Algebraic tensor products revisited: axiomatic approach (Q2045262)

Taking the so-called `universal property' as the axiomatic starting point, this article offers a brief survey on the algebraic theory of tensor products of linear spaces. Recall that one of the many possibilities to define algebraic tensor products is as follows: Given two linear spaces \(X\) and \(Y\) (over the field \(\mathbb{F}\)), a pair \((T,\psi)\), where \(T\) is a linear space (over \(\mathbb{F}\)) and \(\psi: X \times Y \to T\) a bilinear mapping, is said to be a tensor product of \(X\) and \(Y\), whenever for each bilinear mapping \(\varphi:X \times Y \to E\) there is a unique linear mapping \(u: T \to E \) such that \(\varphi(x,y) = u \big(\psi(x,y)\big)\) for all \(x \in X\) and \(y \in Y\). Such objects exist, are (in a certain sense) unique, and usually denoted by \(X\otimes Y\). The content of this article is absolutely standard, and (in one way or the other) presented in many standard textbooks.