“The importance of Teotihuacan has been exaggerated”: thus I ended a review in Antiquity some 23 years ago (Hammond Reference Hammond1993: 950). George Cowgill's excellent, synoptic and sometimes argumentative new book confirms that view, although he does not necessarily share it.

Teotihuacan, ‘city of the gods’ to the Aztecs, and ‘Los Pirámides’ to Mexico City tour operators in reference to the immense Sun and Moon pyramids that dominate the site, was, from around the first to the sixth centuries AD, perhaps the largest pre-Columbian community in Mesoamerica. The city was based on a grid-like pattern of streets enclosing multi-apartment residential compounds, and divided into quadrants by the ‘Street of the Dead’, running south from the Moon pyramid, and twin avenues going east and west from the central giant courtyards of the Ciudadela and Great Compound. This layout gives the semblance of a master plan unfulfilled: the densest and oldest settlement is in the north-west, with sparser habitation and fewer ceremonial structures south of the centreline.

The ruins have never been forgotten: the Aztecs mined them for artefacts to put into ritual caches in their capital, Tenochtitlan, eight centuries after Teotihuacan fell into semi-desuetude following internal or external disruption around AD 550, and Spanish colonial artists made them examples of the picturesque. The first excavation took place as early as 1675, with serious investigations beginning under Manuel Gamio and Ignacio Marquina a century ago and hardly ceasing since. One of the major efforts was the Teotihuacan Mapping Project (TMP) from 1962 onwards, directed by René Millon and with George Cowgill as one of his senior collaborators (Millon Reference Millon1973; Millon et al. Reference Millon, Drewtt and Cowgill1973). The combination of intensive mapping of some 30km², the systematic collection of more than 830000 sherds, 60000 ceramic objects and 230000 obsidian fragments from more than 5000 survey tracts, supplemented by pioneering computer analyses of the vast database thus accumulated, plus judicious excavations, has given an unparalleled picture of a Prehispanic city, which Cowgill is more qualified than anybody to summarise.

He does it within the remit of the Cambridge University Press's ‘Case Studies in Early Societies’ series, which “aims to introduce students to early societies that have been the subject of sustained archaeological research” and to “demonstrate a contemporary method of archaeological analysis in action [. . .] tracing long-term developments” (p. iii). Cowgill says “I have tried to tell a story about all aspects of Teotihuacan society, including technology, politics, economics, environmental interactions, religion, and what we can infer about the texture of life—both everyday and on exceptional occasions”, dealing with “inconspicuous commoners as well as the elite and powerful [. . .] for a society that had no full-blown writing, all this is a daunting challenge” (p. 2). Even though specialists will therefore find a lot of familiar material at the beginning, and students may find the occasional side-swipes at colleagues’ interpretations puzzling, he succeeds rather well.

There are ten chapters. ‘Preliminaries’ defines terms and research questions, and retells the history of research there; the chronological chart (p. 10, see also pp. 245–46) offers a developmental history earlier than some previous estimates. ‘Situating Teotihuacan’ outlines the Mesoamerican environment, and then ‘Urbanism begins in Central Mexico’ shows what was going on in the first millennium BC at places such as Cuicuilco. The latter was a putative rival, formerly believed to have been wiped out by the eruption of Xitle in the first century BC, stimulating the rise of Teotihuacan. Cowgill shows how new volcanic and ceramic chronologies throw this into doubt, while noting that a quasi-coeval eruption of Popocatépetl may well have driven people northwards out of the lake valley of Mexico and into the Teotihuacan pocket.

The chapter ‘Teotihuacan takes off’ shows the small size of Cuanalan-phase villages before 200 BC, followed, in the first century BC, by rapid Patlachique-phase growth in the north-western quadrant. Here, sherds were found over an area of 6–8km², with the first constructions at the Sun and Moon pyramids on its eastern margin, and Cowgill estimates a population rising to at least 20000 at >1.5 per cent per year under “strong, although possibly collective, leadership” (p. 57) and due to a combination of in-migration, successful warfare and trade, “perhaps compelling new religious ideas and very possibly the sheer luck of having a few unusually able and ambitious leaders”—the ‘great man’ theory of prehistory.

Those leaders remained resolutely anonymous throughout the city's history: even where we have élite tombs, there is no indication that the defunct were rulers. The parallel I suggested with the Indus civilisation remains valid (Hammond Reference Hammond1993: 948). Whoever ran Teotihuacan—king or committee—it grew in the Tzacualli phase (AD 1–100) to 60000–80000 inhabitants, at a slower rate of around 0.9 per cent per year and mainly by immigration, but with abundant labour to raise the Sun pyramid to almost (216m², 63m high) its present scale. The fifth chapter on ‘Teotihuacan supremacy’ shows some details of ceramics and architecture, with more in Chapter 6, ‘Great pyramids and early grandeur: 100–250 CE’, when the Sun pyramid grew a bit and the Moon pyramid underwent three successive phases of dramatic enlargement.

Excavations in the latter by Saburo Sugiyama have revealed numerous tombs, with preciosities from as far away as the Maya area; coeval development of the Feathered Serpent pyramid in the Ciudadela (“probably built as a new political center” p. 110), traced by Rubén Cabrera C.’s excavations, also revealed multiple burials: some of these were collective, many were apparently sacrificed warriors or presented as such, and some, from their dental ornamentation and cranial modelling, may in my view have been Maya. The amount of readable detail that Cowgill packs in here, aided by text-boxes on topics as diverse as stable isotopes, architectural styles and the curious absence of ball-courts at Teotihuacan, is truly impressive, although under-labelling of fig. 6.26 makes the status categories of residential compounds from Ian Robertson's research (pp. 126–27) undecipherable.

The discussion of ‘Teotihuacan at its height’ spans the period AD 25–550, when Xolalpan-phase ceramics document population expansion in the southern quarters of the grid (while the greatest density persists north of the Ciudadela to beyond the Moon pyramid). Cowgill estimates peak population to have been perhaps 85000, a long way down from Millon's 125000–200000; he then deals with variation in residential compound sizes—greater than previously understood—and possible social group correlations. ‘Neighborhoods’ such as the Oaxaca enclave, and the presence of West Mexicans nearby, and the ‘Enclave of the Merchants’, with its Gulf Coast and Maya links, suggest commercial relations over generations, rather than formal ‘embassies’. Some Teotihuacano presences in iconography, as well as artefacts at Maya cities such as Copan, Kaminaljuyu and Tikal, and also the recently discovered murals of Sufricaya, indicate impact at higher social levels too, with suggestions by some scholars that this city in Central Mexico had the power to change governments hundreds of kilometres to the east. Interaction between equals, as is more clearly suggested in the Oaxacan archaeological record, seems to me the more probable scenario.

In this chapter, Cowgill also runs through craft categories, including obsidian, pottery and ceramic objects such as figurines, ground stone and lapidary work, and then the “thorny problem” (p. 190) of rulership, which he is no more able to resolve than anybody else. In Chapter 8, he discusses ‘ideation and religion’, with useful summation rather than revelation, followed by an epigonal chapter on the decline of Teotihuacan after AD 550, where the meticulous data-collection of the TMP allows documentation of population shrinkage in the Metepec (AD 550–650) and Coyotlatelco (750–850) periods, and, conversely, the surprisingly large Late Aztec rural population after 1300. (196000 sherds collected by the TMP were from post-Teotihuacan periods.)

Cowgill proposes Teotihuacan hegemony over the basin of Mexico and parts of Morelos and other adjacent regions, but sensibly admits that Cholula to the east and its neighbours were probably independent states. The idea of a Teotihuacan imperium, popular a few years ago, must be replaced by one of a powerful regional state with long-distance commercial and diplomatic connections, and with an iconography reflecting powerful ideas that resonated as far as the eastern Maya lowlands: but from Olmec times onwards, Mesoamerica has been a mosaic of regions in constant contact, some flourishing as others declined. Even the Aztec ‘empire’ did not extend far beyond Central Mexico and its flanking coastal lowlands east to Xoconusco and Coatzacoalcos (it might have gone farther had it not been for Cortez). George Cowgill has given us a worthy summary of one of the most fascinating cultures in the prehispanic New World.