I recently finished reading Philip Jenkin's book The Lost History of Christianity: The Thousand-Year Golden Age of the Church in the Middle East, Africa, and Asia--and How It Died. All told an interesting book, even if I wasn't a big fan of the author's attitude at time.
Anyways, one of the factors which Jenkins attributed the disappearance of the church in Africa made me think back to arguments that folks like Tim Keller have made regarding the importance of cities to the church:
Where the African Church failed was in not carrying Christianity beyond the Romanized inhabitants of the cities and the great estates, and not sinking roots into the world of the native peoples. Like most regions of the Western empire, such as Gaul and Spain, Africa was divided between Latin-speaking provincials and old-stock natives, who spoke their ancient languages - in this case, varieties of Berber. Unlike these other provinces, though, the African church had made next to no progress in taking the faith to the villages and the neighboring tribes, nor, critically, had they tried to evangelize in the local languages. This would not have been an unrealistic assumption, in that already by the fourth century missionaries elsewhere were translating the scriptures into Gothic, and Hunnie languages followed by the sixth century. Evidence of the neglect of the countryside can be found in the letters of Saint Augustine, by far the best known of African bishops, whose vision was sharply focused on the cities of Rome and Carthage; he expressed no interest in the rural areas or peoples of his diocese. (p. 229)
Basically, both/and rather than either/or. (Although Keller wouldn't fit in a hypothetical "either/or" camp).
