Character formulas for some classes of atypical \(gl(m+n\epsilon)\)- and \({\mathfrak p}(m)\)-modules (Q1110645)

In the present paper the authors discuss character formulas for some particular modules of the general linear Lie superalgebra \(gl(m+n\epsilon)\), more often denoted by gl(m/n) or A(m-1,n-1). This is a very important and interesting problem, solved by V. Kac for the so- called typical modules but remaining open for the atypical modules, for which the methods used in representation theory of simple Lie algebras over the field of complex numbers have proven to be insufficiently adequate so far. There have been several publications on this subject, but none has produced a complete solution to the problem. In this respect, the present paper does not differ from others: again only a set of particular modules are analysed. One of the main results (Theorem 2) gives the character of a highest weight module V(\(\lambda)\) for nondegenerate \(\lambda\) (the definition of a nondegenerate highest weight is rather technical). Unfortunately, this character formula still contains the character of ``the annihilator in the supersymmetric algebra of a highest weight vector \(v_{\lambda}''\) in its expression, and therefore the given formula is very hard to use or evaluate. In fact, it turns out that the calculation of the character of the annihilator is a problem almost as hard as the calculation of the character of the module itself. Nevertheless, the authors succeed in determining this character for some very particular situations. These cases are special in the sense that the matrix of atypicality values, i.e. the matrix of numbers \(<\lambda | \epsilon_ i-\delta_ j>\) (where \(\epsilon_ i-\delta_ j\) are the odd roots), satisfies some particular properties (such as having a block structure etc.). It is for these special cases, which are rather peculiar and not necessarily the cases one is interested in, that the authors obtain a character formula. Anyhow, the paper is a step forward towards the solution of a very difficult problem. The proofs of the authors' results will be published elsewhere.