The seven areas covered are Nubia and the middle Nile (nearly 5000 years: Kerma, Napata, Meroe, and the Christian and Islamic eras), the Ethiopian highlands (the "pre-Aksum", Aksum, and medieval Christian periods), the West African savannah (Senegal, the Inland Niger Delta, northern Nigeria, south-west Chad), the West African forest and fringes (sites in Ghana and Nigeria including Old Oyo, Benin City, and Igbo-Ukwu), the East African coast and islands (from Mogadishu through Kilwa to the Comoro islands and Madagascar), Great Zimbabwe and related sites (the Zimbabwe Plateau), and the Upemba Depression and the Interlacustrine Region (two areas in Central Africa, thousands of kilometres apart, where fieldwork happens to have been carried out).
Each of the chapters follows the same format. A brief introduction is followed by a look at "geographical location and environmental factors", covering ecology, climate, soils, diseases, and barriers to movement. The core of each chapter is a section "sources of information", typically around twenty pages long. This describes the historical sources in just a page or two, then surveys the archaeological record in much greater depth. Here Connah presents details of key excavations and sites, with a nice selection of site plans, maps, chronologies, and halftones of artifacts and landscapes. His approach follows the modern trend towards settlement studies and away from the narrower traditional focus on monuments and elite sites.
Six shorter sections in each chapter then consider the key themes of subsistence economy, technology, social system, population pressures, ideology, and external trade, focusing on the role of these factors in urbanization and state formation. Connah generally describes the theories of others rather than offering his own, and his conclusions are almost always tentative and qualified. In the chapter on the West African forest, for example, he writes "there is ... good reason to suspect there would have been a growth in functional specialization", "it seems likely that by the early second millennium AD the subsistence economy of the rainforest was able to produce a surplus", and "it is possible that quite localized population pressure was one of the factors which led to an increasing elaboration of social hierarchies" (to pick just a few examples). While this may frustrate those after simple answers, it seems like a wholesome approach given the limited evidence and the patchiness of the archaeological fieldwork.
Nor does Connah make any attempt to fit a single theoretical framework across the different areas. The overall introduction touches on theories of states, notably that of Haas, and a brief six page conclusion looks at "common denominators", but the emphasis is on the diversity of African environments and the resulting variety in the development of cities and the formation of states. "[T]he physical evidence for African social complexity [merits] far more attention from archaeologists than it has yet had and might have a far greater role in the formation of social theory than has yet been the case".
African Civilizations is solidly referenced and has a thirty five page bibliography, but some further reading suggestions would have been nice: as it is the reader intrigued by any of the specific regions covered isn't offered any obvious starting points in the broader literature.