Comment by jesuslop

I'm reading and liking his book just called "Trees" (some bites), it is about group actions on trees, written in the '80. It talks about the modular group among others, the tree action also being used in Shai Haran work on the "real" prime (if the AIs let us dream the old way about classic problems). Serre et al graph of groups idea is very categorical. He could well have been present in the very seminars when Grothendiek was breweing what is now called the Grothendiek construction (absurdly, since G was a serial constructor). That wraps that graph of groups thing.