Environmental Thresholds and the Empirical Reality of State Collapse: A Response to Erickson (1999)

Article excerpt

We are grateful for the opportunity to respond to Erickson's (1999) critique of our articles on human-environment interactions in the Lake Titicaca basin of Bolivia (Ortloff & Kolata 1993; Abbott et al. 1997; Binford et al. 1997). His decision to publish this critique in ANTIQUITY, rather than in the journals in which our articles appeared, permits us to reach a broader audience. Erickson labels our interpretations a form of `neo-environmental determinism', but his rejection of our conclusions stems from serious misunderstandings and is misleading to readers who have not examined our original data. He (p. 634) claims:

1 our research represents `simplistic reductionist thinking' that treats humans as `passive pawns' of environmental change;

2 our dating of the chronic drought in the Andean altiplano after AD 1150 is imprecise and not correlated with the 12th-century disintegration of the Tiwanaku state; and

3 the drought did not affect intensive agricultural production.

We address these claims in reverse order.

Erickson confuses the distinction between short-term and long-term impacts of precipitation change when invoking modern drought records. He states that `many of these historic droughts and floods are on a par with' the profound climate change that we documented between the 12th and 15th centuries. However, these were short-term dry spells (2-5 years) with lake-level fluctuations of 2-3 m around the 20th-century mean. The cited 17th-18th-century droughts were only 1-3 decades in duration. Erickson fails to grasp the order-of-magnitude difference in duration and ecological impacts between these short-term events and the centuries-long drought we documented. He then uses comparisons of cultural response to modern droughts to argue that 12th-century populations would have readily coped with protracted drought, but his ethnographic analogies are invalid. Contemporary agricultural infrastructures and technologies are significantly different from those of the Tiwanaku period: modern populations do not depend on water intensive, raised-field agriculture. Demographic conditions and socio-economic structures of the pre-Hispanic and modern periods are incommensurate. Erickson's analysis assumes that the political economies of these periods are comparable. Structural disjunctions between 12th- and 20th-century rural society far exceed any assumed continuity.

According to Erickson, the `severe drought of 1983-84 provided an excellent test of raised-field agriculture'. In reality, this brief dry spell constitutes no test at all. The level of Lake Titicaca remained near the 20th-century mean throughout this `severe drought'. Lake level did not fall 15-20 m as it did during the 12th-15th-century drought (Binford & Kolata 1996; Abbott et al. 1997; Binford et al. 1997: 239-40, figure 7). Although raised fields mitigate the impacts of short-term dry spells, they cannot sustain production under chronic aridity.

Erickson asserts that raised fields could be cultivated on lands exposed by the receding lake during drought conditions. His comments reveal a deep misunderstanding of water budgets and their agricultural consequences. Water in a drainage basin is finite. Each year, some water enters the system through rainfall and shallow groundwater; some is lost to evapotranspiration, seepage to deep aquifers, and outflow; and some remains in the basin. When input decreases but output remains the same for several years, water volume necessarily decreases and lake level declines. Decreased infiltration to shallow groundwater causes a longer-term reduction in spring and river flow, the major sources of water for raised fields. The area of land that seems suitable for raised-field cultivation does increase with lowered lake level, but water available to fill canals significantly decreases. Fresh water is the fundamental condition for the ecological functioning of the raised fields (Biesboer et al. …