Cannibals and Kings

by Marvin Harris

  But more about the moral implications of this tale later on. The immediate task before us is to trace out the further consequences of the rise in the state in the context of different regional patterns of intensifications, depletions, and ecological crises. I turn first to the tragic history of Mesoamerica.

  8

  The Pre-Columbian States of Mesoamerica

  Some archaeologists claim that ecology and reproductive pressure had little to do with the rise of the state in Mesoamenca. They believe that the transition to statehood occurred first among the Olmec and Maya, who lived in lowland swamps and jungles where there was neither an opportunity to practice intensive forms of agriculture nor barriers to the dispersion of population. Supposedly these jungle states evolved as a result of spiritual impulses peculiar to the Olmec and Maya conception of the world. Believing that the rains, crops, and continuity of life were dispensations of the gods, the Olmec and Maya felt the urge to build ceremonial centers and to house and provision a priestly class of nonfood producers. Because they happened to be more religious than other pre-state village peoples, they built larger temples and showed uncommonly great respect and devotion to their priests and officials. Cost benefits were irrelevant. Their political organization did not result from population growth, declining efficiencies, warfare, impaction—or anything else so crass. Rather, it evolved from voluntary submission to a benevolent theocracy.

  Archaeologists who advance this kind of explanation for the origin of the state of Mesoamerica seem to be exhilarated by the notion that human faith and ingenuity triumphed over adverse ecological conditions. While I sympathize with the sentiment that lies behind this celebration of the creative achievements of cultures like the Olmec and Maya, I think it is far more urgent that we understand the limitations placed by ecological and reproductive factors on even the most inspired forms of human activity.

  The Olmec are indeed a puzzling case. Described by the Mexican archaeologist Covarrubias as the “mother civilization” of the New World, the Olmec inhabited the humid lowlands and coastal plains of the Mexican Gulf Coast states of Vera Cruz and Tabasco. Between 1200 and 800 B.C. they erected a number of widely separated temple centers—the earliest in the New World—on top of artificial mounds two or three acres in extent. The best-known site is La Venta in Tabasco, on an island in the middle of a swamp. La Venta’s most imposing structure is an earthen cone 420 feet in diameter and about 105 feet high. Monumental sculptures consisting of fifty-ton carved stone slabs called stelae, altars, and huge, round human heads that appear to be wearing football helmets lie strewn about the site.

  While the Olmec ceremonial centers contain impressive evidence of the ability of redistributor-chiefs to organize cooperative projects and to support artisans skilled in sculpture, masonry, and the making of jade jewelry and fine ceramics, the scale of their endeavors falls short of what one would expect of a state-level polity. Each site could easily have been constructed by a population of no more than two or three thousand people and each is too far from the others to constitute a single interconnected political system.

  To keep the Olmec in perspective, one must consider the scale of construction characteristic of sites that are historically known to have reached the threshold of state formation. When the first French explorers ascended the Mississippi Valley, for example, they found populous “towns” and huge earthen platforms supporting wooden temples and the houses of priests and nobles. A remnant of the largest of these structures, the Cahokia mound, still exists on the outskirts of east St. Louis. Before being chewed up by bulldozers, it was over a hundred feet high and covered fifteen acres, as compared with the two or three acres typical of the Olmec sites. Moreover, we know that impressive feats of construction can be carried out under the auspices of “big man” redistributor-chiefs who lack the capacity to tax, conscript, and punish their followers. Even the nonagricultural Kwakiutl and Haida of the Pacific Northwest, led by redistributor-chiefs, were capable of a certain amount of monument-making in the form of totem poles and carved house posts. At Stonehenge and other early ceremonial centers in Europe associated with the spread of farming, pre-state chieftainships managed to erect elaborate astronomically oriented monuments out of blocks of stone that weighed considerably more than those found at La Venta. And the Olmec sites are actually puny by comparison with the great highland centers of the central plateau of Mexico. At best they represent a stage of development which was arrested at the level of incipient statehood. Their failure to develop further was clearly related to the fact that because of ecological circumstances their regional population densities remained low and unimpacted.

  I should also mention the possibility that ceremonial structures indicative of incipient statehood older than the Olmec may yet be discovered in the central plateau highlands. Recent excavations by Ronald Grennes-Ravitz and G. Coleman indicate that Olmec-type figurines found in Morelos and the Valley of Mexico are as old as those found in Vera Cruz and Tabasco. Moreover, at these highland sites Olmec artifacts occur above strata containing indigenous highland ceramic traditions that predate the Olmec period by as much as 400 years. Olmec temple centers may therefore yet be shown to have been partially dependent on the growth of the first highland states. It is even possible that the Olmec sites represent colonial outposts—perhaps pilgrimage centers, as Grennes-Ravitz and Coleman have suggested—around which trade between the tropical lowlands and the arid central plateau was organized.

  To the east of the Olmec heartland lies the Yucatán Peninsula, another region in which the path toward statehood seems to flout ecological principles. Here lived the Maya, a people who invented a complex system of hieroglyphic writing and mathematical numeration, wrote their history in accordion-shaped books, made precise astronomical observations, developed a highly accurate solar calendar, and were masters of the arts of stone sculpture and masonry.

  And yet the lower half of the Yucatán Peninsula is covered by a dense jungle region called the Peten. From A.D. 300 to A.D. 900 the Maya busied themselves with the construction of numerous ceremonial centers right in the middle of this region. Norman Hammond has counted eighty-three major sites in the southern portion of the Yucatán, separated by an average distance of only 15 kilometers (9.3 miles). In these centers are elaborately ornamented multi-room buildings grouped symmetrically about paved central plazas; ball courts for ritual games; stone slab stelae with commemorative dates, genealogies of the rulers, and other historical information not yet decoded; altars incised with additional hieroglyphic texts; and massive statues of the gods and the nobility. Towering over all are great truncated pyramids faced with cut stone and topped by stone temples. The largest site is Tikal, whose temple pyramids rise precipitously 190 feet above the plaza floor. At its maximum, during the 19th century A.D., Tikal may have had as many as 40,000 inhabitants in its rural perimeter while the overall regional density has been estimated at 250 people per square mile. This would make the Petén as thickly populated as modern-day Europe. There is no doubt that the largest of the Maya centers were the administrative capitals of small states. But there is no chance that the Maya achieved statehood entirely independently of preexisting states in the highland region. Teotihuacán, which I’ll describe in a moment, already contained several tens of thousands of inhabitants when Tikal was just beginning to rise above the treetops. Teotihuacán is more than 600 miles from Tikal, but the military and economic shock waves sent out by the great highland empires regularly reached even more remote regions. We know that by A.D. 300 Kaminaljuyu, a Maya city in the Guatemalan highlands overlooking the Peten, had come under the influence of Teotihuacán. Kaminaljuyu probably contained a military garrison which controlled the trade routes between the Peten, the Pacific Coast, and the Central Mexican Plateau. After A.D. 300 trade goods, painting styles, and architectural motifs in the Peten centers themselves leave no doubt that the Maya were being affected by events in the central plateau highlands. Actual military engagements between late formative or
early classic highland states and incipient Maya states in the Peten are not to be ruled out.

  Trade between the Maya and their highland neighbors may also have moved the Maya closer to statehood. The Peten region lacks indigenous sources of rocks suitable for making metates and manos or knives and projectile points. These items were crucial for grinding corn and for military weaponry. Along with salt, they were obtained through trade with the highlands. This trade may have widened the distance between the early Maya redistributor-chiefs and commoners in two ways: more effective terms of trade could be obtained by more powerful individuals who were the equals of the state-level nobility with whom they had to deal, and the control over these additional strategic resources could have added to the potential for controlling the incipient peasant food producers. In general, the larger the volume of trade, the greater the flow through the redistributive system and the greater the power of the individuals who are in charge of the redistributive process.

  The evidence allowing for an interpretation of the Maya centers as secondary states does not rule out the possibility that reproductive and ecological pressures generated within the Peten region itself might also have contributed to the process of state formation. The Peten “jungle,” on close inspection, turns out to be full of surprises. The first aspect that needs to be clarified is its size—only 30,000 square miles, compared with 2 million square miles for the Amazon-Orinoco. Next, there is its peculiar pattern of rainfall. As one moves northward from the Peten to the tip of the Yucatán Peninsula, annual rainfall decreases and forests are replaced by thorny shrubs, cactus, and other drought-resistant plants. Within the central Peten forest itself, annual precipitation is only about half that of the Amazon-Orinoco. The Peten dry season is exceptionally severe, and both the annual and seasonal totals are subject to extreme variations. A single drop of rain may not fall during the months of March and April. Drought conditions frequently prevail during February and May, and even during the rainy season itself. In the words of C. L. Lundell:

  The vegetation does not have the luxuriance of true rain forest, hence it may be designated a quasi-rain forest. The rainfall averages less than 1800 mm. [71 in], a maximum not sufficient to maintain true rain forest in a region with a pronounced dry season.

  Many of the Peten trees shed their leaves every dry season, a tendency that is accentuated during droughts. This “jungle,” in fact, sometimes gets so dry that farmers don’t even have to “slash” in order to clear next season’s garden plots by setting fire to the underbrush. Preventing fires from spreading is the major preoccupation on such occasions.

  And now we come to the fact that the Yucatán Peninsula has a peculiar geological structure. Its bedrock consists almost exclusively of porous limestone (hence the need to import rocks for grinding corn from the highlands). This results in there being few permanent rivers and lakes since most of the rainfall percolates rapidly down through the limestone and disappears entirely without any surface runoff. During the dry season there is even a shortage of drinking water except where there are natural clay-bottomed water holes in the limestone whose interior drainage has gotten clogged.

  As one might expect, the earliest Maya villages were located near the only two permanent rivers on the Yucatán Peninsula: The Usumacinta on the southwest and the Belize on the southeast. Around 600 B.C. the region surrounding Tikal appears to have been uninhabited, suggesting that it was only after the favorable riverine locales had filled up that farmers began to colonize the interior of the forest. These colonists must have resembled the Yanomamo and other canoe-less “foot Indians” who live in the protein-deficient zones of the Amazon-Orinoco basin away from the main rivers. But in a short while the peculiar geomorphology and climate of the Peten region would have created a situation which has no parallel in Amazonia.

  The early Peten farmers were not free to spread out evenly through the forest. Settlements would have had to be located near water holes which could be counted on not to dry up during a severe drought. We know that later on entirely artificial cisterns called chultuns were dug as deep as sixty-six feet into the limestone bedrock and plastered with lime cement in order to assure supplies of fresh water. Some chultuns were built under the paved plazas of ceremonial centers, which acted as catchment basins during rainstorms. At one modern village in Campeche, dry-season drinking water had to be obtained by descending 450 feet below the surface through an underground cavern. All of the classic Maya sites, including Tikal and other Peten centers, were built next to artificial or natural storage wells or reservoirs. The most famous of the natural water holes, or cenotes, is located next to Chichén Itzá, a late Maya center in northern Yucatán. Large quantities of human bones and gold artifacts dredged up from its bottom suggest that people and ritual objects were thrown into it to appease the water gods. And so the lively possibility exists that the early settlements in the Peten tended to increase beyond the normal breaking-up point of tropical forest villages. This theory removes the problem of the initial growth of Maya ceremonial centers from the realm of heaven to the realm of earth and water. The Maya farmers had a very practical reason for not fleeing into the forests when their redistributor-chiefs started acting like kings instead of like mumis.

  The next question to be confronted is how the Maya under the direction of their redistributor-chiefs managed to raise their population density to a level that was 250 times greater than that achieved in the interfluvial zones of the Amazon-Orinoco. Archaeologists have generally assumed that the ancient Maya farmed the Peten the way their modern descendants do—by means of the system known as slash-and-burn. But this is clearly an impossibility.

  Slash-and-burn is a form of agriculture that is well-suited for regions that have abundant forest cover and high rates of regeneration. The object of the slash-and-burn system is to use a section of forest for a few years, let it lie fallow long enough for trees to grow back, and then use it again. “Slash” refers to the practice of cutting down small trees, vines, and shrubs and letting them dry before setting fire to them. The burning, usually carried out just before the onset of the rainy season, creates a layer of ash that acts as fertilizer. Crops are planted directly into the ash-covered soil in holes or small mounds without the need for tillage. High yields of corn, beans, squash, and other crops can be obtained for two or three seasons. Thereafter weeds spread from the surrounding uncut forest and infest the field; at the same time the ash fertilizer is leached away by rainfall. Soon a new plot must be found. Slash-and-burn agriculture is capable of high returns per acre and man-hour provided that an interval sufficient to permit a substantial regrowth of trees and shrubs is maintained between successive burnings. The greater the quantity of ash, the higher the yields. The longer the interval during which a forest is left fallow, the more wood there is to make ashes with. For this reason, slash-and-burn farmers in Southeast Asia think of themselves as “the people who eat forests.” The shorter the fallow period, the lower the yields. In tropical forests the decline can be precipitous not only because the concentrated heavy rainfall rapidly leaches away the soil nutrients but because weeds grow thicker each year the field remains in continuous use.

  Slash-and-burn was undoubtedly the system used by the earliest farming peoples who entered the Petén, but it could not have remained the principal mode of subsistence during and after the transition to the state. By counting the ruins of house sites, Dennis Puleston of the University of Minnesota estimates that there were 2,250 persons per square mile in the residential zone around Tikal and 750 persons per square mile in the zone between Tikal and its neighbor, Uaxactun. It is impossible for slash-and-burn systems to support such densities. Considering the entire Peten area, Sherburne Cook shows that enough maize, beans, and squash could have been grown with slash-and-burn techniques to support the estimated overall population of 1.5 million. But these calculations assume that the farmers were evenly spread throughout the forest and that they were free to move to new clearings as the old ones were exh
austed. Neither of these assumptions is valid since the limiting effect of the dry season on the availability of drinking water is not taken into account. Furthermore, during the rainy season low-lying areas face the opposite problems—too much water—and are too swampy to be used without digging drainage ditches.

  On theoretical grounds, the picture of what must have happened seems clear. As the population of the Peten increased, the slash-and-burn cycle must have been intensified, resulting in shorter fallows between burning and hence declining efficiency. This set the stage for the adoption and spread of a more efficient system involving higher start-up costs, which in turn provided the basis for still higher population densities and the emergence of the first statelets. But what was the nature of the new and more productive system? I fear that my theory has run ahead of the archaeological facts, but there are some hopeful signs that the facts are about to catch up.

  One of the measures taken by the Maya when the efficiency of slash-and-burn declined was to plant groves of breadnut trees (Brosimum alicastrum). As C. L. Lundell pointed out back in the 1930’s, the breadnut is the most common tree covering the ruins of the Peten ceremonial centers. When archaeologists speak dramatically of having to hack away the jungle in order to expose the wonders of Maya architecture and sculpture, they generally neglect to say that they were hacking away at an overgrown orchard. Tree crops, of course, do have high start-up costs—one must wait several years before they begin to return the labor invested in them—but they are highly productive per acre and per man-hour. Recently, Dennis Puleston, having discovered that each house site at Tikal was surrounded by a grove of breadnut trees, reached the conclusion that breadnuts provided 80 percent of the calories consumed by the people of Tikal during the ninth century A.D. There are other alternatives, however, which may simply have been overlooked by the generation of archaeologists who preferred to think the Maya temples were let down from heaven on golden threads rather than built on the backs of people who wanted to know where the next meal was coming from. In this connection, one of the most important discoveries ever made about the Maya may prove to be the one made in 1975 by Ray Mathenay at Edzná in Campeche. Working with aerial photographs taken during the rainy season (others had limited their aerial photography to the dry season, when conditions were “better”), Mathenay detected a network of canals, moats, and reservoirs radiating out from the ceremonial center. Because of the dense foliage covering them during the rainy season and the fact that the water in them dries up during the dry season, these constructions are difficult to detect from ground surveys alone.