Categorical quotients for actions of groupoids on varieties
Ian M. Musson · 2025
Read Categorical quotients for actions of groupoids on varieties by Ian M. Musson (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-13510.txt.
Introduction
Ian Musson studies in 2025 how to form clean quotient spaces when a group-like object acts on a geometric one. Groupoids widen groups by letting moves start and end at varied points. The case in view is a named Weyl groupoid acting on a plain affine variety, building on Musson's own earlier work.
The map to the quotient was known to carry good geometric traits. Musson now shows, when the base field is uncountable, that the map is a geometric quotient with the strongest sharing trait: it stays a quotient after any base change among schemes. A classic book result is reshaped to fit groupoids and even plain equivalence relations. Closed points and Jacobson schemes do much of the work in the proof.
Catalog record
| CREATOR | Ian M. Musson |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 18 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-13510
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Categorical quotients for actions of groupoids on varieties Ian M. Musson Abstract For certain actions of the Weyl groupoid 𝔚 \mathfrak{W} from [ 17 ] on an affine variety X X , geometric properties of the map π : X ⟶ Y = Spec 𝒪 ( X ) 𝔚 \pi:X\longrightarrow Y={\operatorname{Spec}\;}\mathcal{O}(X)^{\mathfrak{W}} were studied in [ 13 ] , In this paper we show that if the base field 𝚔 \mathtt{k} is uncountable, the map π \pi is a geometric quotient which is universal in the category of 𝚔 \mathtt{k} -schemes. To do this we adapt a result from [ 12 ] showing that a geometric quotient is universal in the category of 𝚔 \mathtt{k} -schemes, to quotients by groupoids and more generally by equivalence relations. In our approach a key role is played by the closed points and Jacobson schemes. 1 Introduction Throughout \mathtt{k} denotes an algebraically closed field and iff means if and only if. For a \mathtt{k} -algebra R , let {\operatorname{Max}\;}R denote the space of maximal ideals of R with the subspace topology inherited from {\operatorname{Spec}\;}R . Consider an extension R\subseteq S of \mathtt{k} -algebras with S finitely generated. If M\in{\operatorname{Max}\;}S , then S/M\cong\mathtt{k} , so if m=M\cap R , then \mathtt{k}\hookrightarrow R/m\cong(R+M)/M\subseteq S/M\cong\mathtt{k}. (1.1) Thus m\in{\operatorname{Max}\;}R and the map \pi:X={\operatorname{Spec}\;}S\longrightarrow Y={\operatorname{Spec}\;}R restricts to a map \pi_{\operatorname{max}}:{\operatorname{Max}\;}S\longrightarrow{\operatorname{Max}\;}R,\quad M\mapsto M\cap R. (1.2) As in [ 13 ] , we say the relative weak Nullstellensatz (RWN) for the pair (R,S) if \pi_{\operatorname{max}} is onto. By Zorn’s Lemma an equivalent condition is: if I\neq R is an ideal of R , then IS\neq S . If RWN holds for the pair (R,S) , and m\in{\operatorname{Max}\;}R , then for any M\in{\operatorname{Max}\;}S containing mS , we have M\cap R=m . Thus we can identify {\operatorname{Max}\;}R with the set of equivalence classes in {\operatorname{Max}\;}S for the equivalence relation determined by the fibers of \pi_{\operatorname{max}} . Since the residue fields are \mathtt{k} in both cases, maximal ideals of R,S correspond to \mathtt{k} -points so we have identifications X(\mathtt{k})={\operatorname{Max}\;}S,\quad Y(\mathtt{k})={\operatorname{Max}\;}R. (1.3) Recall that a ring is Jacobson if any prime ideal is an intersection of maximal ideals. To say anything about \pi it is natural to assume that R is Jacobson, since then as observed in [ 4 ] , the Zariski topology on Y is determined by its restriction to the set of \mathtt{k} -points, see Subsection 2.2 . A case of particular interest arises for invariant rings. For actions affine algebraic groups there is a vast literature, [ 2 ] , [ 3 ] , [ 11 ] , [ 10 ] , [ 12 ] . However it is widely recognized that symmetries of geometric structures are often best understood using groupoids, and for the action of a groupoid on X , there is little general theory. More generally, suppose \sim is an equivalence relation on the \mathtt{k} -points of an affine variety X . We are interested in the map \pi in the case S=\mathcal{O}(X) and the invariant ring R=\mathcal{O}(X)^{\sim} consists of all functions in S that are constant on equivalence classes in X(\mathtt{k}) . In [ 12 ] Proposition 0.1 it is shown that for group actions, a geometric quotient is universal in the category of \mathtt{k} -schemes. Our first main result Theorem 2.7 , gives an analog of this result for quotients by equivalence relations, provided the quotient is a Jacobson scheme. We note that quotients by equivalence relations are studied in [ 2 ] and [ 9 ] . We remark that the definitions in [ 12 ] are formulated for an action of a group pre-scheme over a base. In particular [ 12 ] Definition 0.3 gives the axioms for the action using commutative diagrams. This does not seem appropriate for actions of groupoids. Instead, for a variety X consider the groupoid \mathfrak{A}(X) with base all subvarieties of X , and morphisms all isomorphisms between subvarieties. We say the groupoid \mathfrak{G} acts on X if there is a functor \tau:\mathfrak{G}\longrightarrow\mathfrak{A}(X). (1.4) Consider the equivalence relation \sim on X(\mathtt{k}) which is generated by the symmetric relation x\sim y iff there is a morphism f:A\longrightarrow B in \mathfrak{G} such that x\in\tau(A) and y=\tau(f)(x) . The equivalence classes under \sim are called \mathfrak{G} -orbits . We denote \mathcal{O}(X)^{\sim} by \mathcal{O}(X)^{\mathfrak{G}} in this setting. Thus \mathcal{O}(X)^{\mathfrak{G}}=\{f\in\mathcal{O}(X)|f\mbox{ is constant on }\mathfrak{G}\mbox{-orbits in }X(\mathtt{k})\}. If R=S^{\mathfrak{G}} is the fixed ring under certain groupoid actions, we showed in [ 13 ] that the map \pi:X={\operatorname{Spec}\;}S\longrightarrow Y={\operatorname{Spec}\;}R has some desirable geometric properties. Here S is either a polynomial algebra or a Laurent polynomial algebra in a finite number of variables, and \mathfrak{G} is closely related to the Weyl groupoid introduced in [ 17 ] . The fixed ring arises in the study of the enveloping algebra of a classical simple Lie superalgebra. Our second result shows that provided \mathtt{k} is uncountable, \pi satisfies the conditions of Theorem 2.7 and is thus a geometric quotient and universal in the category of \mathtt{k} -schemes. I thank Hanspeter Kraft for some helpful correspondence. 2 Geometric and Categorical Quotients Let us recall some definitions. Throughout \mathtt{k} is an algebraically closed field and a variety X is an integral quasi-projective \mathtt{k} -scheme of finite type, [ 7 ] Proposition II.4.10. Each closed point of X has residue field isomorphic to \mathtt{k} . We denote the set of such points by X(\mathtt{k}) . Note that for every open subset V of X , the algebra {\mathcal{O}}_{X}(V) can be identified with the algebra of \mathtt{k} -valued functions on V(\mathtt{k}) . This is because the set of closed points of X is dense by [ 7 ] Exercise II.3.14. Suppose that \sim is an equivalence relation on X(\mathtt{k}) . Let Y be a \mathtt{k} -scheme. The morphism \pi:X\longrightarrow Y is universal in the category \mathcal{C} or a categorical quotient of X by \sim in the category \mathcal{C} if ( a ) \pi is constant on equivalence classes of closed points. ( b ) \pi is universal with respect to the property in (a). That is if \mu:X\longrightarrow Z is a morphism in \mathcal{C} which is constant on equivalence classes of closed points, then \mu factors uniquely through \pi . Suppose that \pi is as in (a) above and that U is open in Y . If V=\pi^{-1}(U) , then V(\mathtt{k}) is a union of equivalence classes. We denote by {\mathcal{O}}_{X}(\pi^{-1}(U))^{\sim} the functions in {\mathcal{O}}_{X}(\pi^{-1}(U)) which are constant on equivalence classes. The map \pi_{U}^{*}:\mathcal{O}_{Y}(U)\longrightarrow{\mathcal{O}}_{X}(\pi^{-1}(U))^{\sim} given by \pi_{U}^{*}(f)=f\pi is injective. Each part of the following definition is an analog of the corresponding part of [ 12 ] Defintion 0.6. Definition 2.1 . A morphism of \mathtt{k} -schemes \pi:X\longrightarrow Y is a geometric quotient if the following conditions hold (i) The induced map \pi(\mathtt{k}):X(\mathtt{k})\longrightarrow Y(\mathtt{k}) is constant on equivalence classes. (ii) \pi(\mathtt{k}) is onto and each fibre of \pi(\mathtt{k}) is an equivalence class. (iii) If A\subseteq X(\mathtt{k}) is a \sim -invariant closed subset, then \pi(A)\subseteq Y(\mathtt{k}) is closed. (iv) If U is any open set in Y then \pi_{U}^{*} is an isomorphism. Note that if \pi(\mathtt{k}) is onto, then (iii’) is equivalent to (iii´) If B\subseteq X(\mathtt{k}) is a \sim -invariant open subset, then \pi(B)\subseteq Y(\mathtt{k}) is open. 2.1 The affine case We need