Most Mayanists are not strong believers in Gill’s “mega drought” hypothesis. Demarest and Webster have pointed out some of its flaws before. Here I intend to show some of his methodological problems, some of which relate to his use of direct historical analogies.
Gill (2000) uses the rain charts from 20th century Merida and correlates decreased precipitation noticed in the meteorological record with droughts and famines mentioned in textual sources. He discusses similar phenomena during the Colonial period when droughts and famines were recorded and for the meteorological data he largely is dependent on Farriss’ account (1984), but he has used this source selectively. When Farriss explains demographic declines during or after the Colonial period droughts as the result of populations moving into uncontrolled areas in order to escape demands of the Colonial regime (1984:75), Gill rather sees people dying from starvation and thirst despite that he mentions Farriss information (2000:310).
Thus, population decline can partially be described by movement out of areas of Spanish control and not only by death from dehydration or starvation (which for sure also occurred). In fact, to quote Farriss: “we have no way of comparing the mortality caused by famines before and after the conquest. But it is difficult to escape the conclusion that the Maya would have fared better without the sometimes well meaning but almost invariably disastrous intervention of the Spanish” (1984:63). She shows that the Spanish “innovations” both raised the mortality of the Maya and lowered their fertility. These “innovations” were the congregation program, changing marriage customs and family residential patterns, and the imported cattle that destroyed crops. These factors introduced by the Spaniards are ignored or downplayed in Gill’s account.
Colonial communication technology, such as ships, pack animals and wheeled vehicles, could potentially have facilitated transport of subsistence and made the Colonial population better prepared for droughts compared to those of the Prehispanic era. The Colonial power could have used these improvements to come to rescue with imported subsistence in the case of famine. However, Farriss (1984:62) suggests that this help would only have benefited the Spanish population in the Colonial cities of Merida and Campeche. When shortage was approaching, private grain merchants and officials went through the villages and bought grains to sell in the cities. This resulted in a food shortage for the people outside the major Colonial cities. Gill (2000:305) gives us several examples from Colonial sources mentioning people dying in streets and parks, but he ultimately fails to mention that these are Colonial political landscapes and that Prehispanic political landscapes lacked these patterns. Neither the examples he gives where people are starving in the rural areas are useful for understanding Prehispanic times (Gill 2000:308). The rural areas had been depleted to sustain large Colonial towns. No such pattern is known from Prehispanic times.
In short, Gill’s argument for how people and society reacts to droughts rests on shaky foundations. He imagines that people, despite their knowledge of past droughts, had no strategy to cope with them, they died when the drought was hard enough. This scenario may fit the Colonial period but it does not explain the Terminal Classic demographic shifts.
Gill projects his modern and Colonial scenario to the Terminal Classic and even older periods. Here he relies on sediment cores from different lakes, including Lake Chichancanab in the Cochuah region, which only is 7 km from the westernmost site mapped (Tabasquito). The sediment cores indicate an extended dry period during the Terminal Classic (Curtis and Hodell 1996). During this period there were several droughts which are fairly contemporary with the “collapse” and therefore the droughts are seen as the main causes for the “collapse”. But as shown above, this is largely based on Colonial analogies.
To sum up, Gill explains the Terminal Classic “collapse” by analogy of how droughts affect modern and Colonial settlement. Add to this that his model almost completely ignores detailed Terminal Classic settlement data, apart from Chichen Itza which is a somewhat anomalous site in this context (Dahlin 2002), and we have a too generalizing view where climate determines settlement.
What the CRAS survey(s) show is that there is a dense settlement extending from at least San Felipe in the east to Candelaria in the west (10-15 km from Lake Chichancanab). This settlement appears to be Terminal Classic, particularly Candelaria, and it is located south of the area controlled by the later Spaniards. There are no known water sources that can be securely dated to the Terminal Classic. Even if the few wells in this area are of Prehispanic origin they are far fewer per person compared to Colonial towns. How do we explain such a settlement density without large water sources? Maybe there are other dynamics in settlement that we have overlooked?
The Terminal Classic settlements were not the same as during the Colonial period. Settlement patterns were changed by the Spaniards, but this is a change that seldom is noticed even by the historians of the Colonial period. The early Spanish congregation of settlement is well known, but what the historians (and Gill) have missed, or at least underestimated, is how the Prehispanic settlement fluctuated in relation to climate changes in a non-catastrophic way. It changed back and forth in between the “collapses” as well. The large centralized sites that “collapsed” in the Terminal Classic are also the result of a less drastic “congregation” following various cycles (Marcus 1998). Colonial settlement appears not to have been capable to change along with these climate changes. Instead changes in settlement during Colonial droughts are explained as migration, as lines of flight out of the Colonial regime. This is probably not the result of a collapse or a desperate search for food, but rather an age-old strategy to cope with frequent drier conditions. People did not stay in place to await an approaching drought. As in Colonial times, the Prehispanic population probably had relatives living at other sites and there may have been seasonal movement, at least during long-term droughts. However, since we tend to visualize the geopolitical landscape as consisting of polities with contiguous spaces we may argue that it was impossible to move anywhere one wanted and that such migrations would lead to conflicts and warfare. But it is important to note that Prehispanic geopolitics did not consist of contiguous spaces as Martin and Grube (2000) have shown. The Colonial power on the other hand tried to homogenize the territory into a contiguous space. Space was striated on a grand scale (Deleuze and Guattari 1987).