Templo Mayor

On 21 February 1978 a crew from the electricity company, digging behind the cathedral in the centre of Mexico City, struck a carved stone disc more than three metres across. It showed a woman dismembered: Coyolxauhqui, the moon, cut apart by her brother Huitzilopochtli at the foot of the mountain of Coatepec. The disc lay where the myth said it should lie, at the base of the stair of the Templo Mayor, and it settled the location of the temple that the Spanish had levelled and built over four and a half centuries earlier.

What followed was not a single excavation but a long one. The temple had been enlarged at least seven times, and Mesoamerican practice was to build the new stage over the old rather than to clear it, so each rebuilding sealed the one beneath. The result is a stack of datable envelopes, and buried in them are more than two hundred offering caches: worked stone, sacrificial knives, sea creatures brought from both coasts, jaguars, eagles, and human remains.

The building carried two shrines on one platform. The southern one was for Huitzilopochtli, the tutelary god of the Mexica and the god of the sun and of war. The northern one was for Tlaloc, far older than the Mexica themselves, the god of rain and of the mountains that hold it. Two gods on one summit is a statement about what the city thought it was for. The site museum opened in 1987, and excavation has continued outward into the surrounding streets ever since, under a programme that works around the standing colonial city rather than through it.

Writing about sacrifice at this site means keeping three questions apart, because they are routinely collapsed into one and they have three different answers.

The first is whether it happened. Here the site answers plainly and in its own voice. Human remains are present in dated, stratified offerings, some of them decapitated crania, some of them children associated with the shrine of the rain god. None of this depends on a Spaniard having written anything down. The excavated deposits are the evidence, and the practice is attested.

The second is at what scale. Here the site answers much less, and honestly so. An offering cache tells you what was placed in it, not how often anything happened across a century in a city of that size. The totals that circulate come from colonial-era narratives, and the arithmetic in them is not arithmetic that anyone performed. Modern estimates differ by more than an order of magnitude, and that spread is a real state of knowledge rather than a failure to look it up.

The third is whether the source that tells you is trustworthy, and this question has to be asked separately of each source. A conquistador writing a justification of his own campaign, a friar compiling testimony from Nahua informants a generation after the conquest under Church supervision, and a stratified deposit under a colonial street are three different kinds of witness with three different failure modes. Treating them as one body called the sources is how a figure invented in a polemic ends up quoted as a measurement.