Classification of \(\big(\widetilde{\text{Sp}}(n,\mathbb{R})\times\widetilde{\text{Sp}}(1,\mathbb{R})\big )\)-manifolds (Q2010854)

Suppose a Lie group \(G\) acts effectively on a smooth manifold \(M\). Following the Zimmer Program, one expects rigidity if \(G\) is ``large'' compared to \(M\). The authors establish such rigidity when \(G\) is isomorphic to \(\widetilde{Sp}(n,\mathbb{R})\times \widetilde{Sp}(1,\mathbb{R})\). Specifically, assuming \(\dim M \leq (2n+3)(n+1) = \dim G + 4n\) and some technical hypothesis on \(M\) and the action, they show that either \(M\) is weakly reducible, or \(\dim M = (2n+3)(n+1)\) and \(M\) is covered by a space of the form \(\widetilde{Sp}(n+1,\mathbb{R})/\Gamma\) where \(\Gamma\subseteq \widetilde{Sp}(n+1,\mathbb{R})\) a lattice. The idea of the proof is to study the action of the Lie algebra \(\mathfrak{g}\) of \(G\) via Killing fields on \(M\) on the Killing fields on \(M\). The authors show that for most \(x\in M\), \(\mathfrak{g}\) embeds into the collection of Killing fields vanishing at \(x\). This, in turn, allows them to understand the action of \(\mathfrak{g}\) on the normal space to the orbits. Representation theory and the dimension estimates allow the authors to extend the \(\mathfrak{g}\) action to an action by \(\mathfrak{sp}(n+1,\mathbb{R})\), and thus, extend the \(G\) action to one by \(\widetilde{Sp}(n+1,\mathbb{R})\).