MESOAMERICA ON THE EVE OF THE SPANISH CONQUEST

It really is a “numbers game.” The only real evidence is archaeological, in the number and density of settlements in the Late Postclassic. There has not been found, and probably will not be, a single number from a Pre-Columbian text about any population. One reason for revisions to the population size in 1492 or 1519 is that as new archaeological information and context become available, there is more with which to work. However, the translation of archaeological evidence into population numbers is not simple; all ways require assumptions (Hassan 1981). Mesoamerica is fortunate to also have some ethnohistorical information as well. However, it seems to come down to the philosophy and intuitions of individual researchers.

Earlier in the twentieth century, the estimates for Mexico at contact were 3.2 million (Kroeber 1934) and Rosenblat (1954) at 4.5 million. Then in the 1960s, the estimates jumped to 25 million (Borah and Cook 1963) and Dobyns (1966) at 30–37.5 million in central Mexico alone. While more recent estimates of central Mexico are at 16 million (Whitmore 1992) and 17.2 (Denevan 1992), these are definitely higher than the original estimates. What changed? Whether one believes the Pre-Columbian populations were few or many seems to be affected by opinions on how complex or “civilized” Pre-Columbian societies were and how devastating European conquest and colonization were (Alchon 2003). Mesoamerica definitely had complex, hierarchical societies, so some dense populations and significant numbers of people would have been present.

Central Mexico has been the most common area for estimating the precontact population. The most famous is Borah and Cook’s (1963) estimate of over 25 million, which has been accepted but also severely criticized. Using pre-conquest tribute lists (post-conquest copies) and other documents, they calculated the number of tributaries in central Mexico. For total population, one multiplies that figure by a probable average family size (4.5 people) and then adds the proportion exempt from tribute (35–40 percent). However, the relationship between Aztec tribute and population is rather tenuous, because there is no evidence that such tribute was based on the number of taxpayers in a province; it was levied on conquered rulers and how they collected it is unknown (Sanders 1976). The exempt proportion also seems inflated. Borah and Cook pile assumption on assumption and have received more support than warranted by any evidence (Henige 1998).

Sanders (1976; Sanders et al. 1979) criticized Borah and Cook but also did his own calculations based on documentary and archaeological evidence, especially for the Basin of Mexico. He began with numbers for 1568, the first reliable census evidence, and then calculated a depopulation ratio back to 1519 for an estimate of 1–1.2 million for the basin, and 2.6–3.1 million for the basin and adjacent areas. He also has archaeological evidence from the detailed survey of the basin (Sanders et al. 1979). The Aztec period had 1,636 sites, compared to only 398 the period before, indicating a dense settlement. Using artifact density and size of settlements, the calculation was 1–1.2 million at contact, very similar to his documentary estimate. Using Sanders’s estimates, Denevan (1976) calculated a figure of 11.4 million for central Mexico (45 percent of Borah and Cook’s figure).

Whitmore (1992) also tried to calculate the basin population in 1519 but used computer simulation based on censuses in 1530s and 1560s, while modeling fertility, mortality, and migration based on epidemiological models of morbidity and mortality for newly introduced diseases. He ran the model multiple times, varying the demographic and epidemiological parameters, and then compared with the historical data and what seemed probable. His figure was 1.5 million for the basin in 1519, close to Sanders, and 16 million for all of Mexico (64 percent of Borah and Cook’s figure).

The rest of Mesoamerica (the Yucatán Peninsula and northern Central America) has received less attention. That is because the dense population is thought, on archaeological grounds and perhaps unjustly, to be present during the earlier Classic Maya times rather than in the early sixteenth century. Estimates here also vary widely, from 800,000 to 8 million for just the Yucatán (Denevan 1976). Frederic Lange (1971) used documentary evidence to count the large Maya cities, mostly coastal, for an estimate of almost 2.3 million for the northern Yucatán. An estimate for Guatemala (Lovell and Swezey 1982) was calculated three ways: by using a tribute assessment from the mid-sixteenth century, by using the average population density as calculated for central Mexico, and by aggregating available population data for subregions. This latter calculation is the most reasonable method and yields a figure of 2 million. Chiapas (275,000–200,000) and Soconusco (80,000–90,000) are based on similar methods, such as the use of early tribute documents and information on settlements and population from archives (Lovell and Lutz 1995). If all are added together, including that for Yucatán from Lange, the estimate is 4,665,000. As Lovell and Lutz (1995) note, this puts the estimates among the high numbers, but not at the extreme high end. The accumulating evidence that this part of Mesoamerica had complex, hierarchical societies at contact continues the trend toward higher population estimates.

The recent totals for Mesoamerica seem to range from 16 million to near 30 million, a rejection of the earlier small estimates. As Henige (1998) reminds us, many sources have not been critically evaluated. Much more archaeological research is also needed.

CASUALTIES DURING THE CONQUEST

The question of how much credence should be given to early accounts by the Spanish underlies the discussion of mortality during the conquest. Undoubtedly, the Spaniards were able to take over Mesoamerica with relatively few Spanish soldiers: the interest is in the details. Several accounts by eyewitness conquistadors contain estimations of the sizes of the armies faced as well as the sizes of towns and provinces. There are a few accounts by native Mesoamericans, mostly of the conquest of central Mexico, but these were written long after the events and are “sketchier” in detail (Hassig 2006). It is generally agreed that the Spanish accounts were written for self-serving reasons (Hassig 2006; Henige 1998). Thus, the numbers of warriors and sizes of settlements encountered come mostly from the Spaniards (and also influenced the estimation of pre-contact numbers), and there is no reason to think that any numbers were meant to be accurate, especially about native demography.

For example, Cortés faced and defeated armies at Tlaxcala he estimated to be at “more than 149,000” with only a few hundred Spaniards (Cortés 1988). Díaz del Castillo (1982) argued that such numbers are inflated; they faced only 30,000, 40,000, or 50,000. Henige (1998) questioned whether any of these numbers were reasonable or were just common battlefield inflations for the glory of the teller. After the conquest of Tenochtitlan, Cortés had nine hundred surviving Spaniards (Hassig 2006). However, “what made the conquest of Mexico possible was not the Spaniards’ military might, which was always modest, but the assistance of tens and even hundreds of thousands of Indian allies—laborers, porters, cooks, and especially soldiers” (Hassig 2006: 175). Cortés’s success was due to “the pivotal role . . . played by his two hundred thousand Indian allies, even though they went virtually unacknowledged and certainly unrewarded” (Hassig 2006: 178). While Hassig does depend on Cortés’s account, exactly how these numbers were determined is not clear. More investigation of native archives and other Spanish sources resulted in a calculation of circa 24,000 native allies at the siege of Tenochtitlan (Oudijk and Restall 2007), a distinctly lower number. While Hassig (2006) does give credit, as do others, to the Spanish advantage in horses and metal technology, they were highly outnumbered by their allies and their adversaries. The native allies prevented the Spaniards from being overwhelmed. The bulk of the warrior casualties, as well as the number of residents wounded and killed by the attacks on settlements, would have been borne by the natives. There is just no information presently available to allow a quantification of the losses, but it probably was significant for some groups.