John Reader's Cities is more of a collection of essays about different aspects of urbanism than it is a coherent statement about the city in history. The fact that you don't, by any means, need to read this one cover to cover from front to back to get into it was a pleasant surprise. Things learned: the economic centrality of small-scale urban gardening in Cuba due to trade embargos, Nairobi is sinking because it was built on a river bed, Stockholm's utopically planned suburbs have actually offered more collectivity than its inhabitants have desired....
Reader's text is interesting and, while its heavy on the data and statistics, it's extremely readable. But, it is most pointedly NOT a history of "the city" (as some amazon reviewers were quick to point out and completely miss the point of the book). True, Reader starts in Mesopotamia, dabbles a bit in greece and rome and then moves on to more modern stories about cities, but each chapter or essay in the book remains a discrete statement about aspects of urbanism studied in different contexts: urban food economies, urban housing, migration into cities, the relationship between the urban and the rural, city planning, cities, contagation and disease, etc. And this approach underlies well one of Reader's main points: that the city is an ancient and integral part of human habitation, that it is by no means a "new" formation, and that we should not see our booming metropolises of today as some sort of apex of the development of urbanism.