The idea the author seems to try to hint at is what i would call a Functor if we were talking about Categories:

- Language games have structure

- Structures can be compared by correspondences/functions

- A particular concept in one language game therefore could be transported along such a correspondence into another language game.

- We don't have a good correspondence between the language games of biology and physics that allows us to transport "life" from biology to physics.

- Or do we!???

P.S. It occurs to me that my analogy with categories and functors is itself a transport between the language game of category theory and the language game of language games.