On the (non)existence of symplectic resolutions of linear quotients (Q2396645)

The main goal of the paper is to classify all finite subgroups \(G\) of \(\mathrm{Sp}(V)\) for which \(V/G\) admits a symplectic resolution. Let \(V\) be a symplectic vector space and \(G \subset \mathrm{Sp}(V)\) a finite group. The authors study singularities of the quotient \(V/G\). If there exists a projective resolution of singularities \(X \to V/G\) such that \(X\) is a symplectic manifold, then they say that \(V/G\) admits a projective symplectic resolution. Consider \(V\) as a symplectically irreducible representation of \(G\). The authors classify all such pairs \((V,G)\), which admit a projective symplectic resolution, in the case \(\dim V \neq 4\), except for four singularities, occurring in dimensions at most 10, for which the question remains open. Section 2 contains the definition of symplectic variety and symplectic resolutions, and some criteria are given for the (non)existence of projective symplectic resolutions. In Section 3, they recall the Kleinian group, in Section 4, Cohen's classification of symplectic reflection groups. Section 5 deals with two general criteria for the non-existence of projective symplectic reflections for the group \(G(K,H,\alpha)\) with \(H \neq {1}\). In the next section, they prove that the symplectically imprimitive, irreducible symplectic reflection groups of the above type satisfy at least one of the criteria. Section 7 gives the following result: If \(n>2\), then the symplectic quotient \(C^{2n}/G_n(K,H)\) admits a projective symplectic resolution if and only if \(K=H\). The paper ends with some open questions.