Hermitian symmetric spaces, cycle spaces, and the Barlet-Koziarz intersection method for construction of holomorphic functions (Q5929459)

This paper by a well-known expert, studies Hermitian symmetric spaces, complex flag manifolds, flag domains, and cycle spaces. Its goal is to give a new proof (which is much more direct and potentially more general than the earlier proofs) that a certain linear cycle space \(M_D\) is a Stein manifold. Let \(G_0\) be a real semisimple Lie group, \(G\) its complexification, \(Q\) a parabolic subgroup of \(G\), \(X=G/Q\) the corresponding complex flag manifold, and \(D=G_0(x_0)\) an open \(G_0\)-orbit on \(X\), \(Y_0=K_0(x_0)=K(x_0)\), where \(K_0\) is a suitable maximal compact subgroup of \(G_0\), \(K\) is the complexification of \(K_0\). Then \(Y_0\) is a complex flag manifold sitting as a maximal compact subvariety of \(D\). The linear cycle space \(M_D\) of \(D\) is, by definition, the connected component containing \(Y_0\) of the space \(\{gY_0:g\in G, gY_0\subset D\}\) of translates of \(Y_0\) by \(G\) that lie in \(D\). Then \(M_D\) is an open submanifold of the complex manifold \(M_X=\{gY_0:g\in G\}\cong G/E\) of all translates of \(Y_0\) by \(G\), where \(E=\{g\in G:gY_0=Y_0\}\) a closed subgroup of \(G\) is the stabilizer of \(Y_0\) in \(G\). The theorem that \(M_D\) is a Stein manifold was proved earlier gradually in the following papers. \textit{R. O. Wells} and \textit{J. A. Wolf} [Ann. Math. (2) 105, 397-448 (1977; Zbl 0448.32015)] stated the result and gave a proof (in slightly more restricted context than the present one), which was found later on problematic, then in [Ann. Math. (2) 136, 541-555 (1992; Zbl 0771.32016)], \textit{J. A. Wolf} proved that \(M_D\) is Stein if \(D\) is a so-called measurable orbit. Here measurability of \(D\) is a special condition meaning that \(D\) carries a \(G_0\)-invariant positive Radon measure (there are also algebraic conditions equivalent to that). Since nonmeasurable orbits \(D\) also occur in practical situations, it was desirable to extend the result that \(M_D\) is Stein to nonmeasurable orbits \(D\), too. This was done by \textit{J. A. Wolf} [Math. Res. Lett. 2, 179-191 (1995; Zbl 0856.22007)]. The paper under review proves that \(M_D\) is Stein by a direct argument (without separating cases whether \(D\) is measurable or not) relying on (an adaptation of) a new method of \textit{D. Barlet} and \textit{V. Koziarz} [Math. Res. Lett. 7, 537-549 (2000; Zbl 0978.32009)], a paper which directly precedes in the journal and is also reviewed in this volume. Denote by \(s\) the complex dimension of \(Y_0\), and let \({\mathcal C}_s^0(D)\), \({\mathcal C}_s^0(X)\) be the irreducible component of the Barlet cycle space of \(D\), \(X\) that contains \(M_D\), \(M_X\), respectively. The Barlet-Koziarz intersection method is adapted to show that if there is a closed complex submanifold \(X'\subset X\) which satisfies a certain transversality condition \((2.11)\), then \({\mathcal C}_s^0(D)\) is holomorphically convex. The condition \((2.11)\) reads: (i) \(\text{codim}(X'\subset X)=s\), (ii) \(x_0\in X'\cap Y_0\), and \(X'\) is transversal to \(Y_0\) at \(x_0\), (iii) \(X''=X'\cap D\) is a Stein manifold, and (iv) \(X'\) meets every \(G_0\)-orbit on the boundary of \(D\) in \(X\). Proposition 2.13 (still assuming a transversal \(X'\) with \((2.11)\)) then transfers the Stein property from (the component of) the Barlet cycle space \({\mathcal C}_s^0(D)\) to the linear cycle space \(M_D\). Before constructing a transversal \(X'\) satisfying \((2.11)\) in the general case, the author discusses in \S 3 the example of the Grassmann manifold \(X=\text{Gr}(p,q)\), \(G_0=\text{SU}(p,q)\), in which a classical calculation provides a suitable \(X'\), and which example, in fact, led the author to the general construction of \(X'\) in \S 4 and \S 5. While \S 3 is completely self-contained, \S 4 and \S 5 draw upon a detailed knowledge of orbit structure and Cayley transforms theory (to which the author is a main contributor). It is also stated that a slightly more precise result than the Stein condition for \(M_D\) appeared in [Math.~Ann. 316, 529-545 (2000; Zbl 0963.32015)] by \textit{J. A. Wolf} and \textit{R.~Zierau}. Despite its technical nature, the paper is very readable and informative.