Thursday, November 21, 2013

Drought and the collapse of Teotihuacan

While archaeologists have been working for decades to figure out the collapse of Teotihuacan, in come the geologists to decide the matter with a graph of rainfall patterns from a cave >300 km away (Lachniet et al 2012). Guess what? Drought caused the collapse of Teotihuacan ((sarcasm)).

Although the authors claimed to "test hypotheses of climate and cultural change in the highland Basin of Mexico," (p.259), I didn't see any testing in the paper. Instead of testing a model, they employ what Lewis Binford called a post-hoc accomodative argument. They generated some findings, then scratched their head and thought about how to interpret them without reference to explicit hypotheses.

They should have derived the implications for different collapse scenarios ahead of time, and tested for these. What is the most likely timing and nature for a killer drought to generate urban catastrophe? A sudden major drought, or a prolonged less severe event, or a sequence of sporadic droughts? Is a collapse more likely to occur from a drought early in a city's development, or after a long period of urban success? How might drought-induced collapse relate to demographic trends? I have no idea what the answers are, but I would guess there are suggestions in the literatures on urban disasters and societal collapse.

From their graph (above), it looks like Teotihuacan weathered a major dry spike in the early 400s, and then collapsed in a period of fluctuating slightly dryer episodes. Is that enough to claim that the climate events were "associated" with the decline of Teotihuacan population (p. 260), or that "rainfall variability over the past 2.5 k.y was linked to Basin of Mexico cultural changes" (p. 261)? I don't buy it. But in the absence of an explicit model there is no way to determine objectively whether there is any association between these things or not. We will all have our opinions, with no rigorous way to choose among them.

Lachniet et al. note a "megadrought" at AD 690-780 , but by that time Teotihuacan had already shrunk from its Classic period height to a smaller Epiclassic town. But wait. The authors claim that the "Teotihuacan period" didn't end until AD 800 (according to the caption to the graph), and perhaps they think that this claim salvages the drought interpretation. I guess if you are a geologist, you feel free to make up archaeological data to fit your models.

Would it really have been too difficult to consult an archaeologist who knows something about Teotihuacan, its chronology, and its environmental context and dynamics? I'm sure Linda Manzanilla, Emily McClung de Tapia, George Cowgill, or David Carballo would have been happy to help these geologists get their archaeological data and interpretations straight. Instead, some very important new data are slighted by facile and unlikely cultural interpretations. I think findings like this are of utmost importance for understanding the past, but not if interpreted in a faulty and archaeologically-clueless fashion.

Lachniet, Matthew S, Juan Pablo Bernal, Yemane Asmerom, Victor Polyak, and Dolores Piperno
    2012    A 2400 yr Mesoamerican rainfall reconstruction links climate and cultural change. Geology 40(3): 259-262.

See some of my earlier posts on drought and collapse:

The Maya collapse: When theoretical preconceptions get in the way of understanding

Collapse? What Collapse?


Anonymous said...

Off topic but have you heard about this? How pervasisve is this practice?

Michael E. Smith said...

This is all too common. I blogged about it on Wide Urban World: