The join of categories can also be described to be the cograph of the unique profunctor sending all objects to the terminal set (the definition speaks for itself).
In terms of adjoint functors
Consider the inclusion of the boundary of the standard 1-simplex, as a functor between the discrete category with two elements and the walking arrow . It induces a functor
which admits a right adjoint. This right adjoint is precisely the bifunctor , once we noticed that the category comes naturally equipped with an arrow induced by (bi)functoriality of , starting from the canonical arrows to the terminal category.
Proof
It is quite clear that is defined by sending to the pair of categories . The bijection
is now rather obvious, since any functor determines a functor and viceversa.
Properties
If the small categories are two posets, their join consists of their ordinal sum;
The monoidal structure induced on by is not symmetric (if it was, then the right cone of would be equivalent to the left cone, which is blatantly false);
The monoidal structure induced on by is not closed, since the functor does not preserve colimits.
The functor is a left adjoint, and similarly is the functor ; see Joyal, Ch. 3.1.1-2 for a detailed description.
Examples
The cone below a category is the join . The cone above is the join .