finally, the appendices hold fairly technical results on categories enriched over closed complete and cocomplete symmetric monoidal categories and on idempotents that are used to prove the main results of the paper.