A characterization of certain families of strong parapolar spaces (Q1126973)

\textit{A. M. Cohen} and \textit{B. N. Cooperstein} [Geom. Dedicata 15, 73-105 (1983; Zbl 0541.51010)] characterized certain parapolar spaces of finite singular rank \(s\) with thick lines such that \(\{x,y\}^\perp\) is a nondegenerate polar space of rank \(k\) and such that for every point-symplecton pair \((x,S)\) the spaces \(x^\perp\cap S\) are empty, a single point or a maximal singular subspace of \(S\). The present authors call this last unnatural sounding condition the 'gap phenomenon' since it postulates a gap in the spectrum of projective ranks of \(x^\perp\cap S\). Cohen and Cooperstein obtained in particular the following point-line geometries: the Grassmannians \(A_{n,d}(D)\) and \(A_{2s-1,s}(D)/\langle \sigma \rangle\) where \(D\) is a division ring and \(\sigma\) is a polarity of the underlying building of Witt index at least five, the exceptional coset geometries \(E_{6,1}(K)\) or \(E_{7,1}(K)\) where \(K\) is a field, and homomorphic images \(\Gamma/A\) of a half-spin geometry \(\Gamma\) by a subgroup \(A\) of Aut\((\Gamma)\) such that for every element \(\alpha\circ A\) and point \(p\) such that \(p\neq p^\alpha\) the distance from \(p\) to \(p^\alpha\) is at least 5. In the paper under review the authors build on this result. They characterize a broad class of strong parapolar spaces by the properties of having all symplecta of rank at least 3, having finite singular rank, and possessing the weak hexagon property on 6-circuits in the collinearity graph of the geometry. They obtain the geometries listed above except \(E_{6,1}(K)\); furthermore, \(A_{2s-1,s}(D)/\langle \sigma \rangle\) is not possible under certain weak additional assumptions. In order to prove their main theorem the authors formalize the condition in the gap phenomenon as follows. \((CC)_k\) For every point \(x\) not in a symplecton \(S\) of \(\Gamma\), the singular subspace \(x^\perp\cap S\) either has projective rank at most \(k\), or is a maximal singular subspace of \(S\). \(k=0\) is just the Cohen-Cooperstein condition. They show that \((CC)_k\) is satisfied in certain rather natural situations, thus establishing a perfectly natural background context for the gap. More precisely, a parapolar space \(\Gamma\) satisfies \((CC)_{k-2}\) if for any two distinct symplecta \(A\) and \(B\) intersecting in a projective space of rank at least \(k\geq 3\) and for each point \(x\in A\setminus(A\cap B)^\perp\) there exists a point \(v\in x^\perp\) such that \(v^\perp\cap B\) and \(\Gamma_3(v)\cap A\cap B\) are nonempty where \(\Gamma_3(v)\) is the set of all points of distance 3 from \(v\). (This condition is extended to \(k=2\) but requires an aditional assumption.) In a number of steps and under varying assumptions the authors show that \((CC)_0\) is satisfied and then apply the Cohen-Cooperstein theorem.