On the free Jordan algebras

From MaRDI portal
Publication:2020405



Abstract: A conjecture for the dimension and the character of the homogenous components of the free Jordan algebras is proposed. As a support of the conjecture, some numerical evidences are generated by a computer and some new theoretical results are proved. One of them is the cyclicity of the Jordan operad.


For a natural number \(D\), let \(J(D)\) be the free (nonunital) Jordan algebra on \(D\) generators over a field of characteristic \(0\). This is \(\mathbb{Z}\)-graded by assigning degree \(1\) to the generators, so that \(J(D)=\bigoplus_{n\geq 1}J_n(D)\). The general linear group \(GL(D)\) acts naturally on \(J(D)\) by automorphisms preserving the grading. The paper under review presents an intriguing conjecture for the character (and hence, in particular, for the dimension) of this action of \(GL(D)\) on each homogeneous component \(J_n(D)\). This Conjecture 1 tells us that the classes of the \(GL(D)\)-module \(J(D)\) and of its Lie algebra of inner derivations inside the representation ring \(\mathcal{R}(GL(D))\) appear in a explicit technical way. Some numerical evidences for Conjecture \(1\) are obtained using the free software SAGE together with classical results by Cohn, Glennie, Shirshov, MacDonald and others. Moreover, two more conjectures are given related to the homology of the Lie algebra \(\mathfrak{sl}_2J(D)\) obtained from \(J(D)\) by means of the functorial version, due to \textit{B. N. Allison} and \textit{Y. Gao} [Can. J. Math. 48, No. 3, 449--482 (1996; Zbl 0861.17011)], of the classical Tits-Kantor-Koecher construction. Conjecture 2 asserts that the trivial and adjoint isotypic components: \(H_k(\mathfrak{sl}_2J(D))^{\mathfrak{sl}_2}\) and \(H_k(\mathfrak{sl}_2J(D))^{ad}\), of the homology of the Lie algebra \(\mathfrak{sl}_2J(D)\), are trivial for \(k\geq 2\). This Conjecture 2 is shown to imply Conjecture 1 and some results supporting the conjecture are given too. \par Finally, it is proven that there is an action of the symmetric group of degree \(D+1\), extending the natural action of the symmetric group of degree \(D\) on the multilinear part of \(J(D)\). This is used to weaken Conjecture 2 to a new Conjecture 3, asserting that \(H_k(\mathfrak{sl}_2J(D+1))^{\mathfrak{sl}_2}\) is trivial for all \(k\geq 1\). Again Conjecture 3 is shown to imply Conjecture 1.











This page was built for publication: On the free Jordan algebras

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2020405)