The Ecological Hypothesis of Aztec Sacrifice

In 1977 Michael Harner published an argument in American Ethnologist that Mexica sacrifice, and specifically the ritual consumption of the bodies of the sacrificed, was best explained by ecology. The Valley of Mexico, on this account, had been stripped of large domesticable herbivores; the population had grown past what its protein supply could carry; and an institution that killed people and distributed their flesh was a response to that pressure, dressed in religion. Marvin Harris took the argument to a general readership the same year in Cannibals and Kings, and it travelled far beyond anthropology.

The argument had real appeal. It was materialist where the alternative explanations were theological, it was quantitative in form, and it appeared to explain a practice that had otherwise been described mainly with adjectives. It also had a shape that recurs in writing about non-European religion: the participants are taken to be mistaken about their own reasons, and the analyst supplies the true ones.

It is recorded here as a study rather than as a fact because it is a clean example of a testable proposition about sacrifice. It made claims about calories, about who ate what, and about when in the agricultural year the practice concentrated. Claims of that kind can be checked, and within a year they were.

Bernard R. Ortiz de Montellano answered the protein hypothesis in Science in May 1978, and the answer was arithmetical rather than rhetorical.

Tenochtitlan was not short of food. The city drew very large quantities of maize, beans and amaranth as tribute from the provinces, and it worked chinampa agriculture in the lake beds at high yield. Those two sources between them were sufficient for the population. Maize and beans eaten together supply complementary amino acids, so a diet that looks poor in animal protein is not necessarily poor in protein.

The distribution was wrong for the hypothesis. If the practice were a response to protein scarcity, the flesh would have gone to the people who were short. It went to the elite and to those connected to the ritual, that is, to the best fed people in the city.

The timing was wrong as well. Scarcity in an agricultural year falls before the harvest, not after it. The concentration of the practice fell at harvest, which fits a thanksgiving far better than it fits a famine response.

There is a general lesson in the exchange worth more than the specific result. The proposition was rejected because it was specific enough to be wrong. A great deal of what is written about sacrifice in other people's religions is not specific enough to be tested at all, and that is a defect rather than a form of caution.