No label defined (Q4938570)

Given a partition \(\mu\vdash n\), the Garsia-Haiman module \(M_\mu\) is the linear span of derivatives of a certain bihomogeneous polynomial \(\Delta_\mu(x,y)\) in the variables \(x_1,\ldots,x_n,y_1,\ldots,y_n\). \textit{A. M. Garsia} and \textit{M. Haiman} [J. Comb. 3, No. 2, 541-600 (1996; Zbl 0851.05099)] have conjectured that the dimension of this module is \(n!\), and \textit{A. M. Garsia} and \textit{M. Haiman} have shown that this conjecture implies the \(C=\widetilde H\) conjecture [Proc. Natl. Acad. Sci. USA 90, No. 8, 3607-3610 (1993; Zbl 0831.05062)] which, in turn, implies \textit{I. G. Macdonald}'s conjecture [Symmetric functions and Hall polynomials, Oxford (1998; Zbl 0899.05068)] that the \(q,t\)-Kostka coefficients, \(K_{\lambda\mu}(q,t)\), are polynomials in \(q\) and \(t\) with nonnegative integer coefficients. Using this connection, information about the module \(M_\mu\) can be translated into identities for the Macdonald polynomials. In this paper, the authors study the spaces \(M_\nu\) as \(\nu\) ranges over the immediate predecessors of \(\mu\). Using heuristic arguments to explain the structure of these predecessor modules (the ``science fiction'' in the title), the authors are led to a number of new identities for the Macdonald polynomials, some of which can be proven, others have been checked numerically for significantly large examples. Included in these results is a representation theoretic interpretation for some of the observed symmetries in the computed tables of \(q,t\)-Kostka coefficients. For the entire collection see [Zbl 0922.00020].