Virgin Soil, Widowed Land: A Curious African Counterpoint

Delaney Park at Noon on the Winter Solstice, Anchorage, Alaska

Happy Solstice! I’m still working on continuing my series of posts on the study of Native American population collapse, but as part of this project I’m also reading about other parts of the world that share some of the same kind of history. Lately I’ve been reading a lot about Africa, which has a fascinating history that overlaps with that of the Americas in some respects but not others, and I was struck by a remarkable difference in how the topic of historical demography has been handled by Africanist scholars in the same period when it became a very large focus of attention for Americanists (i.e., the 1960s and 1970s).

Briefly, while there is historical evidence of population declines in at least some parts of Africa of the same magnitude seen in the Americas (50-90% in some cases), the implications of this decline for the societies in question seems to have aroused little interest from scholars. These declines are often mentioned offhand in footnotes or briefly in a text that goes on to discuss other historical or anthropological topics. It’s not that the scholars are dismissing the veracity of either the population numbers or the magnitude of decline mentioned in their sources, as was the case with earlier generations of Americanist scholarship. Instead they just don’t seem to care at all or to see these declines are relevant to their main interests.

But these are huge declines, often over very short periods! For example, the area that is now the Republic of Sudan is said to have lost something like 80% of its population to a combination of warfare and disease in the few years of the Mahdist uprising in the 1890s. That’s enormous! And very recent, so much so that’s the figures are likely to be pretty accurate in the absence of evidence to the contrary. But they’re so huge that it seems likely that they are exaggerated, and I’d be interested to see some informed source criticism looking at how reliable these numbers are. Studies like this from the same period, and more recently, doing just that are common in the Americas but I have hardly found any for Africa covering any period. One interesting study of the coastal Mpongwe people of Gabon, who were in close contact with Europeans for a long time, concluded that the dramatic population collapse reported was in fact plausible, due to similar contact conditions as typically obtained in the Americas (but not in most other parts of Africa). But otherwise this type of study seems to be very rare.

Not that demography was not of interest to Africanists at this time, mind you. But the main focus was on trying to quantify the impact of the slave trade to the Americas, a very different sort of concern from those of Americanists. The background here is that there was a persistent narrative, fostered especially by nineteenth-century abolitionists, that the slave trade was leading to massive depopulation of African people, with subsequent negative effects on societies within Africa. There were counterarguments and the debate went back and forth for decades, forming a crucial backdrop to the advent of direct European colonization and “pacification” of the continent, which was often justified in the name of eliminating slavery.

By the 1960s this age-old debate was starting to get some rigorous quantitative attention. The key figure here is Philip Curtin, a pioneering economic historian of Africa who was then at the University of Wisconsin. In 1969 he published a book called The Atlantic Slave Trade: A Census, which was as revolutionary in the study of the African slave trade as Henry Dobyns’s work on Native American population around the same time was on that field. Curtin tried to compile and evaluate whatever quantitative information was available to come up with as rigorous a set of figures as possible for the trade over the course of centuries.

Unlike Dobyns’s work, however, Curtin’s numbers ended up being significantly lower than previous estimates, which may well have been exaggerated by abolitionists trying to stir up opposition to the slave trade. This caused some controversy and pushback, especially by some African researchers who took issue with aspects of Curtin’s methodology. But overall Curtin’s figures were accepted by the field and have formed a baseline ever since for those trying to evaluate the impact of the trade.

So what was that impact? Well, it’s still hard to say. Curtin’s lower figures suggest that the direct demographic impacts of the trade were likely limited. This is not to say that the reorientation of many societies especially in West Africa, the main area of focus of the Atlantic trade for most of its existence, didn’t have important social and economic impacts. Both the exporting societies on the coast, which tended to be complex state-level entities that increased in complexity over time, and the main supply areas in the interior, which tended to be somewhat less complex, were clearly affected in various ways by slave-raiding and -trading. The exact extent of these effects and the way they shifted over time is the subject of a vast literature of which I have been able to read only a tiny portion, and there doesn’t seem to be much consensus.

One key difference between Africa and the Americas, of course, which lurks in the background of all comparisons like this, is that the “indigenous” people of most parts of Africa managed to survive both the slave trade and the period of colonialism that followed, and the continent is now divided overwhelmingly into independent post-colonial states. European colonization only took place on any significant scale in a few places, and of those societies still fewer managed to persist past independence. The Americas are like a mirror image of this situation, with only a few areas still maintaining a significant Native population and the rest of the hemisphere dominated by settler societies (which may have significantly mixed ancestry of course). Clearly these contrasting modern situations have colored the interests of scholars in complex ways, especially in a period that saw both the fall of the great colonial empires in Africa as well as a revival of interest in Native American history and traditional culture. Why the patterns of scholarship ended up how they did exactly is still a bit of a puzzle to me though.

Read Full Post »

Virgin Soil, Widowed Land: High-Count Hegemony

Arizona Welcome Sign

If Sherburne F. Cook was the dominant figure in the “Berkeley Era” of the study of Native American historical demography, his equivalent in the next period was undoubtedly Henry F. Dobyns. Dobyns was a complicated, controversial figure and his legacy today is decidedly mixed, but his importance to the intellectual history of this issue is unquestionable. It is no exaggeration to say that he single-handedly launched the topic into the scholarly limelight, and he continued to pursue it for decades even as questions about his own methodology and approach began to spur a backlash. Although the history is complicated, for convenience and in recognition of Dobyns’s importance I’ve defined this period as bracketed by what are probably his two most important publications on the subject: his 1966 paper in Current Anthropology that first drew extensive scholarly interest to the subfield and was enormously influential in defining it, and his 1983 book that, while also very influential in some circles, in others was considered a major overreach that threw his whole approach into question.

In between, numerous scholars from a wide variety of disciplines investigated the question of population history and the impact of epidemic disease with an unprecedented fervor. Most of these studies, certainly the most influential ones, ended up falling on the “high-counter” side of the demographic debate. With its obvious political implications during a period of intense political awareness, activism, and dispute, there was a clear sense that the high-counter position was ascendant and revolutionizing the whole world’s understanding of the tragedies of the past and how they led inexorably to the inequities of the present.

As I’ve noted before, though, this is an oversimplification. A parallel tradition of low counts continued as an undercurrent to the high-count hegemony, mostly focused in certain disciplines and institutions. This current would emerge into higher prominence in the 1980s and afterward as the work of Dobyns and his acolytes started to be questioned by a wide variety of detailed empirical studies of particular regions. Ironically, this turn in the scholarship would come as the earlier high-count research was just beginning to be incorporated into more popularized accounts which made it better known among the general public.

This chapter of the story is in some sense the heart of it. In some ways it’s a story of scholarly hubris (particularly on the part of Dobyns personally) that eventually led to a painful reckoning, but again that is too simple. The high counters of this era made enduring contributions to understanding of this issue even if not all of their specific conclusions have stood the test of time. In a qualitative sense, at least, even their strongest critics would mostly concede that they were closer to capturing the historical reality than most of their predecessors. Their greatest flaw, perhaps unsurprising in the context of the mid-century modernist context in which they worked, may have been an excessive confidence in the potential for academic research to provide specific, detailed answers to complicated historical questions.

Dobyns was born and raised in Arizona, and his early anthropological work focused mainly on the tribes there, including providing support to their land claims cases against the federal government in the 1950s. After getting his Ph.D. from Cornell University in 1960, he was hired by Cornell and went to Peru to work on the Vicos Project, an innovative and groundbreaking exercise in applied anthropology. Both in Arizona and Peru, Dobyns also did research in the colonial and mission archives and developed an interest in historical demography. He noticed very severe population declines in the records of both areas, and began to develop a theory for what was behind them that would end up being enormously influential in anthropology and beyond.

Dobyns published some early work on the population dynamics of southern Arizona, but his first widely influential publication was a 1963 article on the history of epidemics in Peru. This article was foundational to the emerging interest in disease history and is still widely cited today. His most important early work, however, was the aforementioned 1966 Current Anthropology article in which he reviewed previous approaches to the issue of precontact population and introduced a new methodology and estimate of his own.

The majority of this article consists of a thorough and detailed analysis of all the methodologies that had been used to estimate pre-contact Native American populations. Dobyns convincingly demonstrates that the various techniques used by the “low-counter” school of Mooney and Kroeber likely underestimated numbers by disregarding the accounts of contemporary observers and relying overly on extrapolations from later ethnography. He is more positive about the innovative estimation and extrapolation methods of Cook and others in his circle at Berkeley. In addition, he forcefully points out the importance of disease in affecting population numbers, a factor largely ignored by the previous researchers he cites, with the partial exception of Cook. In fact, Dobyns discusses in detail Cook’s monograph on the malaria epidemic of the 1830s in Oregon and California, a very important study still widely cited today. Dobyns specifically emphasizes two important but underappreciated aspects of disease in relation to demography:

Cook’s analysis of this California epidemic demonstrated the operation of two very important processes in the human ecology of aboriginal American populations. First, he showed the magnitude of mortality which a single epidemic can cause in a non-resistant population. Second, he called attention to the biological fact that epidemic infection is not limited to tribal populations in immediate face-to-face contact with Europeans. The decimation of native Californians was not limited to missionized Indians, but extended outward as far as disease agent and vector could spread infection from intrusive (white) carriers to aboriginal populations. It is necessary to maintain constant awareness of these two processes or fundamental trends among natives of the New World. Any interpretation of reported native populations during the early years of contact with Europeans which ignores the tremendous mortality caused by epidemics inevitably underestimates the size of the aboriginal populace.

Dobyns would continue to emphasize these factors throughout his career.

Dobyns’s analysis of past work is highly convincing in showing the flaws that led to past underestimation of population numbers, and this accounts in part for the influence this paper has had on subsequent scholarship. He went beyond this, however, and also attempted to devise a new methodology which would avoid those flaws and provide a sounder basis for making estimates. Here he starts to make some assumptions and interpretive leaps that would ultimately lead him into some methodological flaws of his own.

Dobyns starts by defining a “depopulation ratio” comparing the population of a given group at two times in its history. The two times he is particularly concerned with are the group’s precontact population, or as close to it as is possible to get, and the time when it hit its nadir of population and began to grow again (as most Native populations eventually did). He doesn’t really justify his use of these two endpoints or the generalizability of a ratio derived from them. He looks at several well-documented cases, including central Mexico, the Andes, California, the Amazon, and the Piman-speakers of southern Arizona and northern Sonora, and derives to his satisfaction a ratio of 20:1 as a “sound, if perhaps conservative, tool to employ as a hemispheric minimum.” He considers 25:1 an alternative ratio to derive a reasonable range. Using these two ratios and available data on population nadirs, he comes up with an aggregate hemispheric estimate of 90,043,000 to 112,553,750 for the whole Western Hemisphere at contact. Kroeber had estimated 8,400,000.

A hundred million people! This conclusion was striking and contributed to the influence of Dobyns on further research. He notes in the paper that Cook’s Berkeley colleague Woodrow Borah had recently come to a very similar estimate, so he was not totally on his own and in some ways his estimate was a plausible expansion of the increasingly high counts that were coming out of Berkeley in the wake of Cook’s pioneering research. Much of that research was in relatively obscure regional publications, however, and it was Dobyns who brought it to one of the flagship publications of American anthropology and garnered a large audience. As it typically does for its major paper, Current Anthropology sent the paper around to a wide variety of other scholars and published their comments, which were mostly very positive, with occasional reservations on particular points. The combination of bold thinking with seemingly innovative quantitative techniques fit well with certain tendencies within the social sciences at mid-century.

They also fit with an increasing political awareness of the plight of modern Native people, along with the general political tumult of the 1960s. Dobyns himself was a strong advocate for Native rights, going back to his land claim days, and his concept that precontact Native populations were large and (implicitly) successful, only to lose 95% or more of their people to disease introduced by Europeans, fit well with the political mood of the times.

The following year, Dobyns’s conclusions were bolstered by a paper by the historian Alfred Crosby documenting the smallpox epidemic that accompanied the Spanish conquistadors into central Mexico and, per Dobyns’s earlier Andean epidemic research, probably continued to spread ahead of them into the Inca empire, killing the emperor and fatally weakening the empire itself. This was a concrete example of the dynamics of disease that Dobyns had discussed in general terms, with very clear and dramatic consequences for the course of world history. Crosby would later go on to expand his argument in his groundbreaking 1972 book The Columbian Exchange and a 1976 paper outlining the specifics of how “virgin soil epidemics” contributed to depopulation in the Americas.

Dobyns’s and Crosby’s ideas rapidly caught fire in all sorts of disciplines, and led to innumerable studies of specific areas in the coming decades. The historian Wilbur Jacobs, whom we encountered before through his oral history of the Berkeley medical school and its treatment of S. F. Cook, wrote an important overview of the implications for the history of contact. Other studies, too numerous to list, looked at the issue from the vantage points of geography, anthropology and many other disciplines. Throughout the heyday of 1970s the high-count school of Dobyns and Crosby expanded its reach and refined its estimates and arguments.

There were some dissenting voices even at this time, however. The low-counters may have seen their influence eclipsed, but they didn’t disappear, and as I’ve noted before they were particularly concentrated in certain disciplines.

In direct response to Dobyns’s article, the Andeanist C. T. Smith published his own in Current Anthropology in 1970. Following up on some mild criticism by the geographer William Denevan in his comments on Dobyns, Smith notes that the depopulation ratio of 20:1, if truly an average of well-documented cases, probably conceals considerable variation in the underlying data set. With regard to the colonial Andes, one of Dobyns’s most important cases for determining the ratio, Smith points out that the coastal regions Dobyns used in his analysis have markedly higher depopulation ratios than the interior regions in the same data; combining them reduces the depopulation ratio to 4:1. Smith is not necessarily opposed to Dobyns’s ideas, however, and he merely suggests that the dynamics of coastal and interior populations in this region should be considered separately. He also does a detailed analysis of additional colonial census data to further refine the depopulation estimates, and Dobyns’s comments on his paper are largely positive.

One notable discipline that was strikingly absent from the study of this demographic topic is, perhaps surprisingly, demography. The reasons for this were made clear in a 1975 paper, again in Current Anthropology, by the demographer William Petersen. In an astonishing display of saying-the-quiet-part-loud, Petersen took aim at a wide variety of demographic interpretations widespread in the study of prehistory. This mostly takes the form of noting the very slim data available for various methods of estimating prehistoric population parameters. With regard to catastrophic population loss in the wake of contact, he admits that there certainly have been losses, and does seem to agree with Dobyns in giving historical accounts more credence than Kroeber did, though he also notes Smith’s clarification about the differences between coastal and interior populations in the Andes. Overall, he maintains a skeptical position and notes the contemporary political implications of positing much larger precontact populations, which he also analogizes to the nineteenth-century “Mound Builder” legends to explain the mounds of the midwestern US. In his response to what he terms Petersen’s “pontifical perambulation,” Dobyns attempts to refute his accusation that high-counters are political opponents of modern liberal capitalism by noting that he himself invests in the stock market and admires some companies. This is not very convincing in substance, but the polemical tone foreshadows Dobyns’s reactions to subsequent criticisms.

Archaeologists and physical anthropologists also continued to be skeptical about the ideas coming largely from historians and geographers. Many of these holdouts were associated with the Smithsonian Institution and carried on the legacy of Mooney, in some cases quite directly. Douglas Ubelaker, who worked on the editing of the Smithsonian’s updated Handbook of North American Indians, was in part responsible for updates to Mooney’s demographic estimates included in the original version of the Handbook. In an interim report published in 1976, he noted that the estimates for specific tribes and regions submitted by that point were higher than Mooney’s, but well below Dobyns’s. The total estimate approximately double’s Mooney’s total, which implies a hemispheric total of about 16 million. A significant difference indeed!

These quibbles were fairly minor during this period, however. The event that started to bring them out more forcefully was, perhaps ironically, Dobyns’s publication in 1983 of his book Their Number Become Thinned, which focused on the Timucuan people of northern Florida but served also a vehicle for him to promote his revised and expanded methodology for population estimates and the role of disease in depopulation. By this point the high count hegemony had reached its breaking point.

Read Full Post »

Virgin Soil, Widowed Land: The Berkeley Era

California Welcome Sign

The study of Native American depopulation and the role of epidemic disease in it began in earnest around the beginning of the twentieth century, and from that point until 1966 it remained a relatively obscure niche topic across several different disciplines. Interestingly, the few scholars interested in this topic, regardless of disciplinary background or specific position on the substantive issues, were overwhelmingly concentrated at a single institution: the University of California at Berkeley.

It’s not entirely clear why Berkeley became the focus of study for this topic, but the precipitous decline in numbers of the California Indians during the nineteenth century, which happened right before the eyes of many of the early white settlers, had spurred more interest in the general topic there than elsewhere starting with the very earliest attempts at professional anthropology. A key early attempt to estimate the numbers of inhabitants in aboriginal California was published in 1905 by C. Hart Merriam. Merriam noted that the part of the state under the influence of the Spanish missions, which he estimated to be approximately one-fifth of the non-desert area, had relatively good demographic data provided by the missionaries. He further assessed, based on his personal experience doing fieldwork throughout the state, that the resource base of the entire non-desert area was pretty similar and could presumably support a similar population density (though he acknowledged that there was no evidence that it actually did). He therefore took the demographic data from the mission records, adjusted it to account for unconverted Indians within the missionized zone, whom he estimated on no evidence to form one-quarter of the total population, and multiplied it by five to come up with a total estimate for the non-desert portion of the state.

For 1834, the year for which Merriam considered the mission data most complete and reliable, he estimated 30,000 Indians in the missions, 10,000 unconverted Indians in the mission zone, and a non-desert state total of 200,000. He further estimated the desert population at 10,000 (again with no basis), for a total within the current state boundaries of 210,000 in 1834. He further noted the decrease of population over the course of the mission period and estimated a total population of 260,000 as of first contact in the late eighteenth century.

Merriam further looked at estimates for Indian population made by various officials during the American period, starting in the 1850s and made some adjustments for under-counting of Indians living away from reservations. He concluded that the population decline had been dramatic, resulting in a 1900 estimate of just 15,500 people. He identified two periods of particularly sharp decline, following the secularization of the missions in 1834 and the start of the Gold Rush in 1848. These two periods overlapped, and Merriam attributed them to similar causes, primarily the greed and oppressive behavior of white settlers, but he saw them as involving distinct groups of both Indians and settlers (Spanish-Mexican ranchers along the coast in the first case, Anglo and international gold prospectors further inland in the second). He mentioned disease in passing as one factor leading to the demographic decline, but didn’t emphasize it.

Merriam’s methodology was fairly crude by modern standards, but it was pretty sophisticated for its time. It’s interesting to note that Merriam combined aspects of what would eventually come to be the two main methodological approaches to estimating pre-contact populations: working backward from ethnohistorical documentation and estimating carrying capacity of particular types of land based on subsistence resources available.

Soon after Merriam’s work was published, James Mooney at the Smithsonian Institution began work on a more ambitious project to estimate contact population for all of North America. A brief summary of this work appeared in the Bureau of American Ethnology’s first Handbook of North American Indians, but Mooney never published his intended full monograph due to his untimely death in 1921. An abbreviated version based on his notes was published posthumously by the Smithsonian in 1928. Mooney primarily made his estimates based on the earliest surviving population counts of reasonable reliability for each tribe, with adjustments to get from there to an estimate of population at whatever date constituted “contact” for the area in question (ranging from 1600 to 1780). Mooney took historically recorded epidemics into consideration in making these adjustments, along with warfare and other factors. For California, however, unlike every other region, Mooney did not make his own estimates but adopted Merriam’s.

Merriam wasn’t based in California, but his pioneering fieldwork there was influential on the development of a local tradition of anthropology at Berkeley, which was developed largely by Alfred Kroeber, a towering figure in American anthropology generally. Kroeber studied a wide variety of topics, and Native American demography was one of them. Around the same time Mooney was working on his estimates for the “Population” chapter of the Handbook of North American Indians, Kroeber began work on his own using a similar methodology for the California chapter of the same publication. Working backward from the earliest solid counts in the ethnographic record, Kroeber came up with a count of approximately 150,000 at the time of contact. This is a much lower number than Merriam’s, and over the years Kroeber became even more conservative in his estimates. By the time he published his own Handbook of the Indians of California in 1925 his overall estimate had declined to 133,000, approximately half of Merriam’s number.

In 1934 Kroeber published an article discussing Native American contact-era population for all of North America, in which he adopted Mooney’s estimates for most areas but substituted his own estimate for California in place of Merriam’s. This reduced the overall continental estimate a bit, and Kroeber stated in the article that he was using Mooney’s estimates but that he thought they were likely a bit high and would come down as more research was done. John R. Swanton, the Smithsonian anthropologist who edited Mooney’s work for the 1928 publication, had a similar opinion, which he expressed in the footnotes at various points.

All this makes Mooney, Swanton, and Kroeber the main founders of the “low counter” school of thought on these issues. Due to Kroeber’s towering reputation within the discipline of anthropology and his prominent post leading the anthropology department at Berkeley, the low count position would come to be popular in many circles among anthropologists for decades to come. This was particularly true among archaeologists and physical anthropologists, who also had other reasons based on their own research and (sub-)disciplinary perspectives to incline toward low counts. Similarly, Mooney and Swanton’s legacy led to a longstanding tendency toward low counts among Smithsonian Institution anthropologists, again especially among physical anthropologists and archaeologists.

At the same time all this was going on in the 1920s and 1930s, however, a very different perspective on population counts and demographic decline was developing among a different set of researchers, again with Berkeley as a major base and California, along with Mexico, as a major field of investigation. This perspective, which would develop into the “high counter” school with a focus on environmental carrying capacity, was largely led by geographers, with the most prominent figure being Carl O. Sauer at Berkeley’s geography department.

Sauer’s role in the development of geography was parallel in many ways to Kroeber’s in anthropology, and his personal research interests were equally broad as well. He had a particular interest in Mexico, and it was his research in the 1930s with Donald Brand (also known for his research on Chaco Canyon around the same time) on the historical demography of northwestern Mexico that set the tone for the school of thought that would follow him on this topic. After reviewing the available historical and archaeological data available at the time, Sauer and Brand concluded that northwestern Mexico had been home to approximately as many people in pre-contact times as in their own time. Given the historical evidence for much lower populations in the initial centuries after Spanish contact, this implied an immense decline in population after contact that contrasted strongly with the interpretations of Kroeber et al. that posited low pre-contact populations and substantial continuity in demographic trends across the boundary of contact. Sauer’s students in geography at Berkeley, and their own students there and at other institutions, would go on to develop his ideas over the next few decades, with the result that geography would become a bastion of “high-counter” thought just as archaeology and physical anthropology would become centers for “low-counter” thought.

The most prominent high counters to emerge in Sauer’s immediate wake, however, were not geographers at all, though they were still associated with Berkeley. They included the Latin Americanist historians Leslie Byrd Simpson and Woodrow Wilson Borah, but the most prominent figure was a physiologist at Berkeley’s medical school named Sherburne Friend Cook.

Cook is a complex figure with an immense but ambiguous impact on the field of Native American historical demography. His eclectic interests and long career at Berkeley echo those of Kroeber and Sauer, but unlike them he was a marginal figure in his own discipline. He received his Ph.D. from Harvard in 1925 with a dissertation on the toxic effects of heavy metals, and after being hired at Berkeley a few years later he continued to study that topic. He was something of a pioneer in studying the physiological effects of environmental contaminants and his early career looked bright.

As Cook continued his studies on heavy metal toxicity in the early 1930s, however, he veered into political ground equally toxic to his career. He documented the presence of heavy metals in chicken feed and began to trace them up the food chain to the chickens and then to the people who ate them. The public health implications of this line of research were considerable, but so were the economic implications to chicken feed manufacturers. It appears (based on later oral history research among Berkeley academics by the historian Wilbur Jacobs) that those manufacturers influenced a dean at the medical school to try to prevent Cook from publishing his results and to go on to use administrative chicanery to sabotage his career for many years.

Discouraged, Cook began to divert his attention away from physiology and explore other fields of interest. He somehow stumbled upon a transcript in Berkeley’s Bancroft Library of an eighteenth-century description of the diseases of the Indians of Baja California written by a Jesuit missionary, which he translated and published in 1935 in a local medical journal. He went on to publish various other odds and ends of medical history and related topics over the course of the next few years. He developed an interest in Indian demography and population history, and was in friendly contact with both Kroeber and Sauer as they did their studies on this topic in the 1930s.

Cook was an astonishingly productive and creative researcher, and he conducted numerous studies of Indian population history, along with many other topics, over the remaining forty years of his life. He developed numerous ingenious methodological approaches to try to wring population estimates out of the most unlikely sources, including Aztec tribute lists and various environmental productivity estimates in California. His methodology was constantly being adjusted, so his specific population estimates for given areas varied from publication to publication, but in general he was coming up with high numbers more in line with Sauer’s results than Kroeber’s. He also did extensive studies of disease history, including a very important 1955 publication on the “fever and ague” that swept California and Oregon in the early 1930s in which he argued, contra earlier researchers, that it was most likely malaria and its extreme death rate among Indians was due to it being their first exposure to the pathogen. (An early glimpse of the “virgin soil” concept that would become so influential later.)

Cook’s research was extraordinarily wide-ranging, and some of it would be controversial on a variety of grounds especially after his death. Many of his ingenious methodologies for estimating populations relied on assumptions that didn’t hold up well to closer investigation, and his conclusions about the oppressive conditions in the California missions were vigorously contested by pro-mission scholars, especially those affiliated with the Catholic Church. He stands as a towering figure in the field of historical demography of Native Americans, however, and set the tone for the emergence of that field as a focal point of scholarship.

Cook’s final estimate for the contact-era population of California was 310,000. This was somewhat higher than Merriam’s estimate of 260,000, and much higher than Kroeber’s 133,000, putting Cook firmly on the “high-counter” side. The influence of Cook and his collaborators, while fairly limited in what was still an obscure field of study in the 1940s and 1950s, would expand dramatically in the 1960s as a new generation of researchers moved the topic out of the back halls of Berkeley and dramatically into the academic and political spotlight.

Read Full Post »

Virgin Soil, Widowed Land: Overview and Scope Update

Clinic, Pilot Point, Alaska

Well, it’s certainly been a while since I introduced this series! The coronavirus hasn’t gone anywhere, though, and neither have I. I’ve been reading lots of fascinating stuff on New World (de)population and disease history, and I have a pretty good idea of the way the series is going to look overall. I’m still not sure how long it’s going to take (at least several months) or how many posts it will ultimately include. I have enough of a sense now, though, to give a tentative outline of the topics I intend to cover, and that’s what I’ll do in this post. I’ve also decided to make a couple changes to the scope of the series, which I’ll also discuss here.

To take the latter issue first, I initially said I would limit the scope of the history I’m looking at to exclude the Black Death of the fourteenth century and the 1918 “Spanish Flu” pandemic, with the idea that the epidemics of interest for Native American population mostly occurred between the two, each of which has been the subject of such an immense scholarly literature that it would be easy to get bogged down. After digging a bit further into the overall topic, though, I realized there were good reasons to look at both these pandemics, and indeed others both earlier and later. The issue of getting bogged down is a real one, which I’ve managed to mostly avoid with the Spanish Flu but less so with the Black Death. Still, so much of the scholarly literature on disease history and the general impacts of epidemics involves or relies on study of these most prominent examples that it doesn’t really make sense to arbitrarily exclude them.

Furthermore, looking both earlier and later than the core period between these two pandemics turns out to be pretty important. Earlier, there were some important epidemics worthy of study for many reasons, including the light they may shed on the history and evolution of certain diseases (especially now that ancient DNA analysis techniques have reached the point that specific pathogens can be unambiguously identified in ancient remains). Later, there were some specific epidemics postdating 1918 that turn out to be very important for the scholarly history of “virgin soil” epidemics and depopulation, in part because they were directly witnessed by doctors trained in modern scientific medicine. Of these, the most important are a series of epidemics in the Yukon Territory associated with the building of the Alaska Highway in the 1940s and another series in the Amazon in the 1960s. So I’ve essentially abandoned any hard temporal bounds on the scope of the series, although the main focus will of course be on the period from roughly 1500 to 1900.

Okay, on to the outline. To anticipate the overall conclusions a bit, despite a long history of research on the topic of Native American depopulation and the role of epidemic disease, I’ve found that there’s never really been a solid consensus about anything related to these topics and such a consensus is, if anything, further away now than it’s ever been. There have been general trends in the popularity of certain interpretations and methodological approaches, which can be (and have been) conveniently categorized into periods of one approach or another being dominant, but a closer look shows that there has always been a diversity of views that really cluster most clearly by academic discipline. I’ll get into this in more detail later, but the general idea is that rather than “high counters” and “low counters” exchanging periods of hegemony, there have always been both, concentrated in their own disciplinary zones, though their influence as measured by the spread of their ideas into other disciplines does show a certain back-and-forth pattern over time.

With that overall idea in mind, and despite my skepticism about a general chronological patter being most important, I do intend to structure the first part of the series chronologically. I’ll have posts for each major “era” in the modern study of these topics, which will hopefully give a sense of the major players and their ideas through time. After that overview, I’ll do multiple subseries of posts looking at the question from our current perspective, taking account of the various contributions from different disciplines.

These subseries will include one organized nosologically, or by the modern categorization of diseases and the pathogens that we now know to cause them. Another subseries will be geographic, looking at the different regions of the New World and what we know (and don’t know) about their population and disease histories. I may also do a third subseries looking chronologically at the big picture of population and disease throughout the different periods of contact and colonization in the Western Hemisphere, but I’m not as sure about that one yet.

So that’s the general idea. As I mentioned above, the series will definitely take a minimum of several months (it’s already taken more than three months to get to this second post!), but I don’t know how long it will ultimately take or how many posts it will involve. It’s a big but fascinating and important topic.

Read Full Post »

Virgin Soil, Widowed Land: Introduction

Chief Andrew Isaac Health Center, Fairbanks, Alaska

Today is the summer solstice, which is an event I usually mark with a blog post about archaeoastronomy. Today I’m going to do something a little different, though. Since the coronavirus pandemic has been dominating and reshaping life around the world for months now, with no end in sight, I’ve been reading up on past epidemics and their impacts on the populations and societies of the Western Hemisphere, and today I’m launching a series of blog posts discussing these issues. I don’t have a clear sense yet of how long this series will go on or what the frequency of posting will be, but it will likely be pretty extensive. The literature on this subject is huge and fascinating, and I’m still working my way through it.

To keep some control over the scope of this series, I’m setting some basic guidelines in advance for what it will include. The main focus will be on the Western Hemisphere and the impacts of diseases introduced by Europeans on Native American societies, although this may branch out a bit into other geographical areas (e.g., Oceania and Africa) that offer interesting parallels and/or counterpoints to the American experience, and I will also look to some extent at the impact of epidemics on European settler societies as well, and in some cases also at possible epidemic diseases that were transmitted in the opposite direction, the most famous example of which is syphilis. The temporal scope will start with 1492, though with some attention to the epidemiological and demographic landscapes before that that shaped the progress of events afterward, and end before the worldwide flu pandemic of 1918. The literature on the 1918 flu is vast and interesting in its own right, but it’s just too much to incorporate into what is already a very ambitious project.

One of the major issues in this field, which has shaped a lot of the scholarly discussion especially over the past 50 or 60 years, is the question of the total Native American population of the Western Hemisphere before European contact in 1492. Estimates of population have varied immensely over time, with enormous implications for how scholars have understood the nature of Native societies, European colonization, and many other important issues. I’ll go into much more detail about the various estimates and the controversy over them in subsequent posts.

Norton Sound Regional Hospital, Nome, Alaska

I’m calling this series “Virgin Soil, Widowed Land.” Both of those phrases have come up in the scholarly debate over epidemics and demography, and I find it interesting that they both use the same (rather distasteful, to be honest) metaphor in very different ways. “Virgin soil” epidemics are those that impact populations with little or no preexisting immunity to the disease in question, so they cause intensive impacts well beyond those on populations with more immunity. The current COVID-19 epidemic is of this type, since the coronavirus in question is new and no one in the world had immunity to it when it emerged. Similarly, New World populations lacked immunity to most Old World diseases, which therefore had catastrophic impacts on them. (Just how catastrophic and what the exact impacts were is very controversial, of course.)

The “virgin soil” concept refers to the populations that an epidemic impacts, but it intersects with a separate use of the virginity metaphor with a longer history in the study of European colonialism: the “virgin land.” In this concept, the Native people of the Americas were few in number and made limited, superficial use of the land, so the land was essentially unused and available for the taking by European colonists. There is a lot of implicit racism and white-supremacist thinking in this concept, but that’s a lot of the historiography of European colonialism for you. Once some scholars started looking more closely at some of the evidence for pre-Columbian population and the impacts of epidemic disease in the wake of initial contact, the virgin land concept came to seem less and less plausible even descriptively, and in some circles it began to be replaced with the idea of a “widowed land,” in which the land may have been largely empty in many places when European colonists arrived, but this was in large part due to the earlier impacts of virgin soil epidemics spurred by initial European contact.

This makes European colonization look a lot worse in some ways, though it arguably still lets the colonists off the hook too much. One objection to the emphasis on epidemic disease as a factor in Native depopulation is that it seems to imply that depopulation was both inevitable after contact and in some sense not really the colonists’ fault since they didn’t know they were carrying deadly disease with them. As I’ll discuss in future posts, there may be something to this but many researchers have pointed to other more direct impacts from deliberate actions of the Europeans, who definitely attacked, enslaved, and violently displaced Native groups from many areas in ways that probably caused substantial mortality on their own in addition to amplifying the effects of disease.

Anyway, there’s much more to say about these issues both in general, big-picture terms and at the level of individual microhistorical case studies. This may seem a little far afield from my focus on Chaco Canyon, which long predates European contact and the impact of these epidemics, but I see it as all part of the same big story, and it certainly is topical and potentially of interest in our current pandemic-dominated world. I can’t necessarily say there are specific lessons we can take for the COVID-19 pandemic from studying previous ones, but I think it’s always better to understand the past better to inform decisionmaking in the present.

San Juan Regional Medical Center, Farmington, New Mexico

Read Full Post »

Foreshadow: Synthesis

Third-Story Walls with Type I Masonry at Una Vida

The final chapter in Crucible of Pueblos offers, in the words of its title, “a synthesis of sorts.” Authored by the noted Chaco specialist John Kantner, it gives a brief chronological overview of the period covered by this book, combining the information from the other chapters to create a picture as complete as possible given current evidence. As Kantner notes several times, current evidence is very sparse for certain regions and periods, and the resulting synthesis is therefore tentative on many issues.

Kantner starts with the period AD 600 to 725, which some but not all of the regional chapters cover. He focuses on the idea that this period was marked by a “Neolithic Demographic Transition” of the sort seen in other parts of the world following the adoption of agriculture. In this case he sees the catalyst for the transition not being the initial introduction of domesticated plants to the northern Southwest, which an increasing body of evidence has shown was actually much earlier, but on the idea that new varieties of maize that were introduced at this time caused a widespread shift to a farming-based lifestyle, whereas earlier cultigens had just been added into a hunting and gathering system as a minor component. This theory has been advanced by several archaeologists in recent years, and it is certainly plausible, but I think the data is still not quite there to establish it firmly. In any case, Kantner sees the immediate result of the shift to intensive agriculture being a sharp increase in population, which led at least in some areas to increases in site size (but only to slightly larger hamlets in most cases), as well as possibly to violence and warfare, as evidenced by an increasing number of stockaded hamlets. Sites were still generally quite small and loosely clustered around a variety of types of public architecture. He claims not to see much evidence of migration between regions during this period, which sounds dubious to me given how much we see later. As he acknowledges, though, the data for this early period is particularly limited, especially for less-researched areas, and it’s hard to draw any firm conclusions.

His next period, AD 725 to 825, definitely does show a lot of migration, and Kantner sees that and increasing settlement aggregation as being the two major processes evidence in the archaeological record. Data gaps are an issue here as well, however, and the details of these processes are much clearer in some regions (especially the Northern San Juan/Mesa Verde area) than others. All this migration and aggregation seems to have led to increasingly ethnically diverse communities, although identifying “ethnicity” in this sort of context is tricky as material culture traits that might be used to identify groups don’t always cluster neatly. Despite this diversity, Kantner sees less evidence in this period for violence than in the previous one, at least until the very end of it when there are some spectacular examples like the apparent massacre at Sacred Ridge, which may have been ethnically motivated. Less spectacularly, the presence of defensive sites in Southeast Utah also seems to increase at the end of this period, again suggesting conflict. Interestingly, though, there seems to be little or no evidence for this sort of conflict further south, although again it’s important to note that southern regions have seen much less research. This period saw possibly the earliest examples of settlements aggregated enough to call “villages,” although Kantner notes that a large portion of the population was still living in dispersed hamlets. The question of why some but not all people chose to begin living in greater proximity is an important one that remains largely unanswered.

The trends of migration and aggregation continue into Kantner’s next period, AD 825 to 880. This is especially apparent in the well-studied Central Mesa Verde region, but it appears to have continued in other areas as well, with a general trend toward settlement in well-watered areas, which may signify another episode of agricultural intensification. This is also suggested by the increased storage capacity of the new villages, some of which might indicate community-level storage of grain. Kantner notes that larger villages might also have been able to mobilize more people for hunting and therefore increased hunting success, a reminder that even a heavily emphasis on agriculture doesn’t necessarily replace all other subsistence pursuits. There also is some evidence for changes in gendered labor at this time, again likely tied to subsistence changes: greater emphasis on stored food, presumably largely in the form of cornmeal/flour, would require more time spent on particular types of processing work. This would potentially include both grinding itself and other tasks required by new ways of preparing food, especially making more pots in which the ground meal would need to be cooked. These are presumed to have been primarily female tasks, so the increased time investment in them may have affected gender roles and relations between the sexes. This is an interesting idea that I think could use more elaboration.

Great Kivas A and Q, Pueblo Bonito

Public architecture continues to be diverse but there are some interesting patterns in what types occur in the new villages, especially between great kivas and what Kantner (among others) calls “oversized pit structures.” He makes some suggestions about correlations between these architectural forms and other factors, such as the idea that great kivas may be associated with periods of social instability and the idea that oversized pit structures were more exclusive than great kivas but the ritual in them may have been more ostentatious, judging by the unusual deposits left behind, such as exotic animal remains and redware pottery. He notes the theory that oversized pit structures and their accompanying U-shaped roomblocks may have been associated with emerging ritual leaders, who may have competed with each other for status and power based on their increased storage capacity, access to game meat, and possibly capacity to control craft activities as well. I think there’s a lot of merit to this idea, although it does still rely quite heavily on data from the well-studied Central Mesa Verde area and new research elsewhere might complicate it.

It’s worth noting again, however, that despite the many very visible and interesting changes resulting from increased aggregation a large portion of the population was still living outside of villages. How these people would have interacted with the villages and how their lives might have differed are under-studied but important questions.

Kantner refers to his next and last period, starting in AD 880, as “the Dawn of Chaco,” which seems reasonable given the emerging picture. The key change at this time is the abandonment of the villages that arose in the previous period and the almost complete abandonment of the Central and Eastern Mesa Verde regions, with their residents apparently moving both west into Utah and south into New Mexico, where some of them very likely contributed to the early development of the regional center at Chaco Canyon. This may have been associated with a period of favorable rainfall in the Chaco area compared to a difficult time in the north, but the climatic details are not yet clear. Kantner notes that recent evidence has suggested that the prior population in the Chaco area was a lot smaller than had once been thought, but he also notes that there definitely was an existing population in and around Chaco, and that some sites like Pueblo Bonito were already established before this migration. This population seems to have had ties to the south and was likely different ethnically from the people moving in from the north. There is some evidence for violence that might have accompanied the initial stages of the migration, but it appears that the groups reached an accommodation of some sort over time that led to the development and florescence of the Chaco Phenomenon over the next three centuries. Kantner suggests that the instability of the early period, and possible inequities between the groups, may have contributed to this process of “social elaboration,” which is another interesting idea meriting further study. There are some clear continuities in architecture between the earlier villages and the communities that developed at Chaco, but the question of what had changed to make Chaco so much more successful and long-lived than the northern villages remains open.

In closing, Kantner reiterates some of the caveats he has mentioned before about interpreting this emerging picture. Why didn’t everyone join villages? This seems like a particularly important question to me, and one that has not received enough attention in the development of aggregation models. It’s a particular problem for models that emphasis “push” factors like the need for defense in an increasingly crowded landscape, though Kantner suggests that this may have been a bigger factor for immigrant groups entering a potentially hostile new area than for the indigenous groups they encountered. He has more discussion of “pull” factors, such as economies of scale for intensified work on activities like farming, hunting, and craft production, but ultimately suggests that a complex combination of pushes and pulls may account for the notable variation in village forms that we see throughout this period. Another important question is why these early villages failed. Kantner suggests changes in the above-mentioned balance of push/pull factors, as well as the possibility that aggregation created its own new problems and stresses on the emerging social systems. Whatever the details, it seems increasingly clear that the lessons from the complicated processes covered by this book formed the basis for the later emergence of Chaco and the immense changes in the Pueblo world that it would entail.

Sign at State of New Mexico Archives Building, Santa Fe, New Mexico

Read Full Post »

Foreshadow: Context

Phoenix, Arizona

The last two chapters in Crucible of Pueblos try to offer assessments of the rest of the book from very different perspectives. The second-to-last, by Steve Lekson, is the shortest chapter in the book, and much of it is drawn from Lekson’s own recent book giving an overview of Southwestern archaeology (which he is very upfront about admitting, to his credit). Nevertheless, it does manage both to situate the present book in the context of other synthetic works covering specific periods of Southwestern prehistory and to situate the areas covered by the book into their own context given what else was happening in other regions at the time.

First, Lekson notes that this volume is the latest in a series of books bringing together experts on different sub-regions of the northern Southwest to compile and analyze data on the most important sites of a given period. This hasn’t been a formal book series, but more of an ad-hoc process of meetings and resulting publications that have ended up being functionally similar. The most obvious examples of other books like this are two published by the University of Arizona Press and deliberately put in similar formats, covering the Pueblo III and Pueblo IV periods. Lekson claims that he played a major role in organizing the meeting that led to the Pueblo III book, the first to be published, which I didn’t know but have no reason to doubt (he was not a contributor to the published book). I discussed that book a few years back in a series of posts very similar to this one. I’ve also read the Pueblo IV volume but haven’t discussed it in any detail here.

What I found particularly interesting was Lekson’s argument that the Pueblo II period, the time of the florescence of Chaco Canyon, is also covered by a volume in this informal series, in this case a special issue of the journal Kiva that was part of the series of meetings and publications that Lekson organized to synthesize the work of the Chaco Project. I have all of the articles in that issue in electronic format, although I haven’t read them all. This chapter makes me think I should.

Finally, Lekson presents Crucible of Pueblos as the equivalent volume for Pueblo I, which makes a lot of sense for me. As he notes, information on major sites is a lot harder for this period than for subsequent ones, as they are much more subtle on the landscape than in later periods. Nevertheless, he recognizes the amount of information brought together for the first time here and its importance, as do I. I’ve tried to convey the importance of this information in this series of posts and I hope I’ve succeeded.

Moving to his discussion of the Pueblo I period itself, Lekson tries to situate the process of village formation in the northern Southwest in the context of developments elsewhere, particularly in the southern Southwest and to a lesser extent Mesoamerica. This is the part of the chapter that’s drawn largely from his book, so it wasn’t new to me, but it is still an interesting and thought-provoking way to conceptualize Pueblo I. The basic argument he makes is that while village formation was (mostly) new in the northern Southwest in Pueblo I and is in some ways its most interesting development, there had been villages in the deserts of southern Arizona and adjacent areas for centuries at that point, and for even longer further to the south in Mexico, the probable origin of the concept for this part of the world. Specifically, he suggests that the formation of villages in the northern Southwest may have been a reaction to the expansion of the Hohokam out of the Phoenix Basin during the so-called “Colonial Period.” It’s not clear exactly what form this expansion took in terms of migration, diffusion, or other processes, but Lekson argues persuasively that people further north and east on the Colorado Plateau must have been aware of it and that they probably factored it into their own decisions about where and how to live.

In the book, Lekson goes into more detail about all of this, but for this chapter he more or less stops there. As I said before, this is a very short contribution, but an interesting one. And, of course, it is written in Lekson’s inimitable and highly accessible style. Not the most weighty or original part of this book, perhaps, but interesting and worthwhile all the same.

Read Full Post »

Foreshadow: Ideology

Chaco Petroglyph Panel Showing Abstract “Blanket” Designs

Chapter 11 of Crucible of Pueblos, by Rich Wilshusen, Scott Ortman, and Ann Phillips, is called “Processions, Leaders, and Gathering Places,” but I think a more concise description of its main concern is ideology. Specifically, this chapter looks at changes in the ideology of leadership, power, and community organization during the Basketmaker III and Pueblo I periods, as seen through the archaeology of public architecture, the portrayal of processions in rock art, and the reconstruction of related vocabulary through comparative linguistics. Due to this innovative interdisciplinary approach, I found this one of the most interesting chapters in the book. Some of the argumentation and conclusions strike me as either weak or overly speculative, but overall this is a fascinating example of how approaches from very different disciplines can be skillfully combined to produce a more complete picture of the past.

The overall argument in the chapter is fairly straightforward. The authors argue that there were a series of shifts in Pueblo society from the Basketmaker III to Pueblo I periods:

• The overall settlement pattern shifted from dispersed hamlets to aggregated villages.
• The locations for occasional ritual gatherings shifted from symbolically important central locations with public architecture to specific locations within villages that in some cases were likely residences of village leaders who exerted control over rituals they hosted.
• The social ties that sustained communities shifted from personal relationships between individuals to symbolic relationships between abstract corporate entities to which individuals belonged.

The authors see all of these shifts as being ultimately driven by the rapid increase in population from an intensification of agriculture (the so-called “Neolithic Demographic Transition”). The actual evidence for this transition, and its relationship to agriculture, seems a bit thin to me, but at least on a theoretical level it makes sense, and there’s certainly no question that populations were increasing rapidly in the Mesa Verde region (to which this chapter, like several others, essentially confines itself due to the scarcity of comparable data for other areas) during the period they discuss.

From archaeology, the most important shift the authors discuss is the well-known change from dispersed settlements during Basketmaker III to aggregated villages in Pueblo I. (Again, we’re essentially just looking at the greater Mesa Verde area here, without any discussion of the possible Basketmaker III villages at Chaco Canyon.) One aspect of the Basketmaker III settlement pattern that is particularly important is the presence of “isolated” public architecture of presumed ritual function, which in some cases took the form of “great kivas” and in other cases took the form of “dance circles,” the main distinction being whether the structure appears to have had a roof. These structures are generally thought to have hosted occasional rituals that brought in people from throughout the surrounding area and helped to integrate them as a social “community.” In addition to the actual rituals performed, about which we know little to nothing, these events would have provided opportunities to trade, share information, and find marriage partners, all important activities to ensuring the success of the community and its members.

As people began to gather in aggregated villages during the Pueblo I period, the nature of public architecture begins to change. Great kivas are still being used in some villages, but less and less over time, and some villages don’t have them at all. Instead, it appears that some of the integrative functions of the great kivas are being taken over by U-shaped roomblocks with associated “oversized pit structures” that have features suggesting ritual use but, importantly, not in the same way as great kivas. The U-shaped roomblocks appear to have been at least partly residential in function, and they may have served as the residences of emerging village leaders who used the plazas they partly enclosed, as well as the oversized pit structures, to host community rituals that served many of the same functions previously served by great kivas. Unlike the great kivas, however, which appear to have been communal sites not associated with any particular members of the community, these structures would have been under the direct control of the families or kin-groups that owned them, who would therefore have the opportunity to amass ever more status, power, and wealth. There have been suggestions, repeated here, that these structures were the forerunners of the later “great houses” at Chaco and its outlier communities, which seems increasingly plausible as more is known about them. (It’s worth noting, however, that great kivas reappear at Chaco as well.)

So far so good, and this is about as far as the archaeology can take us. These ideas are plausible, but they’re not new. Where this chapter goes further than others, however, is in incorporating evidence from rock art as well. The specific focus is on rock art depicting what appear to be ritual processions. The authors analyze two specific panels in detail. One, from Comb Ridge in southern Utah, is thought to date to the Basketmaker III period and to depict the sort of gathering of dispersed communities at a central ritual site that was argued above to have been typical of this period. The other panel is from near Waterflow in northwestern New Mexico, and it is argued to date to later, after the collapse of the Pueblo I villages in the Central Mesa Verde region but before the rise of Chaco to the south. This site is at a key point along what may have been one of the main routes between those two areas, which may be important.

I won’t go into much detail about the analyses of the two panels, interesting though they are. The main points are that the Comb Ridge appears to depict at least two groups approaching a round great kiva or dance circle site from different directions, possibly reflecting the joining of two previously separate communities into one. The focus is on long lines of human figures, some of which have elaborate regalia or carry possible ritual objects, which may indicate that they represent specific individuals. Referring to an earlier study, the authors suggest that the focus on these rituals in Basketmaker III rock art represents a shift in ideology from earlier Basketmaker II art that focused on life-cycle rituals and individualistic shamanism to a more communal type of ritual associated with the central sites.

There is very little rock art associated with the Pueblo I villages, and no known procession scenes at all. The authors don’t discuss this fact in any detail, but it seems significant as evidence for a shift in ideology associated with the new ritual forms they describe as indicated by the architecture. Yet another shift appears to be indicated by the reappearance of procession scenes during the Pueblo I/Pueblo II transition as represented by the Waterflow panel. Here, the procession is primarily of animals rather than people, and they are approaching a square divided into halves and decorated with abstract designs. The whole panel has much more of an abstract feel, and it includes symbols of authority known from later Pueblo religion such as twin mountain lions who appear to be guarding the square. The authors interpret the square as representing the community, with the animals approaching it possibly being symbols of corporate groups like clans that make it up rather than known individuals. Of particular interest, the authors suggest on the basis of other rock art evidence that the symbols on the square actually represent a specific community, as there are apparently other symbols like this with various abstract symbols that may depict community in a sort of “heraldry” comparable to the city glyphs known from Mesoamerica. There are also intriguing petroglyphs of human figures with these squares as heads, possibly indicating village “heads” or chiefs. This system doesn’t appear to continue into later periods, at least in this form, though it may be worth taking another look at distinctive rock art motifs found at later sites to see if there is any continuity in the symbolism. The so-called “blanket” motifs found in rock art at Chaco are similar at least in form.

So the overall picture from the rock art evidence is of a shift from showing communities as consisting of groups of individual people who gather at a central location on certain occasions to more abstract depictions of communities as consisting of social categories, rather than individuals. This may reflect a further step in the development of community ideology after the first, apparently failed, experiments with village living during Pueblo I. The elaborate system that developed subsequently at Chaco may have been yet another step.

Turning to language, this is a particularly interesting part of the chapter for me given my linguistic background. It is based on Ortman’s dissertation, subsequently turned into a book, which considered linguistics along with other lines of evidence to understand the cultural makeup of the Mesa Verde region in the later Pueblo III period. While several languages from different families are spoken by the modern Pueblos, here the discussion is limited to the Kiowa-Tanoan language family, the only family that is both primarily spoken by Puebloan peoples and complex enough in structure to analyze historically in any detail. The analysis is based on what terms for culturally important items and technologies can be reconstructed to different stages of the language, and how the presence or absence of certain terms relates to when they were introduced in the archaeological record. So, for example, the (Puebloan) Tanoan languages share some terms related to agriculture with the (non-Puebloan) Kiowa language, but lack shared terms for such items as pottery, beans, and the bow and arrow. Since these items were introduced to the northern Southwest in the Basketmaker III period, it appears that Kiowa broke off from the other languages no later than Basketmaker II. The subsequent divisions within Tanoan look a lot shakier to me, but if they do hold up they seem to indicate that the Towa language split off during Basketmaker III, which would have left the language ancestral to Tiwa and Tewa as having been spoken during Pueblo I, possibly in some of the early villages of the Mesa Verde region. Tiwa and Tewa are said to have split after Pueblo I, which the authors of this chapter suggest indicates that it was the collapse of those villages that caused the split.

This is an interesting approach to trying to align the linguistic and archaeological records, and I’m glad people are looking at it. It doesn’t seem to add much to the other two lines of evidence in this specific case, however, and there are some potential issues that make it hard to apply in general. For one, it can be hard to tell if the inability to reconstruct a term to a given protolanguage truly indicates that the item it represents was not present during the period when that protolanguage was spoken, especially in a small language like Kiowa-Tanoan. Terms can be lost in daughter languages in many ways, with the ultimate result being the same in the present language as if it had never existed. However, this is a much more productive approach to the problem of correlating linguistics with archaeology than some others that have been tried, like glottochronology, and it’s definitely worth pursuing to see what insights it can provide.

Another problem, however, is that there are several other Pueblo languages not related to Kiowa-Tanoan, and this type of analysis doesn’t, and can’t, say anything about when and where they might have been spoken. A better approach to try to address the diversity of languages among the Pueblos is to look at loanwords, both between different Pueblo languages and between them and non-Pueblo ones, and try to see what can be inferred about when certain items were introduced to speakers of a given language based on that. There have been some studies along these lines that have given some interesting insights and more work would be useful.

Overall, this chapter is a really interesting approach to trying to correlate different types of analyses to complement each other and get a better answer to a specific question about the past than any one type of analysis individually. At the end the authors call for more work like this, and I second that call. The specific conclusions arrived at in this publication may or may not hold up under further study, but the process it demonstrates for getting them will be helpful in moving forward and getting more complete and reliable answers.

Read Full Post »

Foreshadow: Women

“Pithouse Life” Sign at Mesa Verde

Chapter 10 of Crucible of Pueblos, by Richard Wilshusen and Elizabeth Perry, looks at the position and roles of women in early Pueblo society, with a particular focus on how those roles seem to have changed with the economic and demographic changes of the late Basketmaker III and early Pueblo I periods that recent research is bringing into focus. It’s a thought-provoking chapter but in some ways rather odd, with many of its most intriguing proposals resting on what seems like fairly thin evidence.

The chapter looks at three main topics: food production, human reproduction, and gender relations in social power. Its overall thesis is that in the northern Southwest between AD 650 and 850 interrelated changes in food production systems and human reproductive rates led to major changes in gender roles, particularly regarding the division of labor between men and women, that may have led to settlement aggregation into villages and to changes in social power related to trade and ritual. The resulting social structure, in place by the end of the Pueblo I period in at least some areas, was the earliest form of the “Pueblo” society as known from modern ethnography, with its strict division of labor by gender and extension of this gendered ideology to many other domains of life.

Of the three main topics, the authors devote the most attention to the first, and it’s here that their arguments are strongest and most clearly supported by the archaeological evidence. The key change in food production during the period in question is the intensification (not introduction) of maize agriculture as the primary subsistence activity, supplemented by the growing of other crops like beans and squash, by the raising of domesticated turkeys, and by hunting and gathering of wild foods. Recent research has clarified the sequence and extent of this change, although a lot of questions still remain.

As the authors note, one of the most puzzling results of research on early agriculture in the northern Southwest is that maize is now clearly established as having been introduced as early as 2000 BC in several widely spaced parts of the Colorado Plateau (including Chaco Canyon), but for hundreds of years it appears to have remained a minor part of the diet of the groups that used it. It was only in the Basketmaker II period, between 300 BC and AD 300, that maize use became widespread, and even then, according to Wilshusen and Perry, local groups varied widely in the contribution of maize to their diets. There may have been a distinction between immigrant groups from the south that had a heavily agricultural subsistence base and local hunter-gatherers who were gradually incorporating some farming into their lifestyles.

This slow and incomplete adoption of agriculture is in contrast to the situation in other parts of the world where agriculture, once introduced, spread rapidly and quickly replaced hunting and gathering. It’s still not clear why, although Wilshusen and Perry note that as a tropical plant originating in Mexico, maize would have been poorly suited for the harsher climate of more northern latitudes, and that it would have taken some time for people to breed hardier varieties. It is also apparent that the variety of maize initially introduced was small and not as obviously superior to local wild plant foods as later varieties, and that it was initially introduced without an accompanying “package” of other domesticated foods, as was the case with agricultural spreads in other areas. Domesticated squash seems to have been introduced not long after maize, but separately, and domesticated turkeys appear to have been introduced from a different direction altogether, although the timing and details of their domestication remain very murky.

Be that as it may, the main point Wilshusen and Perry make in regard to this slow adoption of maize is that it is likely that women, who based on cross-cultural studies of hunter-gatherers tend to be responsible for gathering of plant foods, were involved in the initial use of maize in the northern Southwest. However, this small-scale introduction of a new food, even one associated with a new type of food production, probably wouldn’t have had a major impact on existing gender roles or division of labor. That would come later.

The full “Neolithic package” appears to have arrived in the northern Southwest between AD 300 and 600, with components including a larger, more productive variety of maize known as Harinosa de Ocho; beans, newly introduced from the south; greater use of turkeys for both meat and feathers; and greater investment in facilities for food storage and processing. The greater productivity enabled by these innovations led to rapid population growth and the spread of agricultural groups over the landscape, in striking contrast to the lack of such growth with the initial introduction of maize much earlier. Wilshusen and Perry associate these developments with the transition from Basketmaker II to Basketmaker III, as well as with the major changes in the roles of women that they document.

Additional support for these changes in food production come from complementary changes in storage facilities and grinding tools. As documented in the well-studied Dolores area, the importance of storage seems to have risen over the course of the late Basketmaker III and Pueblo I periods from AD 650 to 875. Storage facilities changed from small pit rooms isolated from the main dwellings to more secure and more solidly built storage rooms directly attached to living rooms and only accessible from them. These typically consisted of two storage rooms at the back of each living room, the beginning of the “suite” layout that would continue to be a key architectural feature into Chacoan times. It is possible that at least initially these paired rooms were used to store two years’ harvests, one in each room, so as to provide a subsistence buffer against drought and other unexpected problems.

There was also a shift over this period in grinding implements from basin metates with one-handed manos to much more efficient trough metates with two-handed manos. Beyond this shift, a greater variety of grinding tools became common over time. Together with the storage data, this indicates an increased importance of grinding as a component of food preparation. In the modern Pueblos grinding is a female-associated activity, part of an overall suite of food-preparation tasks accomplished by women that also includes shucking, shelling, drying, and storing corn. Of these tasks, however, grinding is considered particularly important to the female role, and it is an important part of female puberty ceremonies (of which the Navajo Kinaaldá, of clear Pueblo origin, is probably the best known). Men, on the other hand, are responsible for planting and harvesting the corn, as well as protecting the fields. This seems to be a change from the presumed hunter-gatherer system in which women were generally responsible for gathering plant foods as well as processing them, and Wilshusen and Perry suggest that it may have arisen in early Pueblo times as fields at greater distances from residence locations in villages became increasingly vulnerable to attack by enemies, prompting men’s role as warriors to encompass guarding fields and, in time, tending them as well.

Another important female task in modern Pueblos is making pottery, and this too seems to have become increasingly important with the expansion of agriculture in early Pueblo times. With more use of crops and additional cultigens such as beans, pots would have become more important for food preparation, and with the number of vessels needed and their short use-lives of 1 to 6 years women would have had to be constantly making new ones. (This is of course assuming that most pots for domestic use were made by the family unit itself, which may well be true for this early period but was not necessarily later on.) Based on detailed study of an isolated Pueblo I hamlet in the Central Mesa Verde area, Wilshusen and Perry estimate the following assortment of vessels for a typical household at any given time:

• 2 to 7 small cooking jars
• 1 to 4 medium cooking jars
• 0 to 2 large cooking jars
• 1 bowl
• 0 to 3 ollas for water
• 2 to 3 other vessels
• 10 to 20 sherds from broken pots used as containers or tools

Keeping a household supplied with all these pots would have been a major part of a woman’s domestic labor, in addition to the food processing tasks mentioned above, along with other major responsibilities such as caring for children.

And speaking of children, Wilshusen and Perry go on to discuss human reproduction and the apparent changes in it associated with the other changes they identify. The two main changes they note are shifts in the use of cradleboards and an apparent increase in the societal fertility rate. This part is somewhat less thorough than the food production part of the paper, but it does still identify some intriguing evidence for change.

First, cradleboards. The authors note that study of these items, in which infants were bundled while they were very young,  has been surprisingly limited, despite their relevance to an important event that has long been recognized: the beginning of evidence for “cranial deformation,” or the reshaping of skulls as a result of prolonged contact with certain kinds of cradleboards in infancy. The shift from “undeformed” to “deformed” (the terminology is very problematic, as there is no evidence of health problems from the practice) crania is traditionally associated with the transition from Basketmaker III to Pueblo I, and early in the history of Southwestern archaeology the change in head shape was even taken as evidence for a population replacement. (That was in the early twentieth century when anthropologists put much more emphasis on skull shapes in defining populations than they do now.) It is now generally thought that the distinction is actually due to the use of soft versus hard cradleboards, but recent research that Wilshusen and Perry discuss suggests that both types of cradleboards were present in both Basketmaker and Pueblo times. Thus, the shift in cranial shape is actual due not to a change in the type of cradleboard but in how it was used. The main changes that actually seem to have occurred in the Pueblo I period are:

• Foot rests on cradleboards disappear.
• Hoods become more common.
• Construction is more expedient.

According to Wilshusen and Perry, these changes together indicate that women had less need to move while carrying children in cradleboards, but that they needed more cradleboards overall, possibly indicating that they had more children. This part of the paper does not go into much detail about where these conclusions come from, but the overall conclusions is that this is further evidence that women were more tied to the domestic sphere in Pueblo I, and possibly that they had more children.

On that note, demographic data appear to indicate that the population increases seen in at least some part of the Pueblo world during Pueblo I were due largely to natural increase after initial immigration into new areas. The best data come from the Central Mesa Verde and Eastern Mesa Verde areas, both of which seem to show this pattern. Prehistoric demographics are notoriously hard to reconstruct, but based on the large recent data sets from major excavation projects Wilshusen and Perry propose that a Neolithic Demographic Transition (a major increase in fertility associated with the beginning of intensive agriculture) began in the northern Southwest around AD 300, with major consequences over time for women in particular, given their gender-defined economic roles. This is comparable to evidence seen in other parts of the world with the beginning of agriculture. The key point here for the role of women is that with increasing rates of both childbirth and survival of children beyond infancy, families would have become larger, increasing the amount of domestic labor required of women to maintain their households given the gendered division of labor presumed to have developed. This would be one explanation for the increased importance of food processing mentioned above.

Finally, Wilshusen and Perry talk about exchange and social power. The discussion here is very abbreviated, and relies heavily on references to the next chapter in the book (which is a little odd), but the basic idea is that rock art evidence shows a shift in social power to male leadership of ritual in late Basketmaker III, continuing into Pueblo I. Female economic roles expressed in matrilocal residence may have driven men to make external trade alliances, which over time developed into new ritual systems focused on important lineages within villages rather than large public rituals at central places not necessarily associated with a specific lineage or community. Matrilineal lineages were still important, and the focus of key rituals, but changing gender roles may have involved an increased role for the men of the lineage in certain types of rituals. Burial evidence from Ridges Basin may support some of these ideas, with striking differences in male and female burials, particularly in the types of exotic goods included. Both women and men were buried with exotic items occasionally, but the specific types of items varied, suggesting gendered access to different trade systems. There are also geographic differences within the basin suggesting different community connections and ideological systems. This section is intriguing but very sketchy, even compared to the rest of the paper. More detailed discussion of some of these ideas will have to wait for the next chapter.

Overall, the conclusion of this chapter is that over the course of the early Pueblo period gender roles shifted in a way that evolved into the system(s) that are well known from the modern Pueblos. This may have been a response, in part to the demographic shift resulting from the development of intensive agriculture, with its resulting higher birthrates and changes to the roles of women. Women’s labor was key to this transition, but it’s not clear that it was actually good for women as a class on net. There has been some discussion of the idea of “parallel status hierarchies,” in which men and women had different tasks but both allowed meaningful status through high achievement. However, later evidence from Pueblo sites shows that women were often excluded from access to high-value resources such as meat, and that their graves were generally less elaborate than men’s (a contrast to the Pueblo I situation in at least some areas). It doesn’t appear that many strictly comparable studies of these issues have been done of the Pueblo I period itself, so it’s hard to say how these changes felt for the women who were living through them. The authors of this paper seem to lean toward thinking the changes were not actually beneficial for those women, but the evidence is thin enough that it’s not clear.

Above I have summarized the arguments of this chapter as best I could, but it’s worth noting that the argumentation of the chapter itself is highly abbreviated, and summarizing it has required a lot of assumptions and interpretive leaps. It kind of reads like this paper is an abbreviated version of a longer argument, with some important parts left out. Nevertheless, it raises a lot of interesting questions that have rarely been addressed in Southwestern archaeology, especially regarding the early Pueblo period, and for that alone it is valuable.

Read Full Post »

Foreshadow: Perishables

Plaited Sandals at Chaco Museum

Chapter nine of Crucible of Pueblos looks at perishable artifacts (i.e., those made of materials that are often not preserved in the archaeological record, such as yucca fiber, animal hair, and cotton) during the Pueblo I period. Written by Laurie Webster, one of the most prominent experts on prehistoric Southwestern perishables, this chapter functions partly as an inventory and description of all known perishables from Pueblo I sites, and as such it is highly technical in nature and not very accessible for a casual reader. For this summary, therefore, I will focus on the high-level conclusions that can be made about Pueblo I cultural dynamics and relationships from the perishable evidence, rather than the evidence itself.

Those conclusions are quite interesting, as it turns out, especially when it comes to the patterning of different types of artifacts. Webster covers several different types of artifact, but I will focus on two with the most interesting cultural implications: sandals and textiles.

First, however, a note about the data. As Webster notes, the Pueblo I period has historically been poorly represented in the perishable data compared to earlier and later period that are known for extraordinary preservation from caves and rock shelters, especially the Basketmaker II and Pueblo III periods. People made much less use of caves and rock shelters during Pueblo I, and as a result many more of their perishable artifacts have, well, perished, and those that do survive are mostly in poor condition. Indeed, most of the best-preserved Pueblo I perishables are from areas like Tsegi Canyon and Canyon del Muerto in northeastern Arizona where caves did continue to be used in Pueblo I, although the Pueblo I occupation in these areas is poorly understood and it is not always clear that artifacts assigned to Pueblo I by early excavators really do date to this period. Luckily, however, the nature of perishable artifacts means that they can be directly radiocarbon-dated, and Webster mentions several examples that have been and many more that could be.

With that caveat out of the way, sandals. These were generally made out of yucca fiber and appear to have been a key way people at the time signaled their cultural identity, based on the geographic patterning of different types, and they likely had symbolic importance as well at least for some groups, based on the elaboration of some examples, implying an immense amount of labor, as well as the depiction of sandals in rock art and the creation of clay effigies (often called “sandal lasts” although that doesn’t appear to have been their actual function). In particular, highly elaborate twined sandals were common in western areas during Pueblo I, a continuation of a tradition from Basketmaker times. Pueblo I examples are known from northeastern Arizona, the eastern slope of the Chuska Mountains in New Mexico, the Dolores area in Colorado, and Chaco Canyon. In contrast, only one example is known from the Animas River Valley, and none from further east, despite the large recent excavations in this area in conjunction with large development projects.

Animas River, Durango, Colorado

A different type of sandal dominates in these eastern areas, a twill-plaited design that appears to date back to the Basketmaker II sites near Durango, Colorado. This type dominates in the Ridges Basin and Blue Mesa area of the Eastern Mesa Verde region and is also found in the Navajo Reservoir area further south, as well as at Grass Mesa Village in the Dolores area. The last is particularly interesting given that there is other evidence that Grass Mesa was settled by people from areas further east. It is also interesting that McPhee Village, also in the Dolores area, shows mainly twined sandals, again supporting other evidence suggesting western connections for this site. Similarly, the one site in the Animas Valley showing evidence for twined sandals also has other evidence of western connections.

A third type of sandal, plain weave with a rounded or pointed toe, appears to also have a western distribution extending from southern Nevada to northeastern Arizona and southeastern Utah but not into Colorado or New Mexico. Less is known about this type than the other two and its cultural significance is not clear.

While in general Webster concludes that Pueblo I perishables mostly continue Basketmaker III patterns without major innovations, she does note one major innovation by late Pueblo I: the increasing use of cotton. While many of the cotton textiles from northeastern Arizona attributed to Pueblo I have questions about their dating and associations, there is one example of a sash from Obelisk Cave in the Prayer Rock District (extreme northeastern Arizona) that has been directly dated to the AD 700s (early Pueblo I). One particularly interesting thing about this sash is that it actually consists of a mixture of cotton and dog hair, clearly showing the transition from animal hair and cotton for textiles. While the form of this item and the use of mixed materials strongly implies that it was made locally, it is not clear if the cotton was in fact grown locally or imported from the Hohokam in southern Arizona, who had a well-established tradition of cotton agriculture by this time.

By late Pueblo I, however, there is strong evidence that at least some Pueblo groups were growing their own cotton. At Antelope House in Pueblo del Muerto, cotton cloth in contexts dating to the AD 900s was found along with cotton seeds and bolls, clearly implying that cotton was being grown in this area by then, as it continued to be throughout the Pueblo period. Interestingly, there is no evidence for Pueblo I use of cotton textiles further east, again implying some sort of major cultural boundary. This is in contrast to later periods, when cotton grown in northeastern Arizona was traded to various other parts of the Pueblo world.

So anyway, those are the major points of interest about Pueblo I perishables. I find the most interesting point from the perspective of Chaco to be the fact that it patterns with the western rather than the eastern style of sandal, which reinforces other evidence for western connections for at least some of the people who came to Chaco in late Pueblo I and contributed to its rise into a dominant regional center in the northern Southwest.

Chuska Mountains from Peñasco Blanco

Read Full Post »