When $a$ and $b$ are elements of sets, the **pairing** of $a$ and $b$ is the ordered pair $(a,b)$.

It is natural to extend this to generalised elements in any category with binary products.

For products of higher arity, one can say **tripling**, **quadrupling**, etc, or just **tupling**.

Let $X$ and $Y$ be objects of some category $C$, and suppose that the product $X \times Y$ exists in $C$.

Let $G$ be some object of $C$, and let $a\colon G \to X$ and $b\colon G \to Y$ be morphisms of $C$. Then, by definition of product, there exists a unique morphism $(a,b)\colon G \to X \times Y$ such that the obvious diagrams commute.

If we think of $a$ and $b$ as $G$-indexed elements of $X$ and $Y$, then $(a,b)$ is a $G$-indexed element of $X \times Y$.

If $C$ is the category of sets and $G$ is the point, then $a$ and $b$ are simply elements, in the usual sense, of $X$ and $Y$; then $(a,b)$ is an element of $X \times Y$, the usual ordered pair $(a,b)$.

If $Y$ and $G$ are each $X$, with $a$ and $b$ each the identity morphism on $X$, then the pairing $(id_X,id_X)$ is the diagonal morphism $\Delta_X\colon X \to X^2$.

Since taking products (when these always exist) is a functor, we can apply this operation to any two morphisms. That is, if $a\colon G \to X$ and $b\colon H \to Y$ are morphisms in a category $C$, and if the products $G \times H$ and $X \times Y$ exist, then we have a morphism $a \times b\colon G \times H \to X \times Y$. This is *not* the pairing $(a,b)$, for which the source is always $G$.

A pairing is the composite of a product and a diagonal morphism:

$G \overset{\Delta_G}\to G \times G \overset{a \times b}\to X \times Y ;$

conversely, a product is a pairing of two composites:

$\array {
G \times H \to G \overset{a}\to X
,\\
G \times H \to H \overset{b}\to Y
.}$

If $G$ and $H$ are each terminal, however, then $(a,b)$ and $a \times b$ are the same global element of $X \times Y$. Thus, both product morphisms and pairings are generalisations of ordered pairs in Set.

Last revised on November 1, 2011 at 07:19:32. See the history of this page for a list of all contributions to it.