Elizabeth N. Arkush. Hillforts of the Ancient Andes: Colla Warfare, Society, and Landscape. Gainesville: University Press of Florida, 2010. 320 pp. $69.95 (cloth), ISBN 978-0-8130-3526-0.

Reviewed by Paul Roscoe (University of Maine)
Published on H-Urban (February, 2012)
Commissioned by Michael E. Smith

Warfare and Political Devolution in the Andes

This book examines the Colla of the Lake Titicaca Basin of the Peruvian Andes in the Late Intermediate Period, between about 1000 and 1400 AD. This was a time of political devolution and stasis, bookended by the collapse of Wari and Tiwanaku states at the beginning and the extension of Inca hegemony at the end. Notwithstanding the collapse that ushered in this period, the Titicaca Basin seems to have remained relatively peaceful until about 1275, at least if we are to judge by how little evidence there is of fortification and other defensive measures. At this point, though, affairs darkened. There is evidence that climate began to change with increases in drought, which likely caused crop failures and may have fueled growing resource scarcity. Whether for this reason or some other, war became a stark and mordant reality. The population migrated away from the better lands at the center of the basin, dispersing into small settlements in or just below highly fortified table top forts (pukara). With this abrupt shift, the Colla abandoned much of their previous religious and ceremonial life, and powerful state rulers gave way to leaders with but minimally elaborated status and power. This state of affairs persisted until the latter half of the fifteenth century, when the warring polities of the pukara fell to the expansion of the Inca Empire--though not without an armed resistance that the Inca considered among the most formidable they had encountered during their entire expansion.

Arkush’s examination of this place and time draws on extensive archaeological survey and excavation work, coupled with a detailed reading of the chronicles compiled by Spanish clerics and bureaucrats in the early colonial era. The product is a work that is meticulously researched, analytically incisive, and so clearly written that it is a pleasure to read. Its particular worth, though, lies in the two issues it targets, warfare and political evolution. Each is of considerable interest to contemporary archaeology and of self-evident importance to our understanding of the human predicament, but it is their unusual conjunction at this point and period that makes this work especially significant. War has always hovered around scholarly discussions of how political complexity emerged and developed on earth, but for the most part research has focused on their synergy, on how war, its intensification, and the development of political complexity may have co-evolved. Obviously, there is another way to look at this--polities can “de-evolve” in addition to evolving, provoking a further set of questions about how war is causally or consequentially related to this “co-devolution.” This side of the coin has received rather less analytical attention, and what makes this case study so valuable is that it foregrounds this issue. What does the collapse of state level society and the intensification of war tell us about political and military process?

The climate changes that appear to have fueled the flight to pukaras ended in the 1300s. Yet the archaeological evidence shows that the Colla population remained ensconced on its hilltops well into the 1400s, even though climate records indicate that by this time precipitation was again more plentiful. What kept the people on these wind-swept, inaccessible heights? Arkush brings to this question an idea also recently advanced by Mark Allen for the Maori: whether or not war promotes political evolution, it can certainly also stymie the process. The physical and social legacy of warfare (fortifications, defensive settlements, and the formation of local alliances) can entrench political fragmentation. The argument is that small, exceptionally well-defended, independent polities become immune to political integration. Because they are all but impregnable, they are difficult to unite through military force; and, because their leaders know this, they are equally difficult to integrate through nonviolent political processes: why become subservient to the leader of another polity when that other leader cannot enforce his or her rule? Political disintegration gives rise to fortification and other defensive measures, which then sustain political disintegration: the system shunts itself onto a siding from which escape is difficult.

This is a powerful argument with considerable generalizability--the hilltop forts of Iron Age Europe and the mesa-top settlements of the desert southwest come immediately to mind as potential venues. But there is much else to applaud in this book. Chapter 3 provides an important foundation for future modeling of the relationship between political complexity and patterns of fortification and defensive location. Chapter 5 is especially rewarding for its innovative effort to diagnose political structure from an analysis of pukara “viewsheds” and its corroboration of these interpretations with overmapping of ceramic frequencies and of the colonial divisions imposed by early Spanish colonialists.

In an effort that will be especially welcome to social anthropologists, Arkush also makes an attempt to interpret the political form that these pukara polities might have assumed, and though there is much to consider in her book and many points worthy of discussion, it is this interpretation that I should like to pursue here. What we can infer from the evidence is that the permanent or semi-permanent residents of a pukara formed an independent political community that varied in membership from as few as seventy to perhaps more than a thousand members, with a few hundred being a plausible average. These political communities were areally clustered into what Arkush persuasively argues were defensive coalitions or alliance relationships varying in size from about three to twenty pukaras. Within a coalition, there may sometimes have been weak relations of dominance between smaller and large pukaras, and at times some coalitions may have banded together into even large ones. The archaeological evidence is thus very much in line with what we know of the general political structure of “tribal” or petty chiefdom types of society.

But how was this structure organized? Drawing on suggestions from Andean ethnographers and historical anthropologists, Arkush suggests that we may be looking at a “segmentary social organization” of the kind familiar from the work of Edward Evans-Pritchard and Marshall Sahlins. This is a system of “kin, ethnic, and political affiliation that follows a nested hierarchy, in which segments--small equivalent units such as villages or minor lineages--group themselves with others along lines of perceived genealogical relatedness to form larger and larger affiliation blocs: clans and tribal confederations” (p. 176). This is an intriguing and plausible idea, but the segmentary model carries some conceptual freight that deserves to be made more explicit if we are to get a closer hold on Colla political and social life.

As classically conceptualized, the segmentary model emphasizes genealogy as the structuring influence on small-scale social organization. The significance of genealogy could hardly have escaped the Africanists who generated the model. Among the Nuer, the Dinka, the Bedouin of Cyrenaica, the Tallensi, the Tiv, and so on, lineages were exceptionally elaborate structures, sometimes ten generations deep, with stipulated descent frequently adding a few more generational layers. Looking for the structural principles that governed small-scale society, Africanist anthropology found them in these genealogies: genealogical “principles” were the prime influence on multiple domains of behavior in these communities.

African models are certainly plausible ethnographic candidates for South American prehistory, but it is worth noting that the segmentary model hit something of a brick wall when attempts were made to apply it to New Guinea societies. In many parts of lowland New Guinea, descent groups were extremely small by African standards, and what united them into village communities was not genealogy but other devices such as male cults and settlement histories. The large clans and tribes of highland New Guinea better resembled the African model, but here too there were differences. The genealogies that united lineages into clans and clans into tribes were typically considerably shallower than those of African societies. More troubling, though, was the blithe disregard that New Guineans in general seemed to have for genealogical principles, at least in comparison to their African counterparts. In the lowlands, ethnographers sometimes encountered patrilineal clans in which 60 percent of members were not true agnates: invariably, they were referred to and addressed as “brothers” or “sisters,” but these were ideological not biological connections. In the highlands, non-agnatic rates were generally lower but still substantial , on the order of 25 percent.

Perhaps African segmentary systems were actually more like the New Guinea systems than early structural-functionalist ethnographers reported. Or perhaps the difference between the two ethnographic theaters was real. Whatever the case, Melanesianists spilled a lot of ink trying to make the African models fit the New Guinea situation before eventually surrendering. What did emerge from this hand-wringing, however, was the importance of place--location and residence--in structuring political and kinship relations in New Guinea. Quite apart from any genealogical connectedness, where one lived exerted a major influence on whom one considered co-citizens, allies, or kin.

Consider again the clans and tribes of the central and western highlands. Clans were the autonomous political communities, and tribes comprised a set of allied clans. Intratribal or intraphratry fighting between clans was limited warfare in the sense that it was intended to be either nonlethal or, if occasionally lethal, restricted to certain categories of target. “True” warfare was directed at other tribes or phratries, episodes of war alternating with periods of precarious peace. All of this closely parallels what Arkush deduced as Colla political structure, and in line with a segmentary model, descent groups within highland clans, and clans within highland tribes were united by ideologies of common descent.

But the influence of residence and location on the segmentary genealogy was as important as any influence that genealogy had on territorial contiguity. Warfare in highland New Guinea resulted in frequent population displacements, and these dislocations had genealogical consequences. Clans or clan segments forced to flee their tribal homelands gradually lost their genealogical connections to their “natal” tribe over the span of a couple of generations. At the same time, their new hosts were happy to accommodate these immigrants if they had the resources to do so, because numbers were crucial to military strength, which was in turn critical to survival in a hostile world. The integrating idiom was siblingship: so long as immigrants acted as members of their new clan, tribe, or phratry--contributing to their exchanges, fighting in its wars, and so on--they “were” members of the clan and tribe. To put this in another idiom, realpolitik was as much an influence on genealogical relationships as genealogical connection itself.

If the ethnohistorical and ethnographic evidence is correct that the Colla system was originally a segmentary organization, then further research might usefully consider whether it was more like the African or the New Guinea example, whether it was structured primarily by genealogical connection or whether residence and realpolitik were as important if not more so in structuring the system. This sort of fine-grained political question is exceptionally difficult to answer through archaeological research, though DNA work has experienced some recent success in reconstructing kinship relations from ancient skeletal material. Suffice it to say, if anyone can meet such a challenge, we should feel confident on the basis of this impressive piece of work that it is Arkush.

If there is additional discussion of this review, you may access it through the list discussion logs at: http://h-net.msu.edu/cgi-bin/logbrowse.pl.

Citation: Paul Roscoe. Review of Arkush, Elizabeth N., Hillforts of the Ancient Andes: Colla Warfare, Society, and Landscape. H-Urban, H-Net Reviews. February, 2012.
URL: http://www.h-net.org/reviews/showrev.php?id=33519

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.