Inca chiefs were carried in litters by their bearers, along roads that were a triumph of engineering. Deep ravines were crossed by suspension bridges made of twisted plant fibres.
South America was a relatively isolated continent before 1492. It thus provides us with interesting opportunities to compare world-historical processes in different continents that were largely or completely independent of each other. To the extent that we can find parallels between processes in the Old and New Worlds in pre-Columbian times, they may tell us something about recurring patterns in world-historical processes that are independent of culture. Warfare is a phenomenon that can be understood from a world systems perspective, in the sense that the occurrence and forms of warfare are considered expressions of the economic relations between different societies at considerable distances from each other. The comparison of patterns and forms of warfare in the Old and New Worlds can thus reveal something about general properties of world systems. Our information on warfare in pre-Columbian South America comes primarily from three sources: archaeology, ethnohistory (that is, indigenous peoples’ own historiography), and contact period history. The bulk of this information is from the Andes. To make sense of these various kinds of information, it is useful to organize it both regionally and chronologically, as well as in terms of a coherent framework of interpretation. The two macroregional divisions considered here are the Andean area and Amazonia.
The Andean Area
The Andean mountain range that runs along the entire extent of western South America saw the development of many complex societies such as chiefdoms, states, and empires during three and a half millennia before the Europeans arrived. The Spaniards who in 1532 conquered the Inca Empire encountered the last of these complex indigenous societies. Like its many predecessors, the Inca state can be understood as an attempt to control the resources, human labor, and flows of goods in the central Andean area. The varied geography of the region had for thousands of years stimulated trade between the arid coast, the mountains, and the tropical rainforests in the eastern lowlands. Attempts to politically control such trade through military means can in northern Peru be traced to the Initial Period of pottery use (1800-800 BCE). A powerful theocratic state based in the Casma valley on the north coast appears to have maintained trade relations with the north-central highlands and the tropical forest areas beyond. Toward the end of the Initial Period, the Casma polity collapsed and the area was invaded by its former trading partners in the highlands, associated with the site of Chavin de Huantar on the upper Marañon River on the eastern slope of the Andes. This initiated the first of three major pre-Columbian attempts by highland societies to take advantage of their middleman position between coast and jungle to gain political power. The carved monoliths at Cerro Sechin, a ceremonial center from this period in the Casma valley, show armed men and their dismembered victims, indicating an emphasis on violence, but we do not know if they represent war scenes or ritual sacrifice. As sites in this period are not fortified, it appears that social integration was achieved more by economic and ritual means than by violent coercion.
During the Early Horizon (800-200 BCE), the highland Chavin polity exerted influence over much of Peru. The traditional view is that the widespread distribution of Chavin art designs reflects a religious cult that spread by peaceful means, but defensive fortifications on the north coast suggest that military confrontations also occurred. Conflicts may have arisen regarding control of trans-Andean trade routes conveying Ecuadorian Spondylus shell and tropical forest produce, or of coca-producing zones, following a pattern known from later periods. Trophy heads appear not only on north coast pottery, but also on chavinoid Paracas textiles on the south coast and in the art of Pukara in the Titicaca basin. Skull fractures indicating blows to the head occur among the mummies at Paracas. The Chavin interaction sphere began to disintegrate in the third century BCE.
In the Early Intermediate Period (200 BCE-600 CE) the decidedly more militaristic Moche state controlled much of the north coast and built heavy fortifications against a highland polity based at Cajamarca. Such defensive architecture typically consisted of hilltop bastions with walls, moats, and stores of sling stones, but as they lacked water they appear to have been built as protection against brief raids rather than long-term sieges. Trophy heads occur on Moche ornaments, and the pottery shows many realistic war scenes, with warriors in helmets with maces and shields, and the sacrifice of prisoners. Recuay pottery from the adjacent highlands also shows warriors with clubs, shields, or trophy heads. Moche expansion relied on a mixture of military conquest and voluntary conversion of local chiefs, a pattern that continues through later expansions including the Inca. Prestigious older polities were sometimes incorporated peacefully by indirect rule. Examples of this include the Casma area within the Moche state and the oracle of Pachacamac on the central coast, venerated throughout its existence up until the arrival of the Spaniards. On the south coast, Nazca pottery shows trophy heads and later also militaristic motifs, but the area seems not to have been subjected to centralized state rule. Toward the end of the period, the expansion of the Tiwanaku polity in the Titicaca basin shows little direct indication of military conquest, but there is evidence of human sacrifice and the hostage taking of subject peoples’ sacred objects. In northwest Argentina, warriors and trophy heads appear as artistic themes at this time.
In the Middle Horizon (600-1000 CE), a short-lived but powerful state based at Wari in the Mantaro basin, in the south-central highlands, seems to have controlled the south and central coasts in their entirety and almost all of the Peruvian highlands from the city of Pikillaqta in the south to Cajamarca in the north. Like its southern counterpart Tiwanaku, Wari based much of its power at the periphery of its domain on dispersed administrative centers that regularly hosted ceremonial feasts for local leaders. Although some sites suggest a military presence, such as on the boundary of Tiwanaku in the southern highlands, relations with earlier polities in the northern highlands and on the coast show little evidence of militarism. As in earlier periods, however, there is evidence of human sacrifice and ritual decapitation. During this period, the Moche state declined and shifted its capital farther north. The Wari Empire collapsed around 800 CE and Tiwanaku about two centuries later.
In the Late Intermediate Period (1000-1476 CE), the collapse of Wari and Tiwanaku left smaller polities warring with their neighbors throughout southern and central Peru. One of these was the emergent Inca state based in the Cuzco valley on the southern boundary of the former Wari Empire. The endemic warfare selected for militaristic leadership in Cuzco as elsewhere. Most archaeological sites from this period are fortified or located on easily defensible hilltops. Coastal and highland groups in central Peru fought over access to the best coca-producing lands at intermediate elevations. In the north, the old Moche state revived as the expansive coastal empire of Chimor (or Chimu), based at the metropolis of Chan Chan. Like earlier north coast polities, Chimor controlled the crucial maritime trade in Spondylus shell from Ecuador.
In the Late Horizon (1476-1532 CE), the Inca state expanded through a combination of diplomacy and militarism to create an empire stretching from Ecuador to central Chile. The military history of this expansion has been reconstructed in some detail, thanks to native Quechua historiography recorded by sixteenth-century Spanish chroniclers and to archaeology. Over the course of less than a century, the Inca were able to assimilate or subdue hundreds of states, chiefdoms, and ethnic groups, including the old kingdoms of the Titicaca Basin, the Chanka and Wanka polities of the Wari heartland, the empire of Chimor, and finally even most of modern Ecuador, the source of the prestigious Spondylus shell. Further expansion failed when Inca armies faced simpler but warlike societies in unfavorable environments in northern Ecuador, the eastern lowlands, Argentina (Chiriguano), and central Chile (Mapuche). On the frontier in northern Ecuador, the Inca used fortresses from earlier polities in the area, while on the southeast frontier in Argentina they built a series of new fortresses to stave off attacks from the Chaco tribes. When Francisco Pizarro (c. 1475-1541) and his 260 Spanish soldiers arrived, the Inca Empire was divided by a civil war between two sons of the former emperor, of which the victorious Atawallpa, destined to be executed by Pizarro in 1533, was based at Quito. Building on native narratives, the chroniclers’ accounts of this civil war, battle by battle, are the most detailed description of pre-Columbian warfare in existence. Recruitment to Inca armies was by a general draft apparently aligned with the decimal administrative system, which divided the population into groups of ten, one hundred, and so on. Weapons included slings, bows and arrows, bolas, spears, spear-throwers, lances, axes, and bludgeons. Armor and shields were also used.
Some general conclusions on the occurrence of warfare in the pre-Columbian Andes can be drawn. There seems to have been a distinctly Andean tradition of diplomacy through voluntary conversion and indirect rule that, in many cases, offered means of control other than military conquest. In the periods we call horizons, when large parts of the area were well integrated ideologically, whether or not this was achieved with the aid of military force, there was comparatively little warfare, except at the borders of the unified territories. In the intermediate periods, following the disintegration of such polities, there was much more conflict between different local populations and ethnic groups. Conflicts often seem to have involved rivalry over the control of strategic resources or trade routes. The emergence of new power centers generally occurred on the periphery of previous polities, challenging older centers and contributing to their decline. Warfare, like the nonviolent practice of power it complemented, was always embedded in religious or symbolic meanings and often had a ritual aspect, such as headhunting and human sacrifice. This is evident even in the Inca period, for instance from accounts of divination and sacrifice in preparation for battle, and from reports that the skins of defeated enemies (e. g., defiant Kañari warlords from Ecuador or rebellious lords from the Titicaca area) could be fashioned into drums played at festivals, and their skulls into drinking cups. The obsession with ritualized warfare, human sacrifice, and cannibalism that has been attributed to the Chibcha-speaking chiefdoms of the Colombian Andes probably had much in common with early instances of theocratic warfare in the central Andes.