Colength growth functions of nonassociative algebras (Q2037677)

Let \(F\) be a field of characteristic zero. We denote by \(F\{X\}\) the absolutely free algebra over \(F\) with a countable set \(X=\{x_1,x_2,x_3,\ldots\}\) of generators. For every \(n\geq 1\), let \(P_n\) denote the space of multilinear polynomials of \(F\{X\}\) in the first \(n\) variables \(x_1,x_2,\ldots, x_n\). Given an \(F\)-algebra \(A\), let \(\text{Id}(A)\) be the \(T\)-ideal of \(F\{X\}\) of polynomial identities of \(A\). It is known that the ideal \(\text{Id}(A)\) is completely determined by its multilinear components \(\text{Id}(A)\cap P_n\) for \(n\geq 1\). The symmetric group \(S_n\) acts on \(P_n\) by permuting variables and the space \(P_n\cap \text{Id}(A)\) is invariant under this action. Thus, the group \(S_n\) acts on \(P_n(A) = \frac{P_n}{\text{Id}(A)\cap P_n}\) as well. If \(\chi(P_n(A))\) is the \(S_n\)-character of \(P_n(A)\), then we can write \[\chi(P_n(A))= \sum_{\lambda\vdash n}m_\lambda \chi_\lambda\] where \(\chi_\lambda\) is the irreducible character corresponding to the partition \(\lambda\) of \(n\) and \( m_\lambda\geq 0\) is the multiplicity of \(\chi_\lambda\). An invariant associated to the identities of \(A\) is the colength \(l_n(A)= \sum_{\lambda\vdash n}m_\lambda\) defined for all positive integers \(n\). Thus, \(l_n(A)\) is the total number of irreducible components appearing in \(\chi(P_n(A))\). The main result of the paper is about a study of colengths of an algebra \(A(w)\), the algebra associated with an infinite binary word \( w = w_1w_2\ldots\), where all \(w_i\) equal 0 or 1. This algebra is generated by two elements \(a\) and \(b_0\) with basis \(\{a,b_0,b_1,\ldots\}\) and multiplication defined as follows: If \(w_k=1\) then \(b_k=ab_{k-1}\) and if \(w_k=0\) then \(b_k=b_{k-1}a\) for \(k\geq 1\), and all the remaining products are zero. A subword \(u\) of \(w\) is called a subword of the first type if it occurs in \(w\) only after \(0\) or only after \(1\). If \(u\) occurs in \(w\) both after \(0\) and after \(1\), then it is called a subword of the second type. The main result of the present paper is the following result. Theorem 1. Let \(w = w_1w_2\ldots\) be an infinite binary word. Then \(l_n(A(w)) = 2k_{n-2}+ 3t_{n-2}\), where \(k_{m}\) and \(t_{m}\) are, respectively, the numbers of proper subwords of length \(m\) of the first and second types in \(w\).