Standard monomials for wonderful group compactifications. (Q876337)

The author extends standard monomial theory to the wonderful compactification \(X\) of a semisimple group \(G\) of adjoint type. Recall that \textit{R. Chirivì} and \textit{A. Maffei} have already constructed standard monomials for the more general situation of a wonderful compactification of a symmetric space [J. Algebra 261, No. 2, 310-326 (2003; Zbl 1055.14052)]. We fix a dominant weight \(\lambda\) and the corresponding line bundle \(\mathcal L_\lambda\) on \(X\). Then Chirivì and Maffei provide a basis of \(H^0(X,\mathcal L_\lambda)\) consisting of `standard monomials' with certain properties. The author shows that in the present situation one can do more. First of all, the standard monomials are shown to behave well with respect to restriction to \(B\times B\)-orbit closures, not just \(G\times G\)-orbit closures. Recall that there are finitely many \(B\times B\)-orbits and that they have been classified by \textit{T. A. Springer} [J. Algebra 258, No. 1, 71-111 (2002; Zbl 1110.14047)]. There are degrees of freedom in the construction of Chirivì, Maffei and these the author exploits to arrange more properties familiar from the classical standard monomial theory for flag varieties. The basis of \(H^0(X,\mathcal L_\lambda)\) is indexed by LS-paths again. And if \(Z\) is a \(G\times G\)-orbit closure, or more generally a \(B\times B\)-orbit closure, then the standard monomials that do not vanish on \(Z\) form a basis of \(H^0(Z,\mathcal L_\lambda)\). They can be characterized combinatorially. The many ingredients that are needed in the proof are explained clearly.