Isotopes of prime \((-1,1)\)- and Jordan algebras (Q619327): Difference between revisions

The notion of an isotope was introduced in [\textit{A. A. Albert}, Ann. Math. (2) 43, 685--707 (1942; Zbl 0061.04807); 43, 708--723 (1942; Zbl 0061.04901)]. Suppose that algebras \(A\) and \(A_0\) have a common linear space on which right multiplication operators \(R_x\) and \(R_x^0\) are defined (on \(A\) and \(A_0\) respectively). Then \(A_0\) is called an isotope of \(A,\) if there exist invertible linear operators \(\varphi, \psi\) and \(\xi\) such that \(R_x^0=\varphi R_{x\psi}\xi\). Isotopes, such that \(\varphi, \psi\) and \(\xi\) are operators of multiplication by fixed elements of a given algebra, are called \textit{c}-isotopes. This article is devoted to adjoint commutator algebras and adjoint Jordan algebras of \textit{c}-isotopes of prime strictly \((-1,1)\)-algebras. Set \([x,y,\dots ,z]=[[x,y,\dots ],z]\). It is shown that for each of the commutators \([x_0,y,y]\), \([x_0,y,y,x_1]\), \([x_0,y,y,x_1,x_2]\), \([x_0,y,y,x_1,x_2,x_3]\) starting with the second one, there exists a \textit{c}-isotope of a prime \((-1,1)\)-algebra on which it is an identity, but its predecessor is not. Let \(A\) be a prime nonassociative \((-1,1)\)-algebra. An identity which holds in the adjoint Jordan algebra \(A^+\) and fails in some its \textit{c}-isotope is indicated. Many other interesting facts about prime \((-1,1)\)-algebras are obtained. Moreover, using isotopes the author produces new central polynomials of a free strictly \((-1,1)\)-algebra.