Morita equivalence of pseudogroups (Q2049392): Difference between revisions

An \textit{inverse semigroup} is a semigroup \((S,\cdot)\), in which for every element \(s\in S\), there exists a unique element \(s^{-1}\in S\) (called the \textit{inverse} of \(s\) in \(S\)) such that \(s=s\cdot s^{-1}\cdot s\) and \(s^{-1}=s^{-1}\cdot s\cdot s^{-1}\). Given an inverse semigroup \(S\), two elements \(s,t\in S\) are said to be \textit{compatible} provided that both \(s^{-1}\cdot t\) and \(s\cdot t^{-1}\) are idempotent (an element \(s\) of a semigroup \(S\) is called \textit{idempotent} provided that \(s\cdot s=s\)). A subset of an inverse semigroup \(S\) is \textit{compatible} provided that either it is empty, or every two its elements are compatible. An inverse semigroup \(S\) is a \textit{pseudogroup} provided that every its compatible subset has a join, and multiplication distributes over all such joins, namely, if \(A\subseteq S\) is compatible, then \(\bigvee A\) exists, and for every \(s\in S\), it follows that \(s\cdot(\bigvee A)=\bigvee_{a\in A}s\cdot a\) and \((\bigvee A)\cdot s=\bigvee_{a\in A}a\cdot s\). For example, all partial homeomorphisms between the open subsets of a topological space make a pseudogroup. For standard results on inverse semigroups, one can consult~[\textit{M. Lawson}, Inverse semigroups. The theory of partial symmetries. Singapore: World Scientific (1998; Zbl 1079.20505)]. \par It is well known that two unital rings are said to be \textit{Morita equivalent} provided that their categories of left modules are equivalent (see, e.g.,~[F.~W.~Anderson and K.~R.~Fuller, \textit{Rings and Categories of Modules}, 2nd ed., Springer-Verlag, 1992; Zbl~0765.16001] for more detail). More general, two structures \(\mathcal{A}\) and \(\mathcal{B}\) are said to be \textit{Morita equivalent} provided that the category of left modules of \(\mathcal{A}\) is equivalent to the category of left modules of \(\mathcal{B}\) (the concept of ``left module'' should be defined in a suitable way for the structure in question). The present paper describes Morita equivalence of pseudogroups, based in the already existing theory of Morita equivalence of inverse semigroups of~[\textit{J. Funk} et al., J. Pure Appl. Algebra 215, No. 9, 2262--2279 (2011; Zbl 1229.20064)] as well as [\textit{B. Steinberg}, Houston J. Math. 37, No. 3, 895--927 (2011; Zbl 1236.46049)]. \par There exist two distinct approaches to dealing with pseudogroups: one can consider them as a class of inverse semigroups, or they can be connected to quantales, namely, inverse quantal frames~[\textit{P. Resende}, Adv. Math. 208, No. 1, 147--209 (2007; Zbl 1116.06014)]. These two approaches are related via two functors providing an equivalence of the respective categories: from a pseudogroup \(S\) one can construct an inverse quantal frame \(\mathcal{L}^{\vee}(S)\), and from an inverse quantal frame \(Q\) one can get a pseudogroup of partial units \(Q_{\mathcal{I}}\). The present paper describes the Morita theory of pseudogroups within the inverse semigroup approach, where the authors employ the sheaf theory of inverse quantal frames of~[\textit{P. Resende}, J. Pure Appl. Algebra 216, No. 1, 41--70 (2012; Zbl 1231.06020)] as well as their recently developed Morita theory based on Hilsum-Skandalis maps of~[\textit{J. P. Quijano} and \textit{P. Resende}, Appl. Categ. Struct. 29, No. 4, 629--670 (2021; Zbl 1468.18008)]. In particular, the authors manage to prove the following three main theorems. Theorem A. Given pseudogroups \(S\) and \(T\), the following three conditions are equivalent: (1) \(S\) and \(T\) are Morita equivalent pseudogroups; (2) \(\mathcal{L}^{\vee}(S)\) and \(\mathcal{L}^{\vee}(T)\) are Morita equivalent inverse quantal frames; (3) \(\mathsf{G}(S)\) and \(\mathsf{G}(T)\) are Morita equivalent localic étale groupoids. Theorem B. Two pseudogroups \(S\) and \(T\) are Morita equivalent if and only if there exists an \((S,T)\)-equivalence bimodule. Theorem C. Given pseudogroups \(S\) and \(T\), there exists an \((S,T)\)-equiva\-len\-ce bimodule if and only if \(S\) and \(T\) have a joint sup-enlargement. Lastly, a property \(\mathtt{P}\) of pseudogroups is said to be \textit{Morita invariant} provided that whenever a pseudogroup \(S\) has the property \(\mathtt{P}\) and a pseudogroup \(T\) is Morita equivalent to \(S\), then \(T\) has the property \(\mathtt{P}\). Towards the end of the paper, the authors describe two Morita invariant properties. The paper is well written, provides some of its required preliminaries (the omitted stuff can be always found following the carefully provided references), and will be of interest to all those researchers who study Morita equivalence of different algebraic structures and the respective Morita invariant properties.