Endogenies and Linearisation

Adrien Deloro and Frank Wagner · 2024

Read Endogenies and Linearisation by Adrien Deloro and Frank Wagner (2024) 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-2412-13316.txt.

Introduction

Adrien Deloro and Frank Wagner proved that a broad class of symmetry actions in model theory can be recast in linear form, in a 2024 paper. The actions come from two endless, commuting families of maps on a finite-dimensional structure that cannot be split into smaller parts. The team shows that such an action turns into motion on a vector space over a field that the structure itself defines. The same result holds when the maps are loosened to a related notion called endogenies, apart from a finite leftover part.

Catalog record

CREATORAdrien Deloro and Frank Wagner
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE66 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-13316

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Endogenies and Linearisation Adrien Deloro and Frank Wagner Abstract . We show that the action of two infinite commuting invariant rings of endomorphisms of a finite-dimensional virtually connected irreducible bi-module linearizes into a vector space over a definable field. The same holds if the action is merely by strongly commuting endogenies, modulo some finite katakernel. 2000 Mathematics Subject Classification 03C60, 03C45, 13L05 — Para la otra vez que lo mate — replicó Scharlach — le prometo ese laberinto, que consta de una sola línea recta y que es invisible, incesante. La muerte y la brújula Introduction The present work studies—under the model-theoretic hypothesis of finite-dimensionality—irreducible bi-modules, i.e. abelian groups together with two commuting subrings of endomorphisms, and which are irreducible for the bi-action. This enhances the result in [ 2 ] , which itself stems from our initial investigations. In fact, we are working with a more general notion than endomorphism, namely endogenies , which are endomorphisms blurred by a finite subgroup. We work in finite-dimensional theories in the sense of [ 6 ] (see Definition 1.25 ), and contrary to customary practice, in the category of virtually connected groups (i.e., connected-by-finite). We first show that in a finite-dimensional theory, an irreducible virtually connected (\Gamma,\Delta) -bi-module A admits a quotient by a finite subgroup F such that \Gamma and \Delta act on A/F as actual endomorphisms rather than endogenies (Theorem A), and then characterise this action in Theorem B. Remark . Irreducibility is with respect to the bi-action, and in the category of virtually connected bi-modules. In line with common model-theoretic teminology, we shall call this (\Gamma,\Delta) -minimality. In case \Gamma and \Delta consist of endogenies as discussed in Section 1 , invariance splits into a weak and a full form; the former is more important, and minimality is with respect to it (see Section 1.5 ). Commutation is in the sharp sense of Section 1.4 We shall show: Theorem A . In a finite-dimensional theory, let A be a definable, virtually connected, abelian group. Let \Gamma and \Delta be two invariant prerings of definable endogenies of A such that A is (\Gamma,\Delta) -minimal. Suppose \Gamma and \Delta commute sharply, both are essentially infinite, and one of them is essentially unbounded. Then there is a finite fully \Gamma - and \Delta -invariant subgroup F such that the induced action of \Gamma and \Delta on A/F is by endomorphisms. In fact, F is given explicitly. Theorem B . In a finite-dimensional theory, let A be a definable, connected, abelian group. Let \Gamma and \Delta be two invariant rings of definable endomorphisms of A such that A is (\Gamma,\Delta) -minimal. Suppose that \Gamma and \Delta commute, both are infinite and one of them is unbounded. Then there is a definable field \mathbb{K} such that A is definably a \mathbb{K} -vector space of finite linear dimension, and both \Gamma and \Delta act \mathbb{K} -linearly. Endogenies are discussed at length in Section 1 . More specifically, prerings are explained in Section 1.2 ; essential sizes are discussed in Section 1.3 ; sharp commutation is in Section 1.4 ; then the global katakernel is defined in Section 1.6 . Theorem A is proven in Section 2 and Theorem B in Section 3 . To our knowledge, this is the first time endogenies are used for linearisation. Linearisation using endomorphisms is now common and goes back to Zilber (see [ 2 , § 1.2] for an attempted history; we could however not locate Zilber’s famous ‘ viniti preprint’). Acknowledgements. The first author has a variety of institutions to thank: Stellenbosch University, the Hausdorff Research Institute for Mathematics, and Université Claude Bernard Lyon 1. The second author would like to acknowledge the support of the Science Committee of the Ministry of Science and Higher Education of the Republic of Kazakhstan (Grant No. AP19677451). 1. Endogenies This section describes a variation on endomorphisms called endogenies , which are like morphisms but ‘blurred’ by finite subgroups. Some of their properties can disconcert at first and one may wish to spend a little time computing and checking statements in order to become familiar with them. We shall add some definability and connectedness in Sections 1.7 and 1.8 ; the beginning is purely algebraic. Endogenies abound in the theory of elliptic curves, with which we assume no familiarity. A simpler example arises as follows. Let \mathbb{K}\models\mathrm{ACF}_{p} and A=(\mathbb{K},+) . The Artin-Schreier endomorphism of A , viz., f(x)=x^{p}-x , is onto with a finite kernel. Its inverse relation f^{-1} has domain A and is additive (viz., is a subgroup of A\times A ), but behaves like a p -valued polymorphism. For instance it ‘takes’ 0 to the subgroup \mathbb{F}_{p} . To gain intuition through practice, one should compose f^{-1} with itself, and then with f on both sides. To our knowledge endogenies made their appearance in model theory (under the name of quasi-endomorphism ) in the beautiful paper [ 1 ] which proved that \aleph_{0} -categorical \aleph_{0} -stable connected groups are abelian. The arise quite naturally: If an abelian group A decomposes as almost linearly independent sum of definable minimal groups, then any definable group gives rise to endogenies between the components, and is itself characterized by these endogenies. Ever since this has been a major ingredient in the study of \aleph_{0} -categorical tame groups [ 5 , 4 , 3 ] . 1.1. The definition The general definition could be formulated with non-commutative groups, but we shall quickly require abelianity. We therefore assume it from the outset. Let A_{1},A_{2} be abelian groups. Definition 1.1 . An endogeny from A_{1} to A_{2} is a subgroup \gamma\leq A_{1}\times A_{2} with the following properties: • \gamma is global, viz., \pi_{1}(\gamma)=A_{1} (where \pi_{i}\colon A_{1}\times A_{2}\to A_{i} is the canonical projection); • the katakernel \operatorname{kat}\gamma=\{a\in A_{2}:(0,a)\in\gamma\} is finite. If A_{2}=A_{1}=A , we simply call \gamma an endogeny of A . Remarks 1.2 . • In Section 1.3 we shall introduce a natural equivalence relation on endogenies. • (The graph of) a morphism from A_{1} to A_{2} is also an endogeny; the converse does not hold. • An endogeny from A_{1} to A_{2} may also be viewed: – as an additive relation \gamma on A_{1}\times A_{2} with \pi_{1}\gamma=A_{1} and finite fibre over 0\in A_{1} ; – as a multi-but-finite-valued morphism A_{1}\to A_{2} ; – or as a (classical) morphism g\colon A_{1}\to A_{2}/\operatorname{kat}\gamma . Be careful that the image of an element a_{1}\in A_{1} under \gamma is a set , namely a coset of \operatorname{kat}\gamma . To avoid confusion we write \gamma[a_{1}] , with square brackets. Hence \operatorname{kat}\gamma=\gamma[0] and \operatorname{im}\gamma=\gamma[A] . Before going any further, one should check under all interpretations above that \gamma^{-1}[\operatorname{kat}\gamma]=\ker\gamma . 1.2. The prering of endogenies Definition 1.3 . Let A be an abelian group, and \operatorname{Endog}(A) the set of endogenies of A . We equip \operatorname{Endog}(A) with the following two operations: • pointwise addition/subtraction, viz.: (\gamma_{1}+\gamma_{2}):=\{(a,b_{1}+b_{2}):b_{1}\in\gamma_{1}[a]\wedge b_{2}\in\gamma_{2}[a]\}; • product as composition of relations, viz.: (\gamma_{1}\circ\gamma_{2}):=\{(a,c):\exists b\,(b\in\gamma_{2}[a])\wedge(c\in\gamma_{1}[b])\}. One should quickly check that \operatorname{Endog}(A) is closed under these operations. We omit \circ in practice. One sees easily that \operatorname{kat}(\gamma_{1}+\gamma_{2})=\operatorname{kat}\gamma_{1}+\operatorname{kat}\gamma_{2} and \operatorname{kat}(\gamma_{1}\gamma_{2})=\gamma_{1}[\operatorname{kat}\gamma_{2}] . Remark 1.4 . Functional interpretation: • Each \gamma_{i} corresponds to a morphism g_{i}\colon A\to A/\operatorname{kat}\gamma_{i} , which composes to h_{i}\colon A\to A/(\operatorname{kat}\gamma_{1}+\operatorname{kat}\gamma_{2}) .

Read the complete text