Morphisms of polar spaces (Q1875909)

A polar space \(P = (P, {\mathcal L}, I)\) is a pair of sets \(P\) called points and a collection \(\mathcal L\) of subsets of \(P\) called lines, where \(P\) has at least 2 elements, together with an incidence relation \(I\) such that for each point \(p\) and each line \(l\) the point \(p\) is incident with either one or all points of \(l\). For a polar space the authors require an additional axiom, namely that any two distinct lines intersect in at most one point. Thus, polar spaces in this sense are partially linear polar spaces. This point of view leads to the theory of paraprojective spaces. The authors define a paraprojective space \((P,\ast)\) as a non-empty set \(P\) and a map \(\ast : P \times P \to {\mathcal P}(P)\) into the power set of \(P\) satisfying the following axioms: (P1) \(a \ast a = \{ a \}\), (P2) if \(a \ast b \neq \emptyset\) then \(b \in a \ast b\), and (P3) if \(a \in b \ast p\), \( p \in c \ast d\), \(a \neq c\), \(b \ast c \neq \emptyset\) and \(b \ast d \neq \emptyset\), then \((a \ast c) \cap (b \ast d) \neq \emptyset\). An element of \(P\) is called a point. If \(a \ast b \neq \emptyset\) and \(a \neq b\), then the set \(a \ast b\) is called a line of \(P\). A morphism \(g\) between two paraprojective spaces \(P_{1}\) and \(P_{2}\) is a partial map \(g : P_{1} \setminus E \to P_{2}\) such that \(E\) is a subspace of \(P_{1}\), if \(a,b \not\in E\), \(c \in E\), and \(a \in b \ast c\), then \(g(a) = g(b)\), and if \(a,b,c \not\in E\) with \(a \in b \ast c\) then \(g(a) \in g(b) \ast g(c)\). It is shown that paraprojective spaces and their morphisms form a category with polar spaces as a full subcategory. The main result of the paper under review constitutes a splitting theorem of morphisms of polar spaces, i.e. any such morphism can be decomposed into a morphism of nondegenerate polar spaces and a morphism of projective spaces.