Monoid deformations and group representations (Q1295623)

Deformations of a finite monoid \(M\) with zero are introduced. First, the generic monoid \(M(t)\) is defined using an associated operation \(\xi\colon M\times M\to \mathbb{C}(t)M\), where \(\mathbb{C}(t)\) is the field of rational functions. Namely, for \(a,b\in M\) we assume that \(\xi(a,b)=ab\) if \(ab\neq 0\) and, if \(ab=0\), then \(\xi(a,b)=f(t)u\), where \(f(0)=0\) and \(u\in MaM\cap MbM\). If \(c\in\mathbb{C}\) is not a pole of any of the coefficients of \(M(t)\), then \(M(c)\) is called a deformation of \(M\). So \(M(0)=M\) in particular. Such monoids are studied from the point of view of the representations of finite reductive groups \(G\) and their connections with the associated universal monoids of Lie type \(\mathcal M\), constructed on \(G\) by \textit{M. S. Putcha} and \textit{L. E. Renner} [Trans. Am. Math. Soc. 337, No. 1, 305-319 (1993; Zbl 0787.20039)]. It is known that \(\mathcal M\) can also be constructed with the modular Steinberg representation of \(G\). A bigger monoid \(\widetilde{\mathcal M}\) is defined in the paper under review, and it is shown that a deformation \(\widetilde{\mathcal M}(q)\) of \(\widetilde{\mathcal M}\) is contained within the ordinary Steinberg representation. Imposing a semigroup structure on Chevalley's big cell \(B^-B\), and deriving from it a new description of the Steinberg representation, is one of the key ingredients of the proof.