Readable, but selective account of monetary history with communitarian bias. This "historical" approach to money is always distorted by the fact that "quantitatively" for most of human history and cultures, monetary phenomena were linked to cult-based and tribal social relations, whereas "qualitatively" the smaller subset of catallactic monetary phenomena (cooperation between foreigners) are materially and, eventually, culturally much more important. So, this communitarian bias boils down to Graeber's anarcho-primitivist bias. I appreciate him for being much more straight-forward about it than most history romantics.