The finding of cracked and calcined bones in some of the rooms brings up the question of the eating of human flesh by the people of this pueblo. There was no evidence of human bodies having been buried in rooms above the first floor and only portions of skeletons were in evidence in Rooms 61 and 80 which contained broken and charred bones. During the period of our work in Pueblo Bonito some of our Navajo workmen cleaned out a number of rooms in Penasco Blanco and in one of these a great many human bones were found. Some of these including portions of the skull, were charred, and the majority of the long bones had been cracked open and presented the same appearance as do the animal bones that have been treated in a similar way for the extraction of the marrow. It would therefore seem that these Pueblo Indians, either through stress of hunger or for religious reasons, had occasionally resorted to the eating of human flesh.
The report was published in 1920, and ever since then the question of cannibalism has hung over Chaco Canyon like a giant question mark. In the 1920s Frank Roberts found some additional bones at another site at Chaco that were in a similar condition, but these and the ones mentioned by Pepper constitute the entire sample of bones from Chaco itself that have been proposed as evidence for cannibalism. On its own, this is pretty weak stuff, especially since it comes only from sites that were excavated early in the history of archaeological research at Chaco and information on the context of these bones is very limited. Most archaeologists have therefore generally been content to conclude that whatever these bones represent, they don’t have much relevance to explaining Chaco as a whole.
Over the decades since Pepper wrote his report, however, a growing number of other sites throughout the northern Southwest have revealed human bones that are broken, burned, and otherwise suspiciously unlike typical burials. As archaeological techniques have improved, the amount of information about the context of these finds has increased, and as physical anthropologists have gained experienced and added new techniques more information can be gained from both these new finds and the early ones now in museums. One physical anthropologist in particular, Christy Turner of Arizona State, has put an enormous amount of effort into analyzing evidence for cannibalism in the prehistoric Southwest. His overall interpretation of the evidence has led him most recently to propose, particularly in the 1999 book Man Corn, coauthored with his wife, that cannibalism was a core part of the Chaco system, which he sees as a militaristic state led by some sort of sociopathic Toltec leader who came up from Mexico and attempted to institute a Mesoamerican state based on tribute and human sacrifice, or something (I haven’t read the book).
Obviously this theory depends heavily on attempts to equate the evidence for cannibalism across the Southwest with the rise of Chaco, and here it immediately runs into problems. While there is the evidence from Chaco reported by Pepper and Roberts, most of the evidence for cannibalism in the Southwest comes from contexts that are rather distant from Chaco both spatially and temporally. Some are earlier, particularly in the AD 800s and 900s, when the Chaco system may have been gearing up but was certainly not yet in the culturally dominant role it attained by the late 1000s. More are later, mostly in the AD 1100s, after Chaco’s influence declined. Spatially, most of the well-documented examples from recent examples are from southern Colorado, which was certainly under Chacoan influence at one point but probably was not at the times the events apparently occurred. Indeed, most of these well-documented examples are from a strikingly specific time and place: around AD 1150 in the area around Sleeping Ute Mountain, west of the modern town of Cortez, Colorado. There are other examples known from New Mexico and Arizona, but like the Chaco examples they mostly come from poorly documented early excavations and can’t be placed very well temporally.
According to one count, there are 32 sites in the Southwest that have yielded bones that may indicate cannibalism. Of these, 18 are in the greater Mesa Verde region in southwestern Colorado and southeastern Utah and can be dated with some certainty. Fully half of these sites, 9 out of 18, date to between AD 1125 and 1175. This striking spatial and temporal patterning has led some to suggest that these may reflect a single event of some sort that occurred in the context of the multidecadal drought of that period and the social disruption caused by the decline of the Chaco system at the same time. This idea was proposed in a talk I attended at the 2009 Pecos Conference. Several names were on the talk, including Turner’s (although he didn’t seem to actually be there), but much of the actual talking was done by David Breternitz, an archaeologist with long experience with the region. In addition to proposing that these assemblages indicating cannibalism or at least “folks treated inconsiderately” represented a single event that may have taken place over the course of mere weeks, days, or even hours, Breternitz noted the problematic connotations of using the word “cannibalism,” which has made the study of these sites a lightning rod for controversy as well as an irresistible temptation for sensationalistic treatments in the popular press. He referred to it as the “C word” and analogized it to the “N word,” which is not appropriate to use in talking to the NAACP (or, I would add, most other audiences).
Both the puzzling and the controversial aspects of the cannibalism issue are present in the case of one of the best-documented of these assemblages in the Cortez area. This is that found at site 5MT10010 near Cowboy Wash on the southern piedmont of Ute Mountain, which was excavated as part of a salvage operation in connection with an irrigation project of the Ute Mountain Ute tribe, on whose reservation the site is located. This site, which was described in an article in American Antiquity in 2000, was located in an area that was slated to be completely covered by a new irrigated field and was therefore completely excavated. It consisted of three pitstructures, like kivas but of clearly residential function, with associated surface masonry rooms, which do not seem to have been used residentially, outdoor activity areas, and trash middens. Two of these pitstructures contained assemblages of human bones that were heavily damaged and broken into hundreds of small pieces, some of which showed burning, cut marks, and other evidence of cannibalism similar to that found at other sites. These apparently represented the remains of seven individuals: two adult males, one adult female, one adult of uncertain sex but probably male, and three youths of uncertain gender aged approximately 14, 11, and 7. Their bones were strewn all over the two pitstructures and showed no grave goods or evidence of formal burial, in striking contrast to five individuals who had been formally buried in the trash mounds (which is a typical burial location for this area and period).
From the way the bones were broken and burned the authors concluded that all of the individuals in the pitstructures had been killed in a sudden attack at the same time, after which they were cooked and eaten and the site was totally abandoned and not reoccupied. In addition to the cut marks and burning on the bones themselves, two stone tools found in one of the pitstructures tested positive for human blood, implying that they had been used to chop up the bodies. Evidence for the sudden nature of the attack and the resulting abandonment consisted largely of the large number of artifacts left in place in all three pitstructures; normally when a site would be abandoned the people would take most of the usable artifacts with them, but in this case they were left in place.
As if all this wasn’t enough, the third pitstructure, which didn’t contain any bones except two which were heavily weathered and appeared to have been deposited naturally from the surface at some point after abandonment, did contain something even more interesting: a coprolite, unburned, in the hearth. Analysis revealed that the coprolite was human, apparently resulted from a meal consisting entirely of meat, and tested positive for the presence of a human muscle protein. This is about as close to smoking gun as it is possible to get in trying to establish if cannibalism took place in a given instance. Cut marks on bones can be and have been explained in various other ways, which are not generally very convincing but hard to disprove. This coprolite, however, seems to establish clearly that at least in this one instance, cannibalism did take place.
The authors, having established to their satisfaction that this assemblage does in fact represent an episode of cannibalism, went on to situate it in the context of both the local community and the larger region. The regional analysis basically pointed out the rash of likely cannibalism episodes around AD 1150, which may have been associated with the drought of that period, although it also noted the earlier episodes and the likelihood that this behavior constituted a longstanding pattern that for some reason became briefly intense and then subsided. They note that there is no evidence for cannibalism in the Mesa Verde area after AD 1200 (though this may not actually be true; more on that later) and very little evidence anywhere else in the Southwest after that date either. Whatever this event represents, it was clearly pretty temporary.
The community analysis was particularly interesting. This community on Cowboy Wash, which was established around AD 1125 and consisted of 10 sites, four of which have been excavated. Strikingly, all four of the excavated sites showed the same sort of evidence of cannibalism, strongly implying that they were all attacked and destroyed at the same time around AD 1150. This is in contrast to the earlier community at Cowboy Wash, which existed from about AD 1075 to sometime in the 1120s and was abandoned before the new one was established, probably by new people. Another striking characteristic of the newer Cowboy Wash community was the prevalence of ceramics made in the Chuska Mountain area to the south along the Arizona-New Mexico border. Chuska pottery was common at Chaco during its prime and has been found in various other areas as well, but it is extremely rare in the Mesa Verde region and the prevalence of it at Cowboy Wash strongly suggests that the inhabitants were either immigrants from the Chuskas themselves (pretty plausible in the light of the disruptions likely associated with the fall of the Chaco system, with which the Chuska communities were strongly associated) or maintained strong trading ties to Chuska communities rather than other local communities. Either way, the inhabitants of this community would likely have been easily identified as “outsiders” in the area, which may explain why they were targeted by others during this time of scarce resources, extended drought, and increased violence. The Cowboy Wash community was not in a defensive location and does not appear to have had any defensive features such as walls. Its fate may explain why such locations and features became so common not long afterward.
Of course, with something like this there will always be objections, and a critique of the claim of cannibalism by two archaeologists who have been closely identified with the interests of the modern Pueblos, Kurt Dongoske and T. J. Ferguson, along with physical anthropologist Debra Martin, accompanied this article when it was first published. The three of them basically criticized everything they could about the presentation of data and argumentation in the article on the Cowboy Wash site. Some of the criticisms were reasonable, including a caution about being too quick to talk about cannibalism and to keep in mind the likely effect of this in the popular press and on modern Native Americans. Others basically amounted to nitpicking about the presentation of data, often appearing to demand a level of information more appropriate for a full site report than a journal article. In a response, the authors gave more information on the analysis and tried to counter accusations that the blood residue and coprolite analyses in particular may have been contaminated or otherwise problematic. They also conceded some of the points raised in the critique, especially about the importance of thinking about the effect of this kind of research and the way it is presented in public discourse. This response seemed pretty convincing to me. I’ve heard that the coprolite evidence in particular has been challenged, but I’m not sure if any more substantial objections than the easily defused ones in the initial critique have been raised.
Another interesting thing about the critique and response, to me, is the way the critique basically accuses the authors of the initial article of being insufficiently “scientific” in not rigorously testing alternative hypotheses. This is interesting largely because it clearly falls on the “anthropology is and should be a science” side of the “is anthropology a science?” debate, even though these particular anthropologists are coming from a perspective that is usually associated with the other side in their close association with the modern Pueblos and appreciation of their viewpoints. In this context they seem to be mostly raising the science issue as a club with which to (rather ineffectively) attack the initial article, but the use of this approach is still noteworthy. The response basically runs with this assumption and argues that, no, the analysis really was very scientific and they didn’t include all the data and hypotheses because of the space constraints of the journal article, but it’s all in the full report. Both seem to think of “science” mostly as a methodology, particularly associated with hypothesis testing, rather than a body of knowledge, which is also interesting.
Another thing the critique mentions is the desirability of more comprehensive approaches to the issues of violence, migration, warfare, and so on. The critique authors say this in the context of criticizing the initial report for focusing too narrowly on cannibalism specifically, which sits oddly with their criticisms of it for not being specific enough in describing the physical evidence, but this is one of their better points. They specifically mention the possible role of ethnicity in structuring some of these dynamics, and cite the paper on ethnicity at Wupatki by Glenn Davis Stone and Christian Downum that I recently discussed. I know of at least one recent paper that has done just that; I’ll have more on it later.
So what does all of this imply for Chaco? Not much, really. The authors of the initial Cowboy Wash paper note that to the extent that the Cortez-area spate of cannibalism episodes around AD 1150 has anything to do with Chaco, it is most likely in representing one aspect of the social chaos that may have followed the decline of the Chaco system after AD 1130 or so. They also note that Cowboy Wash is about as far from a Chacoan outlier as it was possible to get in the region during this period. The closest proposed outlier is Yucca House, on the other side of Sleeping Ute Mountain toward Cortez, the outlier status of which is rather questionable since it is unexcavated and also has a large later occupation that makes it difficult to identify any Chacoan parts. And speaking of the location of Cowboy Wash, it is important to note that the southern piedmont of Sleeping Ute Mountain is remarkably desolate even by Southwestern standards and would have been very marginal for agriculture. It doesn’t get enough rain for dry farming, so any farming the people at 5MT10100 would have done would have to have been floodwater farming using the wash itself, which would be very vulnerable to fluctuations in rainfall and other climatic conditions. This probably explains why the area was colonized and abandoned several times during prehistory. The time of the attack appears to have particularly bad for the inhabitants, as there was minimal evidence of domestic plants and most of the plant remains found at the site were of wild plants that would have been gathered in the spring to compensate for a poor harvest in the fall.
And what about the bones mentioned by Pepper at Chaco itself? Hard to say. Turner has reexamined some of them and concluded that they do in fact represent evidence of cannibalism, but then again he would conclude that. The fact that we don’t really have any more detailed information about their contexts within the site than the minimal information given by Pepper makes it very difficult to place them within the context of Chacoan prehistory as we know it today. Judging from the better information we now have from Colorado, I think one possibility is that the Chaco bones represent something similar that happened around the same time. The canyon was not totally abandoned after it declined in regional importance around 1130, and it might have participated in some way in the cannibalistic warfare that apparently took place after that, though whether as victim or perpetrator is hard to say. Whether or not this is the actual explanation, it is important to remember that the fact at a big, complex site like Pueblo Bonito not everything found somewhere in the is necessarily closely connected to the original purposes for which the site was built. Things change, after all. Furthermore, Pueblo Bonito has hundreds of rooms, almost all of which have been thoroughly excavated, and possible evidence of cannibalism has been found in two. That’s a pretty thin reed on which to hang a theory postulating that Chaco was all about cannibalism. Some cannibalism may well have taken place at Chaco, but there is really no reason to think that it was a major part, or indeed any part, of the reason Chaco was built or became so influential. At least, that’s the way I see it. Merry Christmas.
Billman, B., Lambert, P., & Leonard, B. (2000). Cannibalism, Warfare, and Drought in the Mesa Verde Region during the Twelfth Century A.D. American Antiquity, 65 (1) DOI: 10.2307/2694812
Dongoske, K., Martin, D., & Ferguson, T. (2000). Critique of the Claim of Cannibalism at Cowboy Wash American Antiquity, 65 (1) DOI: 10.2307/2694813
Lambert, P., Leonard, B., Billman, B., Marlar, R., Newman, M., & Reinhard, K. (2000). Response to Critique of the Claim of Cannibalism at Cowboy Wash American Antiquity, 65 (2) DOI: 10.2307/2694066