Dimensional quasi-Frobenius groups

From MaRDI portal



Abstract: We are interested in a class of groups, quasi-Frobenius groups (with involutions), whose internal structure generalizes that of the classical groups GA1(C), PGL 2(C) and SO3(R) : a subgroup and its conjugates, of finite index in their normalizer and trivial mutual intersection, cover "generically" the ambient group. From the perspective of model theory, we work with the hypothesis of the existence of a good notion of dimension on definable sets (we must distinguish between the o-minimal case and the ranked case). We pay special attention to the ranked case. By studying the geometry of incidence induced by involutions, we sketch a classification of quasi-Frobenius groups and thus determine under which conditionsclassical groups can be identified in a dimensional framework -- -- Nous nous int'eressons `a une classe de groupes, les quasi-groupes de Frobenius (avec involutions), dont la structure interne g'en'eralise celle des groupes classiques GA1(C), PGL2(C) et SO3(R) : un sous-groupe et ses conjugu'es, d'indice fini dans leur normalisateur et d'intersection mutuelle triviale, recouvrent "g'en'eriquement" le groupe ambiant. Dans la perspective de la th'eorie des mod`eles, nous travaillons avec l'hypoth`ese de l'existence d'une bonne notion dimension sur les ensembles d'efinissables (il faut distinguer le cas o-minimal et le cas rang'e). Nous accordons une attention particuli`ere au cas rang'e. En 'etudiant la g'eom'etrie d'incidence induite par les involutions, nous esquissons une classification des quasi-groupes de Frobenius et nous d'eterminons ainsi sous quelles conditions des groupes classiques peuvent ^etre identifi'es dans un cadre dimensionnel.


In this paper, the author studies a generalisation of Frobenius groups -- quasi-Frobenius groups -- in the finite Morley rank and in the o-minimal context, with the aim of generalising results obtained for Frobenius groups of finite Morley rank, and proving identifications theorems as \(\mathrm{PGL}_2({\mathbb C})\) and \(\mathrm{SO}_3({\mathbb R})\). A group \(G\) with a subgroup \(C\) (called the Frobenius quasi-complement) is called a quasi-Frobenius group if \(C\) has finite index in its normaliser, and trivial intersection with its distinct conjugates. The author studies such groups of \textit{odd type}, i.e.\ which contain a Prüfer-\(2\)-subgroup but no infinite elementary abelian \(2\)-subgroup. He moreover distinguishes according to whether the degree \(|N_G(C):C|\) is even or odd. He conjectures: \begin{itemize} \item [(\textit{A. Deloro} and \textit{J. Wiscons}, Bull. Lond. Math . Soc. 52, No. 3, 411--428, 2020; Zbl 1528.20048)] If \(G\) and \(C\) are connected of finite Morley rank and even degree, then \(G\) is isomorphic to \(\mathrm{PGL}_2(K)\) for an algebraically closed field \(K\). \item If \(C<G\) are connected and definable in an o-minimal structure and if all translations (products of two involutions) are in some conjugate of \(C\), then \(G\) is isomorphic to \(\mathrm{SO}_3(R)\) for some real closed field \(R\). \end{itemize} The author shows this under some additional hypotheses. More precisely: {Theorem 1.} If \(C<G\) are connected and definable in an o-minimal structure such that all \(2\)-elements and all translations are contained in some conjugate of \(C\), then \(G\) is semi-simple; if moreover \(C\) is nilpotent, then \(G\) is isomorphic to \(\mathrm{SO}_3(R)\) for some real-closed field \(R\). {Theorem 2.} If \(C<G\) are connected and of finite Morley rank of even degree and such that every Borel subgroup is generous (i.e.\ the union of its conjugates is a generic subset of \(G\)), then \(G\) is isomorphic to \(\mathrm{PGL}_2(K)\) for an algebraically closed field \(K\). {Theorem 3.}If \(C<G\) is a connected Frobenius group of finite Morley rank such that \(C\) is soluble and \(G\) isn't and if the Borel subgroups are generous, then \(C\) is not nilpotent. The proofs are generalisations of the techniques used for Frobenius groups of finite Morley rank. It should be noted that the author tries to unify the proofs for the finite Morley rank and the o-minimal case. A major difference lies in the fact that if an involutive automorphism acts on a connected subgroup with finite fixed subgroup, it is not known in the o-minimal case that the subgroup is abelian and inverted by the automorphism (this is standard in the superstable case).



Cites work









This page was built for publication: Dimensional quasi-Frobenius groups

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6166187)