Monks Mound Sign at Cahokia Mounds

Regardless of exactly how many people lived at Cahokia, it’s clear from recent research that the population of the site and its immediately surrounding area grew immensely in a short period of time in the eleventh century AD. As Timothy Pauketat points out in the 2003 article that I was discussing earlier, the scale of this growth is much too great to be due to natural increase of a local population. Thus, even though there was a local population of some size in the American Bottom in the immediately preceding period that presumably contributed to the growth of the Cahokia site, it’s pretty clear that there must have been quite a bit of immigration in addition to account for the huge influx of people.

Who were these people, and where did they come from? It can be fiendishly difficult to untangle migrations in the archaeological record, even in conditions like this where it seems clear that there must have been some migrating going on. Archaeology depends mainly on the remains of material culture, which can change quite rapidly under conditions of change and stress such as those that often obtain during and immediately after migrations. Nevertheless, there do seem to be some clues to where at least some of the immigrants into the American Bottom may have originated. These clues mostly revolve around the type of material culture on which archaeologists most often rely to understand prehistoric societies: pottery.

Pots can’t automatically be equated with people, of course, but certain types of pottery are sufficiently distinctive and limited in distribution that they may serve as markers for population movements. In the case of Cahokia, Pauketat and others have pointed out that one of the many pottery types found in deposits dating to the period of population growth bears a striking resemblance to a type known as “Varney Red Filmed” found in northeast Arkansas and southeast Missouri around the same time. Importantly, these pots are often tempered with ground shell. Shell-tempering would later become one of the defining traits of Mississippian societies throughout the South and Midwest, but it appears particularly early in the area where Varney Red Filmed was made, starting around AD 800. According to another study reporting on evidence of Varney-like ceramics found at sites in Wisconsin, examples of Varney wares have been found in the American Bottom starting shortly before AD 1000. While these early finds are considered trade wares, by the time Cahokia emerges about 50 years later similar ceramics are apparently being made there, as shown by analysis of the clays in them, and some shell-tempered sherds that seem to be made from American Bottom clays were identified at the aforementioned Wisconsin sites, attributed to a short-lived influx of immigrants from the American Bottom, perhaps those who lost out in the struggles that led to the rise of Cahokia.

Cahokia Interpretive Center from Monks Mound

Returning to Pauketat’s 2003 article, he finds that not only are locally produced but clearly Varney-influenced pots being made at Cahokia itself in the mid-eleventh century, some of the sites in the rural Richland Complex to the east, which was settled rapidly around the same time, show the same high proportions of Varney sherds as some areas of Cahokia itself, implying that at least some of the people who established the Richland Complex came from the American Bottom. The relationship between these people and those at Cahokia remains difficult to entangle.

One of the hallmarks of the American Bottom in the period immediately preceding the rise of Cahokia is remarkable diversity in ceramic style. Varney is only one of the many local and “foreign” ceramic styles present in sites during this period, and it seems clear that people were coming to the area from all over. Nevertheless, the important similarities between Varney ceramics and those that soon begin to spread along with other attributes of Mississippian culture implies that northeast Arkansas and southeast Missouri may have played an important role in developing the complex of ideas that we today call “Mississippian” and that an increasing body of research suggests initially spread through the influence of Cahokia. Studying the archaeology of these areas during the period before and during the rise of Cahokia is therefore potentially important.

Ceramic sourcing using Instrumental Neutron Activation Analysis has been one line of evidence applied to sites in southeast Missouri during this period, not least because the University of Missouri happens to have a research reactor that has been widely used for sourcing of ceramics and other archaeological artifacts. One paper from this effort, published in 2000, showed that this technique can indeed distinguish among sherds made from clays from three major physiographic regions of southeastern Missouri: the Ozark Highlands and the Western and Eastern Lowlands in the Mississippi Valley. The results showed in addition that there was extensive movement of vessels, including shell-tempered Varney wares, between the Western Lowlands and the Ozark Highlands. It can’t be determined from this data if this reflects trade or migration, but it certainly shows interaction between the two. There is also some limited evidence, in the form of possibly transitional sherds tempered with a mix of shell and sand or limestone, that the shell-tempering technique may have developed here rather than being imported from elsewhere. Given the importance of shell tempering later on, the idea that it may have been an indigenous development among Varney potters may have important implications for lines of cultural influence.

This is all very suggestive of the idea that some of the people who immigrated to the American Bottom around AD 1050 came from the Varney area and brought with them some ideas that they had developed there or perhaps borrowed from other groups. People were certainly coming from other directions too, and their ideas may have been just as influential even if they were less visible archaeologically. In addition, there was a longstanding local population in the American Bottom that surely played a role as well. All of these people likely contributed to the unique flowering of the greatest cultural and political center north of Mexico that began soon afterward. Untangling exactly what contributions each group made, and figuring out in more detail who they were, are bigger challenges than simply tracing where they may have come from and what their pots looked like, of course.

Lynott, M., Neff, H., Price, J., Cogswell, J., & Glascock, M. (2000). Inferences about Prehistoric Ceramics and People in Southeast Missouri: Results of Ceramic Compositional Analysis American Antiquity, 65 (1) DOI: 10.2307/2694810

Pauketat, T. (2003). Resettled Farmers and the Making of a Mississippian Polity American Antiquity, 68 (1) DOI: 10.2307/3557032

The Great Plaza and Monks Mound, Cahokia

The greatest of the Mississippian mound centers, by far, is Cahokia. This vast site contains numerous mounds and is located in the American Bottom area of southwestern Illinois, across the Mississippi River from the modern city of St. Louis, Missouri. This is a highly strategic location, very close to the confluence of the two largest rivers on the North American continent (the Mississippi and the Missouri), and about twenty years ago Peter Peregrine published a short paper using graph theory to demonstrate that Cahokia was located at the point of highest centrality in the entire Mississippi drainage, potentially giving Cahokia’s inhabitants the ability to control riverine trade across a large portion of the continent. In addition, the American Bottom is an area of very fertile bottomlands along the Mississippi potentially allowing for substantial agricultural surpluses. As early as 1964 Charles Bareis noted that the configuration of the mounds at Cahokia relative to the meander loops where the Mississippi had changed its course over time indicated that the river had been relatively stable over time in this region, changing its channel much less frequently than in areas further downstream. The combination of fertile soil and a relatively stable, predictable river could have combined with the strategic location within the river system to provide the conditions under which Cahokia rose to prominence.

Ecological/geographical explanations like this for the rise of “complex” societies like Cahokia were quite popular among processual archaeologists from the 1960s to the 1980s, but they have since been challenged by a newer generation of archaeologists influenced by the post-processual movement that began in Europe. As Michelle Hegmon noted in an important summary of the theoretical status of North American archaeology a few years ago, Americanists have generally not been inclined to go all the way over to European-style post-processualism, instead adopting elements of both processual and post-processual approaches in varying ways. Hegmon labels the resulting theoretical orientation “processual-plus,” and that seems like an apt description to me.

With regard to Cahokia specifically, the main voices for the “processual-plus” perspective have been Thomas Emerson and Timothy Pauketat of the University of Illinois at Urbana-Champaign. Pauketat, in particular, has argued vigorously for what he calls a “historical-processual” approach to archaeology, incorporating the insights of processual culture ecology as well as a more historical-particularist approach more associated with post-processualism. One of his more important contributions along these lines has been a 2003 paper discussing both Cahokia and an area to the east, in the uplands surrounding the American Bottom, that contains what he calls the “Richland Complex”: a group of rural agricultural sites that appear to date to right around the time of the rise of Cahokia, circa 1050 AD. Cahokia rose very rapidly, and it’s apparent that the population increase in the mid-eleventh century couldn’t possibly have been due solely to natural increase. Migration from surrounding areas seems to have been a major part of the development of the site, and Pauketat argues that the Richland Complex also shows evidence of having been rapidly settled around the same time. This area was apparently very sparsely populated before this, and the settlements established in the eleventh century show sufficient diversity in material culture to suggest that their resident immigrated from different areas. Some show a particularly distinctive pottery style typical of southeastern Missouri and northeastern Arkansas, also found at Cahokia itself, which suggests migration from that region into the American Bottom at the time of the rise of Cahokia. (This may be particularly important in explaining the Mississippian phenomenon in general, for reasons I’ll discuss in a future post.)

Interpretive Sign at Cahokia

A recent article in Science discussing recent research in the Cahokia area has drawn the notice of Mike Smith at Arizona State, one of the foremost authorities on ancient urbanism. He seems pretty comfortable with the idea that Cahokia was a city (an idea strongly pushed by the interpretive material at the site itself), but he does question the assertion that it had a lower population density than the Maya cities, which apparently had notably low population densities within the urban centers themselves but notably high densities in the rural areas outside the cities:

Now I don’t know the demographic data for Cahokia and its hinterland, but I think the population density within the Cahokia urban center was probably HIGHER than within Maya cities, but the population density of the “Greater Cahokia” region was most likely lower than that of the Maya lowlands.

Maya cities had very low URBAN population densities (even compared to a sprawling modern city like Phoenix):

• Tikal (Maya):    600 persons per square kilometer
• New York City:  9,400
• Phoenix:    1,900

But the Maya lowlands had a very high  REGIONAL population densities:

• Maya lowlands:  180 persons per sq. km
• New York State:   150
• Illinois:   80
• Arizona:  17

The high density of Maya regional populations (how many people lived on the landscape, whether in large or small settlements) is one of the remarkable features of ancient Maya society. I’d be interested to see how Cahokia fits in comparison with these figures.

Population estimates for Cahokia vary, as you might expect, but I figured I could chime in with Pauketat’s estimates from his 2003 paper. It’s important to note here that Pauketat is an advocate for a larger, more urban Cahokia than that seen by some other archaeologists, so his figures are relatively high compared to some other estimates, but he goes into some detail in this paper about where he gets his figures and his procedure seems reasonable to me.

St. Louis from Monks Mound, Cahokia

For the core area of Cahokia, which is about 1.8 square kilometers in area, Pauketat calculates a population range of 10,000 to 16,000 people at the site’s peak during the Lohmann Phase (ca. AD 1050 to 1100), which equates to a population density of about 5500 to 8900 people per square km. As Smith suspected, this is much higher than Classic Maya cities like Tikal. Indeed, it’s much higher than contemporary St. Louis, which has about 2000 people per square km, and at the high end it approaches the density of New York! Calculating the regional density is trickier because of limited survey coverage in some areas, but for the 300-square-km Richland Complex specifically (a relatively dense rural area), Pauketat calculates a population range of 3000 to 7400 people for roughly the same period, which equals a density of 10 to 25 people per square km. Again, as Smith predicts, this is much lower than the Maya regional density and down with the more sparsely populated US states like Arizona. It’s lower than both Illinois (89 people per square km) and Missouri (34 people per square km) today.

There’s way more to say about Cahokia. The literature on this site is vast, not unlike the literature on Chaco. (Observant readers may have noted some striking similarities between the two sites, which I’ll discuss later.) Given that Cahokia is widely acknowledged these days to have been a city, population density is a useful way to look at it and to compare it to other cities. There are other aspects of this fascinating site worthy of discussion, however, and I’ll have much more on it soon.

Bareis, C. (1964). Meander Loops and the Cahokia Site American Antiquity, 30 (1) DOI: 10.2307/277637

Hegmon, M. (2003). Setting Theoretical Egos Aside: Issues and Theory in North American Archaeology American Antiquity, 68 (2) DOI: 10.2307/3557078

Lawler, A. (2011). America’s Lost City Science, 334 (6063), 1618-1623 DOI: 10.1126/science.334.6063.1618

Pauketat, T. (2003). Resettled Farmers and the Making of a Mississippian Polity American Antiquity, 68 (1) DOI: 10.2307/3557032

Peregrine, P. (1991). A Graph-Theoretic Approach to the Evolution of Cahokia American Antiquity, 56 (1) DOI: 10.2307/280973

Monks Mound, Cahokia Mounds, Collinsville, Illinois

One of the major advantages Southwestern archaeologists have over those studying other areas of prehistoric North America is a very solid chronology, based primarily on tree-rings and extended by diagnostic pottery types that in many cases changed rapidly. As a result of this chronology, in many parts of the Southwest unexcavated sites can be dated to within a century or less based purely on the potsherds found on the surface. This is not the case in most other parts of North America, where dating is based primarily on radiocarbon dates, which are inherently less precise than tree-ring dates, and where the pottery types are often less solidly defined and/or less securely tied to the sequence of absolute dates. In recent decades, however, the development of more precise radiocarbon dates through the accelerator mass spectrometry (AMS) technique and the further refinement of artifact typologies have allowed archaeologists in many areas to get a better sense of cultural dynamics over time.

One of the areas where this has been particularly the case is the southeastern US, where the late prehistoric period is marked by the rise and fall of a wide variety of societies noted for their large platform mounds, extensive use of maize agriculture, and distinctive iconography indicative of a widespread set of ideological beliefs and social structures. These societies are known as “Mississippian,” and while much remains to be known about them, we know much more about them now than we did fifty years ago. They are generally considered to have had a “chiefdom” form of society, less complex than a state but more so than a band or tribe, although an increasing number of Southeastern archaeologists have been questioning the whole concept of chiefdom and of this typological approach to societal complexity in general. Be that as it may, it’s quite clear that the leaders of most Mississippian communities had quite a lot of control over their fellows, as shown by the amount of labor they could control to build the platform mounds (which in Mississippian communities generally served as bases for elite residences and temples, a striking difference from the more communal ritual purposes that mounds had served in many earlier societies) and their elaborate burial ceremonies and grave goods.

Closeup of Monks Mound, Cahokia Mounds, Collinsville, Illinois

The greatest Mississippian center, by far, was Cahokia, in the American Bottom region of west-central Illinois across the Mississippi River from what is now St. Louis. I’ve written a bit about some recent work on the origins of Cahokia, but while it’s a really spectacular site (well worth a visit) it’s important to note that there were many other Mississippian centers, none of them quite on the same scale as Cahokia but many of them pretty important and impressive themselves. Among the most important of these in the area surrounding the confluence of the Ohio and Mississippi Rivers, south of Cahokia, were:

• Kincaid, in far southern Illinois, across the Ohio River from the modern city of Paducah, Kentucky
• Angel, further up the Ohio in what is now Evansville, Indiana
• Wickliffe, a smaller center in what is now Wickliffe, Kentucky, right at the confluence of the Mississippi and Ohio
• Shiloh, further south, on the Tennessee River at the site where the Civil War battle of Shiloh would later be fought

After the introduction of radiocarbon dating in the late 1940s, the chronology of occupation at these sites gradually became clearer. It now appears that Cahokia, despite being the biggest and most impressive Mississippian center, was relatively early in the sequence, with its major building period beginning in the eleventh century AD and continuing into the twelfth, but declining after then and becomingly largely empty by about 1350. Kincaid and Angel, which emerged later than Cahokia, lasted a bit longer, but both seem to have been abandoned by about 1450. The other, smaller centers in the region seem to have similar trajectories, with occupation appearing to end at various points but never later than 1450 or so.

Sign at Kincaid Mounds, Massac County, Illinois

It’s notoriously difficult to date the end of anything by archaeological means, but the widespread lack of post-1450 dates throughout the confluence area is certainly suggestive, especially since it is not universal among Mississippian societies. Indeed, many Mississippian centers further south and east continued to be occupied after 1450, and some were still there when the De Soto Expedition showed up in the sixteenth century (and wreaked sufficient havoc to destroy most of them, but that’s another story).

This apparently total but geographically limited abandonment of some of the largest river valleys on the continent led one archaeologist, Stephen Williams of Harvard, to propose in the early 1980s that this region constituted a “Vacant Quarter” starting sometime between 1450 and 1550 and continuing into the historic period. Williams defined the Vacant Quarter as starting with the American Bottom, extending south along the Mississippi to its confluence with the Ohio (but, importantly, no further south), then up the Ohio as far as Angel and up the Tennessee as far as the Mississippi and Alabama borders. Part of his argument for the vacancy of this area was the total lack at major sites within it of certain artifacts he had found at sites in northeastern Arkansas that he considered temporal markers for the very late prehistoric period. While these or similar objects were also found in other areas further away that are known to have been occupied at this time, including Alabama and Wisconsin, they are conspicuously absent from Kincaid, Angel, Cahokia, and other major sites along the Mississippi, Ohio, and Tennessee. In his early formulations of the concept, however, Williams did acknowledge that his information came entirely from large sites in the major river valleys, and that more study might show that smaller sites in the surrounding uplands may have been occupied later.

Charles Cobb and Brian Butler went on to test this idea in their fieldwork in the uplands of southern Illinois, and they published their results in 2002. What they found was that Williams’s Vacant Quarter hypothesis held for these upland areas as well. The sites they excavated were smaller than the large riverine centers, but the many radiocarbon dates they had acquired showed the same pattern: little to no occupation later than 1450, and abandonment possibly even earlier.

Mural of Kincaid Mounds, Paducah, Kentucky

The fact that abandonment appears from this data to have been essentially complete within the Vacant Quarter, regardless of environmental setting, suggests that while climatic changes associated with the Little Ice Age (one explanation that has been offered for the abandonment of the area) may have had something to do with the abandonment, whatever changes occurred must have been sufficiently severe to affect basically everyone in the area, but not people in surrounding areas with generally similar conditions. To me this suggests that environmental changes were not the main cause of the abandonment, but it is quite possible that they played a role in catalyzing changes that also had to do with social and political dynamics. Persistent drought, which can be quite localized, is a possibility, although it would have to take one hell of a drought to force farmers to abandon what was (and is) one of the most productive agricultural areas in the world. More likely, perhaps, as an environmental cause is severe flooding, which as we’ve seen in recent years can be a major problem in this region. When I was traveling through it this summer the effects of recent flooding were very apparent. Still, it’s not clear why this would have affected the Vacant Quarter but not the surrounding areas.

Finally, as ever with abandonment scenarios, there is the question of where the people went. In the case of Angel, this seems to be apparent. Around the same time the Angel center was abandoned, a new occupation seems to arise a short distance downstream, around the confluence of the Ohio and the Wabash near what is now the Illinois-Indiana border. This is known as the Cabon-Wellborn phase, which is generally Mississippian in culture but lacks platforms mounds and extends well into the protohistoric period, as late as 1650 or even later. Cobb and Butler seem to think this represents the descendants of the people who lived at Angel, and I see no reason to doubt this. The other parts of the Vacant Quarter seem to lack comparable descendant communities, which suggests that they went somewhere else, perhaps joining the surviving Mississippian centers to the south or the quasi-Mississippian Oneota groups to the north.

Mound A, Angel Mounds, Evansville, Indiana

Cobb and Butler note that while the Vacant Quarter is a particularly dramatic example of regional abandonment, such events are not unusual in the archaeological record of the Eastern Woodlands. They suggest that this was a fairly common occurrence, and that it should receive more study than it has. They specifically refer to the Southwest as a region where similar abandonments are widely accepted and intensively studied, and imply that Southeasternists could follow this lead in trying to explain these events. Their own guess about the Vacant Quarter specifically is that “climatic deterioration or other causes may have exacerbated a competitive and often hostile political environment that rendered the region less hospitable,” which sounds good to me.

The major mound sites of the US Midwest and South have been objects of considerable interest for well over a century, but it is only in recent decades that modern archaeological techniques have begun to unravel the undoubtedly complicated stories behind them. There’s still a very long way to go before we have a clear picture of what these societies were like and what happened to them, but we’re getting closer all the time.

Cobb, C., & Butler, B. (2002). The Vacant Quarter Revisited: Late Mississippian Abandonment of the Lower Ohio Valley American Antiquity, 67 (4) DOI: 10.2307/1593795

Ortega Hall, University of New Mexico, Albuquerque, New Mexico

Tim De Chant at Per Square Mile has an interesting post discussing an article by Ruth Mace and Mark Pagel in which they did a statistical analysis of the distribution of Native languages at European contact in North America and found that the density of languages correlates inversely with latitude (when controlling for land area) and directly with habitat diversity (even when controlling for latitude). You can see Tim’s post for more details on how they actually did the analysis. Here I just want to point out a few considerations that aren’t really addressed in the article, interesting though the result is.

First, I should note that the article itself seems fine for what it is. I don’t see any glaring problems with the statistical analyses, and the authors are clearly aware of the potential issues with their data. They don’t try to push the statistics too far, which is a welcome change form how studies like this often go. Furthermore, I suspect the conclusions they reach do in fact reflect a real phenomenon despite the issues I raise below. As they point out, species distributions are generally known to follow a similar pattern, with more species per land area closer to the equator and in areas with more diverse habitats, all else being equal. It makes intuitive sense that this would be true for human cultural groups as well; more ecological niches to exploit should tend to result in more specialization and therefore more groups, generally at higher population densities, within a given area when these conditions obtain than when they don’t. Since there is also a general tendency for different cultural groups to have different languages (for a variety of reasons), in the aggregate language density should also show these patterns.

That said, I have some concerns about the data underlying this study. The language distributions they discuss appear to come from an atlas of the world’s languages published by a trade publisher and presumably intended for a popular rather than a scholarly audience. There’s nothing wrong with that, of course, but it makes me really want to know how the authors of that atlas got their data. Mace and Pagel don’t discuss this issue at all, merely taking the data from the atlas as given, but it’s important to note that determining these distributions is a much more difficult problem than it seems at first sight.

Sign for Ortega Hall, University of New Mexico, Albuquerque, New Mexico

First of all, defining where one language begins and another begins can be tricky. In cases where totally unrelated languages border each other, it’s pretty easy, but in cases where large areas are covered by closely related, contiguous languages, as is the case for many parts of North America, the difference between a series of related languages and a chain of dialects within a single language becomes very important, and this can be very difficult to determine, especially for poorly documented languages such as those spoken in many parts of North America. There are reasonably consistent ways to do this, but for a study like this I’d really like to see them spelled out to know which decisions are behind the data being analyzed.

Furthermore, this distribution of languages is apparently intended to represent the situation at “European Contact,” but Mace and Pagel don’t specify what they mean by that. Contact came at different times in different parts of North America, which is a rather large area. The linguistic situation in 1500 was really quite different from that at 1800, but both of these are reasonable dates for initial contact in different parts of the continent. Since contact came at different times in different places (i.e., it was “time-transgressive”), just mapping the situation in each subregion whenever contact occurred there, which is what I suspect the authors of this atlas did, creates a highly artificial construct when viewed at a continental scale. Trying to control for this and fix language boundaries at a single point continent-wide would be really quite difficult and require either a lot of guesswork or some sort of way to account for the effects of European colonization, depending on which time point you chose. Nevertheless, the kind of analysis Mace and Pagel are doing here really requires a single time point to make sense.

Like I said above, I don’t think this is a terrible paper given what it tries to do, and I suspect that its conclusions do actually point to a real pattern despite the important methodological shortcomings mentioned above. Without more detail on the underlying classification that was the basis for the statistical analysis, though, I don’t see it as being all that useful in actually establishing the reality of that pattern.

Mace, R., & Pagel, M. (1995). A Latitudinal Gradient in the Density of Human Languages in North America Proceedings of the Royal Society B: Biological Sciences, 261 (1360), 117-121 DOI: 10.1098/rspb.1995.0125

Welcome Sign Facing the Ohio River, Paducah, Kentucky

In 1827 William Clark, who had attained national fame as co-leader of the Lewis and Clark Expedition more than 20 years earlier and had gone on to a successful career as an Indian Agent and governor of the Missouri Territory, obtained title to 37,000 acres in western Kentucky along the Ohio River that had been part of a grant to his late brother, George Rogers Clark, for service in the Revolutionary War. William Clark immediately went about evicting the handful of settlers on the site, which was then called Pekin, and surveying it as a new town which he named Paducah.

Paducah would go on to become one of the most important towns in the region, and over the years a local tradition grew up attributing the town’s name to a “Chief Paduke” of a local tribe. A letter Clark wrote to his son, however, appears to clearly show that this is not correct, and that Clark instead chose the name to honor a tribe he knew as the “Padoucas” (consistent spelling was not one of Clark’s strong points) that had once been large but had been much reduced since European contact. This condition applies to virtually all tribes in the Americas, of course, but the name is definitely not of any tribe in western Kentucky, which was occupied by the Chickasaw until 1819, when Andrew Jackson negotiated the cession of this area, known as the “Jackson Purchase,” opening it up for settlement and speculation by people like Clark.

Plaque Describing the Naming of Paducah, Kentucky

Who then were these Padoucas, and how did Clark know of them? Presumably he learned of them either during the Lewis and Clark expedition or later, when he was serving as a frontier official in Missouri, so the place to look is well to the west of Kentucky. And, indeed, a tribe known as the Padoucas (under various spellings) appears in records from French Louisiana starting with the early explorers of the late seventeenth century and continuing down to Clark’s day. These reports refer to the Padouca as living somewhere to the west of the Mississippi and Missouri valleys, which is to say, on the Great Plains.

The French were not very familiar with the Plains, and most of their information on them came from tribes in the Mississippi and Missouri valleys with whom they were in close contact as trading partners and military allies. They did mount a few expeditions further west in the seventeenth and eighteenth centuries that provided a bit more information, but these were too few and far between, and their permanent outposts too far away, for them to keep close tabs on developments on the Southern Plains during this period. This fact turns out to be quite important for understanding the identity of the people the French called “Padouca.”

Plaque Describing William Clark's Foundation of Paducah, Kentucky

The Spanish in New Mexico and Texas, on the other hand, were much closer to the Plains, and between their occasional forays onto them and their close diplomatic contacts with various Plains tribes they had much better information than the French about who was out there and what they were doing. As a result, we can track developments on the Southern Plains in the eighteenth century quite clearly through Spanish documents, and as it turns out there was a lot going on out there.

At the time the Spanish entered the Southwest in the sixteenth century, the Plains were occupied by various groups of hunter-gatherers whom the early Spanish explorers called “Querechos.” After the establishment of a permanent colony in New Mexico in 1598, the Spanish began consistently referring to these groups as “Apaches” (the term itself is of obscure origin, perhaps coming from one of the Pueblo languages), with specific politically autonomous subgroups indicated by modifiers such as “Lipan,” “Jicarilla,” and “Faraón.” There were Apaches to the west of New Mexico as well, including the Navajos. Relations among the Spanish, the Pueblos, and these Apaches varied, with the Apaches at times raiding both the Pueblos and the Spanish villages, and at other times allying with one or both of the more settled groups for various purposes. During the Pueblo Revolt of 1680, some Apache groups joined with the Pueblos to kick out the Spanish, but after the Spanish Reconquest in 1692 Apache raids on both the Pueblos and the Spanish continued intermittently. By the early eighteenth century, however, the Spanish had allied with some of the more distant Apaches on the Plains for mutual protection against the French and their allied Indians in the Missouri Valley.

By the middle of the eighteenth century, however, this picture had been totally changed by the movement of a new people onto the Southern Plains. These were the Comanches, who had broken off from the closely related Shoshones in the Rocky Mountains and headed out onto the Plains, where they acquired horses and became the most aggressive and successful military power that region had yet seen. They swiftly pushed the Apaches out of the way and became the main power on the Southern Plains, raiding New Mexico with a ferocity unprecedented in earlier times. The threat of Comanche raids kept New Mexico Hispanics from expanding significantly east of the Rio Grande Valley for over a century, and it likely also led the Pueblos to more completely accept Spanish rule in exchange for the level of protection provided by Spanish arms. At first the Spanish attempted to defend their Apache allies on the Plains by fighting the Comanches, but this approach met with limited success and only encouraged increased Comanche attacks on Hispanic and Pueblo settlements, so in the late eighteenth century Governor Juan Bautista de Anza decided to switch sides, sign a peace treaty with the Comanches, and agree to assist them in their wars on the Apaches. This general pattern, of the Spanish allied with the Comanches against the Apaches, endured until the Americans conquered New Mexico in the 1840s.

Sign Describing William Clark's Purchase of the Future Site of Paducah, Kentucky

So where do the Padouca enter into this picture? In the Spanish documents, not at all. The term “Padouca” was not used by the Spanish for any of the Plains tribes they encountered. Instead, they spoke exclusively of Apaches and Comanches. Given that the French documents referring to the Padouca place them in the same places as these tribes at the same time, that leaves three possibilities for who the French could have meant by the Padouca:

1. The Apaches
2. The Comanches
3. Both, either by not distinguishing them at all or by referring to one group at some times and the other at other times

With the rise of American ethnography in the late nineteenth century, American anthropologists began to look back at these documents and try to connect them to the ethnographic data on contemporary tribes. They found that many of the tribes from the Plains and surrounding areas used terms similar to “Padouca” to refer to the Comanches, and many simply assumed from this that the French had always used the term to refer to the Comanches as well. This became something of the received wisdom by the early twentieth century, but in 1920 the anthropologist George Bird Grinnell published an article taking another look at the issue taking into account the Spanish and early French sources. From these Grinnell concluded that the term had instead referred to the Apaches, whom the Spanish sources clearly showed were the only tribe that the earliest French sources from the seventeenth century could plausibly have been referring to, since the Comanches didn’t enter the area in strength until well into the eighteenth century. Furthermore, there is one incident narrated by both Spanish and French sources that seems to clinch the case.

In 1720 a Spanish expedition led by Pedro de Villasur was sent out onto the Plains to combat French influence. The expedition included some Jicarilla Apaches as guides. Upon reaching what is now Nebraska, the expedition was attacked and soundly defeated by a group of Pawnees, possibly accompanied by French traders. The important thing about this incident for the Padouca question is that Villasur’s guides are consistently referred to as Apaches in Spanish accounts but as Padoucas in French accounts. Grinnell concludes, quite reasonably, that this shows that at least in the 1720s the “Padoucas” in French sources were Apache.

Mural Showing William Clark Surveying the New Town of Paducah, Kentucky

However, Grinnell’s article doesn’t address the evidence from more recent ethnography, and in a subsequent comment Truman Michelson points out that among the Fox (Mesquakie) tribe the term for “Comanche” is something closely approximating “Padouca.”  He declines to discuss Grinnell’s article further, but his tone clearly shows that he considers this information more dispositive than Grinnell’s historical sources. And, indeed, any full solution to the problem must account for the fact that the recent ethnographic data clearly shows that the term was used in the nineteenth century by many tribes for the Comanches.

A solution tying all this evidence together didn’t appear until thirty years later, when Frank Secoy published an article bringing in more evidence, especially from French maps, and considering both the documentary and ethnographic evidence. He concludes from this that both sides were both right and wrong in part. Specifically, Grinnell was right that in the earliest French sources “Padouca” must mean Apache, but Michelson and others were right that later on it meant Comanche for both the French and the tribes in the Mississippi and Missouri Valleys. The switch came in the middle of the eighteenth century when the Comanches took over those parts of the Plains that had formerly been Apache. Since the French only ventured out into that area occasionally, they were not very aware of this process as it was occurring, and later expeditions encountering the Comanches where their maps (put together by cartographers in France grappling with disparate information from French and Spanish sources) showed the “Padoucas” to be naturally assumed that the two groups were the same, not realizing that the expeditions from decades earlier had actually used the term to refer to a different people who had been there at the time but were now elsewhere.

Furthermore, Secoy notes that the etymology of the term itself gives a clue to how its meaning might have changed for the Indians who used it. It seems to come from a verb root common throughout the Siouan language family with the meaning “to pierce.” The rough meaning of “Padouca” in whichever Siouan language the French got it from would then be “piercer,” which Secoy, drawing on his own previous research on warfare on the Plains, links to the use by Plains groups in the early historic period of large piercing lances, which they had adopted during what he calls the “post-horse–pre-gun” period of Plains warfare. That is, since the Plains tribes acquired horses before they acquired guns, there was a period when the key advantage of having a horse was in allowing a warrior to get very close to his enemy very quickly, and a thrusting lance was a very useful weapon in this context that would have made a major impression (so to speak) on the horse-less groups that encountered these groups. Since the first groups like this that the Siouan tribes of the Missouri Valley encountered would have been Apaches, the name “Padouca” or “Piercers” would make sense for them to use and to pass on to the French. Since the Comanches entered the Plains during this same technological period and used the same sorts of lances, however, the name would have made equal sense for them and could have easily been transferred once they showed up in the eighteenth century. Indeed, the new horse-based Plains culture would have made both groups seem pretty alien to the Siouans, and they may even have used “Padouca” to refer to Plains groups in general rather than to specific cultural or political entities. Apaches and Comanches speak unrelated languages, but both are also unrelated to Siouan and represented the furthest east representatives of both their languages families as well, so it’s possible that the Siouans couldn’t easily tell them apart. Secoy notes that the French use of “Padouca,” based on outdated information, for the Comanches could also have influenced Siouan usage.

Downtown Paducah, Kentucky

But what about Clark? Where did he get the term, and what did he mean by it? Secoy discusses the appearances of the term “Padouca” in the records of the various American explorers, including Lewis and Clark, and finds that they are sort of all over the map.  He says:

As they traversed the country completely they came into contact with both the colonial French and the Spanish sources of information, which were, of course, still divergent. There were several possible conclusions which could be drawn from such a situation and the American explorers, collectively, arrived at all of them, the differences being caused by the varying specific conditions under which a given explorer obtained his information.

However, when it comes to Lewis and Clark specifically, there is some evidence that the people they identified as “Padoucas” were in fact Apaches, or at least Athapaskan-speakers, despite the fact that their information came mostly from French or Siouan sources. Secoy notes this without explanation, but Grinnell goes into a bit more detail. From a table of tribes that the expedition apparently compiled from information gained at the villages of the Mandans (Siouan-speakers), Grinnell summarizes several entries listing tribes said to speak the Padouca language. The “Padouca” entry is most informative:

Padoucas—English name, French nickname Padoo, Padoucies is their own tongue. Live in villages on heads of Platte and Arkansas, trade with New Mexico; many horses. Yet almost immediately Clark says he could get no definite information about this once powerful nation, and quotes French writers. Speaks of a fork of the Platte bearing the name of the tribe and conjectures that the nation had broken up and become individual small tribes.

It’s not clear how much, if any, of this is quoted rather than paraphrased by Grinnell. Nevertheless, the bit about the Padouca being a “once powerful nation” is presumably the source of the information on them in Clark’s later letter to his son. The fact that Clark couldn’t get much information on them from his current sources and had to rely on (highly outdated) French written sources is one possible reason to think that the group referred to here is Apache rather than Comanche, and other references in the Expedition’s journals to a Padouca-speaking group called the “Cataka” have often been thought to refer to the Na’isha or “Kiowa Apaches,” including by Secoy. Still, if the French by this time were using the term “Padouca” to refer to the Comanches, and Lewis and Clark were getting their information from French informants, why would they call the Apaches “Padoucas”?

Neither Grinnell nor Secoy addresses this question, but I think the answer lies in where Lewis and Clark where when they got this information. Remember, all that stuff about the Comanches pushing out the Apaches and the French inadvertently switching the meaning of the word “Padouca” is about the Southern Plains. Lewis and Clark, however, were traveling across the Northern Plains, following the Missouri River. The Mandan villages where they compiled the list of tribes including the Padoucas were in what is now North Dakota. The Comanches never came close to this part of the Plains, which had a rather different cultural sequence in the historic period. Part of the difference, I suspect, is that the French traders and their Siouan trading partners on the Upper Missouri continued to use the term “Padouca” to refer to Athapaskan groups to the south and southwest, including those who had, unbeknownst to them, recently been displaced by the Comanches from their homelands in what is now eastern Colorado and those that might have still been somewhere on the Northern or Central Plains at this point, north of the Comanches. The former are probably the ones that Clark refers to in the section quoted above, and the lack of information on them is probably due to the fact that any ties they had had to the north had been severed by the Comanche advance without the northern groups necessarily knowing what had happened. The latter would probably include the ancestors of the Na’isha, who may or may not already have been affiliated politically with the Kiowas at this point but probably were somewhere in the Northern Plains. Grinnell notes an account of a Na’isha man who claimed to have been born on the Missouri River northeast of the Black Hills around 1810.

Mural of the Lewis and Clark Expedition Passing the Future Site of Paducah, Kentucky

Thus, I think the most reasonable interpretation of the Padouca question is that the term, meaning “Piercers” in some Siouan language, was used for the Apaches of the Plains by various Siouan-speakers (and others?) living along the Mississippi and Missouri Rivers in the seventeenth century, and was taken over by French explorers at this time. Later French officials in Louisiana continued to use the term for Plains groups without realizing that the Southern Plains had been taken over by Comanches in the mid-eighteenth century, and subsequent French expeditions applied it to the Comanches rather than the Apaches. At the same time, the more southerly Siouan-speakers shifted the referent of the term as well, perhaps under French influence but not necessarily. Further north, where both the French and the Siouans were too far away from the Comanches to encounter them, the term continued to refer to Plains Athapaskans, both the nearby Na’isha and the more distant Apaches about whom the northerners knew little. Clark learned about the Padoucas from Upper Missouri groups who didn’t know much about them, and supplemented this information with French written sources which were also not very well informed. From this mix he came up with an idea of the Padoucas as a tribe that had once been numerous and powerful but had been decimated by European contact. Perhaps entranced by this romantic idea, when he acquired a bunch of land many years later and decided to build a town on it he decided to name it after this “lost” tribe. And that’s the story of Paducah.

Grinnell, G. (1920). Who Were the Padouca? American Anthropologist, 22 (3), 248-260 DOI: 10.1525/aa.1920.22.3.02a00050

Michelson, T. (1921). Who Were the Padouca? American Anthropologist, 23 (1), 101-101 DOI: 10.1525/aa.1921.23.1.02a00120

Secoy, F. (1951). The Identity of the “Paduca”; An Ethnohistorical Analysis American Anthropologist, 53 (4), 525-542 DOI: 10.1525/aa.1951.53.4.02a00060

Sign at Kluane Lake, Yukon Territory

In 1898 Washington Matthews, the US Army physician who was one of the earliest and best recorders of ethnographic information on the Navajos, published an article in the Journal of American Folklore entitled “Ichthyophobia.” It’s an interesting piece of scholarship for a number of reasons, not least its florid Victorian prose style. Matthews begins thus:

By the term Ichthyophobia I mean, of course, fear of fish; but I do not mean that proper fear, based upon actual knowledge, which the native diver of certain tropic seas feels, who will not venture into deep water lest he be torn to pieces by sharks, nor that equally rational fear that leads us to discard tainted fish, which so often proves poisonous as an article of food. I refer to the fear which results from superstition, and which prohibits all fish as an article of food; in short, to the taboo of fish.

He soon goes on to explain the background to his interest in this issue:

In the year 1866, after I had spent about twelve months on the Upper Missouri among some of the most primitive tribes then within our borders, I came on to Chicago, and there made the acquaintance of a gentleman who had recently returned from New Mexico, having spent a year or more among the Navaho Indians. Oddly enough the gentleman’s name was Fish, although this fact, like the vernal blossoms, had nothing to do with the case, since the Indians did not fear him. In comparing notes of our experience among the Indians, he asked me, “Do the tribes of the Upper Missouri eat fish?” “Of course they do,” I said. “Is there any one in the world who will not eat a good fish if he can get it ?” “Yes,” he replied; “the Navahoes will not eat fish; they will not even touch a fish, and I have known them to refuse candies that were shaped like fish.” At the time, although I had every reason to believe that my friend was a truthful person, I was half inclined to believe that his was a “fish story” in more senses than one, or that he had made some error in observation. But that was in the days of my youthful ignorance. I knew not then the extent and nature of the customs of taboo. I did not realize that I was myself the victim of taboo practices just as unreasonable as that of the Navaho fish-haters.

Fourteen years later I found myself a neighbor of these same Navaho Indians, and one of the first subjects I proceeded to investigate was the fish taboo, of which I had learned years before. I found that my friend, Mr. Fish, had told me the truth,  but had not expressed his case as strongly as he might have done. I found that the Navahoes not only tabooed fish, but all things connected with the water, including aquatic birds. Speaking of the Navaho repugnance to fish with the landlady of the Cornucopia Hotel (a slab shanty) at Fort Wingate, she related the following as a good joke on the Indian. She employed  a young Navaho warrior to do chores around her kitchen. The Navaho warrior has no pride about the performance of menial labor. He will do almost anything at which he can earn money, and this one would do any work for her but clean fish. He would eat, too, almost anything in her kitchen except fish. Noticing his aversion to the finny tribe, she one day sportively emptied over his head a pan of water in which salt fish had been soaked. The Indian screamed in terror, and, running a short distance, tore in haste every shred of clothing from his body and threw it all away. She learned that he afterwards bathed and “made a lot of medicine” to purify himself of the pollution. He never returned to work for her, so this little trick cost her a good servant.

Anthropologists don’t write like this anymore, which seems like a pity to me. That aside, while Matthews certainly shows his fair share of nineteenth-century racism in what he writes here, he goes on to draw an interesting conclusion that is in some ways well ahead of his time:

Our philanthropists wonder at the reluctance of Indians to send their children to a distance to school, and think it is but foolish stubbornness. They cannot realize that, in addition to many practical and sentimental reasons, there are long-cherished religious scruples to be overcome—reasons which are the most potent of all—and, among these, not the least is that they know their children will be obliged to violate tribal taboos. The Navahoes have heard from returning pilgrims that the boy who goes to the Indian school in the East may be obliged to eat geese, ducks, and fish, or go hungry; or that, if  he eats not at first of these abominations, he may be ridiculed and chided till he changes his customs.

“What foolish scruples!” we say, and yet fail to realize that we all refuse certain edible and wholesome articles as food for no good reason that we can assign. What civilized father would send his child to a distant boarding-school where he might be obliged to eat stewed puppy? Yet I have been informed by those who have tasted it that it is a very palatable dish. But we can find a better illustration of our case than this: There are many among the most cultured of our Christian communities who, for religious reasons, refrain on certain days and at certain seasons from articles of food which at other  times are eaten. Such persons would not willingly send their children to places where they would be compelled to disregard these fasts. We may all understand and approve the sentiments which actuate them; yet we seem unable to extend an equal consideration to savages who are, perhaps, actuated by equally worthy motives. Often among the Navahoes children returning from eastern schools fall into feeble health. Their illness is almost always attributed to the violation of taboo while they were away from home, and costly healing ceremonies are performed in order to remove the evil effects of the transgression.

Matthews was unusual in his generation, even among anthropologists, in taking Indians seriously as people and trying to look at things from their perspective. He is best known for his extensive documentation of Navajo legends and ceremonial practices, including sandpainting designs. His success at eliciting this information was likely due in part to his status as a physician and a sense among the medicine men he talked to that he was in some sense engaged in the same trade as they were, that of healing. His respectful attitude toward his informants, evident in his discussion here of their concerns about the possible effects of boarding school on their children, likely also played a role.

Although Matthews clearly considered it important to emphasize to his audience that the Navajo taboo on eating fish is not some absurd result of “savages” being “irrational,” the main point of his article is to look into how this taboo may have arisen. He notes that the Apaches have a similar taboo, and quotes at length a then-recent article by one P. C. Bicknell, who had recently explored the mountains of east-central Arizona, inhabited by the Western Apaches, and had noticed the same taboo among them and inquired into the reasons for it.  Bicknell eventually got an explanation that the Apaches had long ago, when there was not enough food for everyone in the region, made an agreement with the Mohave and Yuma tribes who lived along the Colorado River to the west. Under this agreement, the Apaches would eat no fish, while the river tribes would eat no venison, and therefore everyone would have enough to eat. Neither Bicknell nor Matthews found this explanation convincing, and Matthews uses it as another example of the parallels between Indian and Anglo-American society:

The story here related, which is wisely discredited by Mr. Bicknell, may have been coined for the occasion; but it is more likely that it has been current for some time among the Indians. White men are not the only ones who are importunate to know the why and the wherefore. The inquisitive small boy whose business in life it is to ask questions exists among the savage as well as among the civilized; and there are boys of older growth who pester their seniors for explanations. To satisfy the mind of the inquirer with something in accord with his mode of thought, with the grade of philosophy which he has reached, is the aim of the man, in all ages of the world, who would gain and retain a reputation for wisdom. Milton’s Adam explains everything to Milton’s Eve according to the philosophy of Milton’s time. Modern science has its myth-makers, no less than the wild Apache.

Matthews does however find a possible clue to the actual origin of the taboo among both Navajos and Apaches in another of the explanations given to Bicknell. One of the Apaches he asked, who apparently had very limited proficiency in English, just said “all same water.” Bicknell interpreted this to mean that fish is just as insubstantial and tasteless as water, but Matthews considers this implausible, since if the Apaches have been avoiding fish for generations they presumably don’t actually know what it tastes like. He sees this instead as an indication that it is the association of the fish with water, which is so scarce and precious in a desert land, that accounts for its avoidance. He checks in with Frank Cushing, the preeminent expert at the time on the Zunis, to see if they have a similar taboo, and receives a reply which he quotes:

The Zuñis, like the Navahoes, will not, under any circumstances, eat fish or any other water animal. The reason is this: Abiding in a desert land, where water is scarce, they regard it as especially sacred; hence all things really or apparently belonging to it, and in particular all creatures living in it, are sacred or deified. But, in the case of the fishes, they eat water, chew it, and are therefore, since they also breathe water and the currents or breaths of water, especially tabooed. The Zuñi name for the Isletas is Kyas-i-ta(w)-kwe, Fish Cannibals, because they ate fish formerly.

Matthews considers this ample confirmation of his own conclusions regarding the Navajo avoidance of fish, which seems to be a common (but not universal) trait of Southwestern groups. Furthermore, he notes that the northern Athapaskans, who speak languages quite similar to those of the Navajos and Apaches but live far to the north, in generally well-watered areas of Canada and Alaska, do not have any sort of taboo on fish consumption. To ensure that he hasn’t missed something on this topic he checks with Franz Boas, the towering figure in American anthropology who had done extensive fieldwork among the Northern Athapaskans. Boas replies:

The northern Athapascan tribes have no taboo against fish; on the contrary, they almost subsist on fish for a considerable  part of the year.

Indeed, salmon in particular is hugely important to the diet of many of the Athapaskan groups in Alaska to this day. Matthews draws the reasonable conclusion from all this that the Navajos and Apaches likely acquired their fish taboos after reaching the Southwest, probably under the influence of the Pueblos, although the arid environment itself may have played a role directly as well.

Many years after Matthews’s paper, Herbert Landar presented further thoughts on the linguistic implications of all this. He notes that Navajo basically has only one word for “fish”: łóó’, a generic term referring to all fish.  This echoes the situation in Hopi and Zuni, both of which only have one general “fish” term, but is quite different from the extensive inventory of terms for various fish found in northern Athapaskan languages. These languages do tend to have terms cognate to the Navajo one, and these terms usually refer either to fish in general or to salmon or whitefish specifically. They also have a wide variety of terms for other specific fish and aquatic creatures, cognates for which are apparently totally missing in the southern languages. Landar concludes from this, in keeping with Matthews’s conclusion (though rather oddly he does not cite Matthews’s article) that “a prehistoric southwestern fish taboo led to the truncation of the Apachean fish vocabulary.”

At the time Landar was writing in 1960, the Alaska Athapaskan languages were still not very well documented compared to those in Canada and the Southwest. It was not until the establishment of the Alaska Native Language Center in 1972 that extensive, systematic documentation of all these languages began. The data collected by the ANLC has greatly increased both the ease and the reliability of the kind of comparative study done by Landar, which in his case given the material he had to work with was necessarily very tentative.  As far as I know no one has yet looked at this exact issue using that data, but it would be interesting to see exactly how many “fish” words each Athapaskan language has and how specific they are. Be that as it may, however, the conclusions reached by Matthews and Landar using much less information are likely to stand the test of time.

Landar, H. (1960). The Loss of Athapaskan Words for Fish in the Southwest International Journal of American Linguistics, 26 (1) DOI: 10.1086/464559

Matthews, W. (1898). Ichthyophobia The Journal of American Folklore, 11 (41) DOI: 10.2307/533215

Hazel Gates Woodruff Cottage, University of Colorado, Boulder, Colorado

This video has attracted some attention in certain corners of the internet.  It features a (very talented) male actor doing a pitch-perfect impersonation of a young woman saying various expressions that are strongly stereotyped as “female” in contemporary American English.  One thing that struck me about watching the video was how it shows how language reflects social relations and cultural norms.  In English gender roles are encoded into language primarily through prosodic and syntactic differences between typical male and female speech, but in some other languages gender differences are much more rigid and formalized, in that women and men not only speak with different intonation patterns and sentence structure but with different words and even different grammatical constructions entirely.

There are many examples of this, particularly in Asia, and it is present in Native American languages as well.  Probably the best documented example is in Koasati, a Muskogean language of the Southeastern United States, which has very distinct male and female versions that have been studied for several decades.  I haven’t read any of that research, but I have found some papers on similar phenomena in the languages of the Southwestern Pueblos.

The Pueblos, despite their very close cultural similarities, speak several distinct languages, only some of which are related to each other.  Many of the Pueblos of the Rio Grande valley in north-central New Mexico speak languages belonging to the Tanoan language family, the surviving members of which are known as Tiwa, Tewa, and Towa.  (There may have once been other languages belonging to this group that have died out since the Spanish conquest, but they are not well documented and they could well have instead been dialects of one or more of the extant languages).  Also spoken in the Rio Grande Valley, as well as at Acoma and Laguna further west in New Mexico, is Keres, a linguistic isolate not known to be related to any other language.  Zuni, spoken only at Zuni Pueblo in western New Mexico, is another isolate.  Finally, the Hopis in northeastern Arizona speak a language belonging to the large and far-flung Uto-Aztecan family.

Although these languages are distinct and unrelated to each other, they share certain features suggesting that they have influenced each other over the centuries or millennia that their speakers have been in contact with each other.  Gendered language, by which I mean different linguistic forms used by men and women, may be one example of this “areal” influence.  Paul Kroskrity, whom we saw earlier discussing an instance of possible influence on one of these languages, Tewa, by non-Pueblo Athapaskans, wrote a short article in 1983 pulling together some documentation of different terms used for common, everyday expressions by men and women in three Pueblo languages.  The languages are Hopi, Tewa, and Keres, and the terms are mostly for phrases like “thank you,” “yes,” and other things that people would likely say often, which Kroskrity notes would tend to reinforce the distinction between male and female speech even if the actual differences were few, as they seem to be in at least some of these languages.

The terms themselves don’t correspond closely at all among the different languages, which leads Kroskrity to conclude that they likely were not borrowed from one language to another as words, although the concept of distinct male and female speech forms may have diffused among the Pueblos.  Interestingly, however, he notes that Zuni apparently lacks any comparable distinction.  Unlike in some other languages with similar distinctions, there don’t seem to be consistent sound correspondences between the male and female forms within any given language, although the Tewa and Keres forms do tend to show some general similarities.  The Hopi forms mostly seem to be completely different.

A few years after Kroskrity’s article, Christine Sims and Hilaire Viloquette published one giving more data on the distinction in Western Keres specifically and challenging some of Kroskrity’s conclusions.  While Kroskrity had thought there were no particular patterns to the relationships between male and female forms, Sims and Viloquette show that there were some errors in Kroskrity’s data and present more of their own showing that at least for the set of what they call “cue words,” expressing “the speaker’s emotional relationship to the content of a sentence,” (i.e., love, discomfort, fear) there is a consistent distinction, in that the male versions consistently have a long vowel with a falling tone while the female versions have a short vowel.  This prosodic distinction reminds me in an interesting way of the intonational differences between male and female speech in English being lampooned in the video I linked at the beginning of the post.  They also note some other gendered distinctions, some of which don’t follow this same pattern and appear to be largely obsolete or archaic, appearing primarily in old recorded texts and among a few older speakers.  As for how this came about, they are oddly hostile to Kroskrity’s conclusion that it resulted primarily from linguistic diffusion among the Pueblos.  (Indeed, throughout the article they seem oddly hostile to Kroskrity for reasons that are unclear.)  They don’t really have a better explanation, though, except to suggest that it was not the linguistic construction per se that diffused but rather the social structure that ended up being reflected linguistically, which is a fair point.  Pueblo cultures are certainly much more like each other than Pueblo languages.

If Keres is typical of Pueblo languages in this respect (and it’s not at all clear that it is), it seems to indicate that gendered language may be a relatively recent development, not deeply entrenched in the grammar as it is in some Asian languages, and that it may in turn have arisen at one of the various inflection points in Pueblo (pre)history, perhaps at one of the times that various groups who had previously lived separately began to rapidly aggregate into much larger communities, necessitating new social relationships and perhaps some changes in language.  There’s been a fair amount of work on understanding the transformation of material culture during these periods of aggregation, but much less attention to their possible linguistic consequences.  It’s quite plausible, however, that these aggregating groups didn’t all speak the same language initially, and some of the oddities of the modern Pueblo languages may be best explained as relics of one or more periods of rapid linguistic change spurred by ecological, political, or cultural pressures.

Kroskrity, P. (1983). On Male and Female Speech in the Pueblo Southwest International Journal of American Linguistics, 49 (1) DOI: 10.1086/465769

Sims, C., & Valiquette, H. (1990). More on Male and Female Speech in (Acoma and Laguna) Keresan International Journal of American Linguistics, 56 (1) DOI: 10.1086/466144

Looking South from Kin Ya'a

One of the most notable examples of an assemblage of highly mutilated human remains from the Southwest being attributed to witchcraft execution rather than cannibalism, in accordance with J. Andrew Darling’s theory discussed in the previous post, is Ram Mesa, southwest of Chaco Canyon near Gallup, NM.  This site was excavated by the University of New Mexico as a salvage project, and the relevant assemblage was reported by Marsha Ogilvie and Charles Hilton in 2000.

The Ram Mesa assemblage, consisting of 13 individuals, is pretty similar to many other assemblages in the Southwest attributed to cannibalism, but Ogilvie and Hilton make a plausible case that while the remains are clearly highly “processed” there isn’t a whole lot tying this dismemberment and mutilation to actual consumption of the remains.  Few of the bones showed any evidence of burning, a condition which applies to several other cases of alleged cannibalism as well.  The few cut marks, which were mostly found on children’s skulls and lower jaws, weren’t particularly indicative of the removal of large muscles that might be expected if consumption were the object.  On the other hand, however, relatively few of the bone fragments were sufficiently large to be identified to body part, and any diagnostic evidence from these tiny fragments was clearly destroyed by the thoroughness of the processing.  It’s not clear, therefore, how representative the larger fragments with surviving evidence of burning and cutting are of the entire assemblage.  The most I would say about this site is that the evidence is not sufficient to make a positive diagnosis of cannibalism, and other explanations are therefore plausible.

However, as I noted before in discussing Darling’s arguments, witchcraft execution and cannibalism are not necessarily mutually exclusive.  Indeed, the execution of suspected witches may well have involved some level of cannibalism among some Southwestern groups in prehistoric times, thought there is certainly no evidence that it did in historic times as documented by ethnographers.  There are some other oddities about the Ram Mesa site that suggest that it might not be expected to pattern with the majority of the suspected cannibalism assemblages, so it is certainly possible that it represents a variation on the same behavior that may not have included cannibalism.

For one thing, this is an odd place for one of these assemblages.  Although some early excavations at Chaco Canyon and in northern Arizona have been proposed as showing evidence of cannibalism, the vast majority of the well-documented cases are in southwestern Colorado, especially around the modern town of Cortez.  This includes the Cowboy Wash site, the site with the best evidence for cannibalism of any of them.  Given the known cultural differences between prehistoric populations at the northern and southern edges of the San Juan Basin (the San Juan and Cibola Anasazi, respectively), it’s quite possible that the cultural activities resulting in similar assemblages in these two areas may have been somewhat different, with the San Juan groups practicing cannibalism and the Cibola groups not.

Furthermore, there may be differences in the dating of the sites.  Most of the well-documented Cortez-area sites date to right around AD 1150, and they may all represent part of a single event at that time, which was in the midst of a severe drought when social structures were likely under extreme stress.  The Ram Mesa site is dated by six radiocarbon dates to a period that Ogilvie and Hilton describe as “AD 978 to 1161.”  They do clarify that these are calibrated dates, which is helpful, but it would have been better if they had shown the ranges for the individual dates, as well as the materials that were dated, which would give a better idea of the most likely dating for the human remains.  On the assumption that the remains date to the latest period of occupation, which seems plausible based on comparison to similar assemblages elsewhere, this puts the latest date at 1161, which is interestingly close to the dates for the similar Cortez sites.  Due to the lack of information of the dates, however, it’s not clear is this is an intercept (i.e., most likely) date or the late end of a range; if the latter, it’s possible that the assemblage dates to somewhat earlier than the Cortez sites.  In that case it would not be part of the same phenomenon, whatever that was, and the postulated lack of cannibalism may be related to that.

In any case, this site definitely seems to have been within the Chacoan sphere of influence, which makes the interpretation of the remains there important for understanding the relationship of alleged cannibalistic events to the rise and fall of Chaco.  Christy Turner has famously argued that they represent the expansion of the Chacoan system and the use of brutal force by the rulers of Chaco (hypothesized on very dubious evidence to be Toltec immigrants from central Mexico) to ensure that outlying communities were incorporated into the system and supplied tribute to the canyon.  This idea is pretty implausible based on the evidence from the Cortez area, where most of the assemblages date to the period of Chaco’s decline rather than its rise.  If Ram Mesa dates to the same period it would support that evidence, whereas if it dates to earlier it could conceivably either support Turner’s ideas or point to a different interpretation, perhaps having something to do with the well-known fact that the outlying Chacoan communities to the south of Chaco seem to have been abandoned beginning much earlier than those in other directions.  There are a lot of outlying Chacoan great houses in this area, including Casamero and Kin Ya’a, but they seem to have rather different histories than those to the north, such as Aztec Ruins and Yellow Jacket.

Like most research related to Chaco, this paper ultimately raises more questions than it answers.  That’s not necessarily a bad thing, however, especially when it comes to a topic as controversial and poorly understood as these assemblages suggesting cannibalism.

Ogilvie, M., & Hilton, C. (2000). Ritualized violence in the prehistoric American Southwest International Journal of Osteoarchaeology, 10 (1), 27-48 DOI: 10.1002/(SICI)1099-1212(200001/02)10:13.0.CO;2-M

Fort Smith Gallows, Fort Smith, Arkansas

Last year around Christmas I did a series of posts on the evidence for cannibalism in the prehistoric Southwest.  I didn’t cover nearly all that there is to say about this important but controversial issue then, so I figured it would be a good idea to discuss it a bit more this year.  In this post I would like to take a look at what has become the main alternative explanation for the assemblages of human remains that are often seen as evidence of cannibalism: witchcraft execution.

The idea that the assemblages of broken, burned, and otherwise heavily “processed” human bones found in various parts of the Southwest represent the remains of the execution of alleged witches originates mainly in an article by J. Andrew Darling published in 1998.  In it, Darling notes that most of the evidence for cannibalism in these assemblages has come from osteological analyses of the bones conducted by physical anthropologists, with little contextual analysis by archaeologists or ethnographers that would allow better understandings of what the assemblages mean in cultural terms.  This is more or less true as of when he was writing, although more recent analyses have been much more thorough (perhaps in part as a response to critiques like Darling’s).  Importantly, Darling does not contest the osteological evidence.  He accepts the conclusions of physical anthropologists that it represents substantial “processing” of human remains around the time of death or shortly afterward, and he even agrees that the specific actions indicated are strongly reminiscent of the evidence for butchering found on animal bones.

Where Darling parts ways with the physical anthropologists, however, is in inferring from this evidence for butchering that the intent of this processing was to prepare human flesh for consumption.  He points out, again rightly, that little is known about the butchering of animals for purposes other than consumption, and that it cannot therefore be assumed that any processing of remains, human or animal, necessarily indicates consumption of those remains.  In other words, Darling agrees that people were butchering and mutilating the bodies of other people, but he does not agree that this means they were eating them.

At this point I think it is worth pointing out that Darling’s acceptance of this much of the physical anthropological evidence indicates how much the cannibalism side had already won the debate at this point.  Indeed, the objections of Darling and those who cite him may reveal more about the societal mores of modern anthropologists than those of ancient Puebloans.  ”Yes, okay, they did gruesomely butcher and dismember people, but that doesn’t mean they ate them” is a rather odd line of defense to offer against accusations of cannibalism.

Note also that Darling was writing before the publication of the data on the Cowboy Wash assemblage, including the infamous coprolite that tested positive for human muscle tissue.  His arguments that the evidence of butchery on the bones doesn’t necessarily indicate consumption of the flesh therefore stand on their own, but they don’t apply at all to the coprolite evidence, which is just about the most solid evidence for actual consumption of human flesh imaginable.

Leaving all that aside for the moment, Darling’s alternative explanation for the heavily processed assemblages is that they represent the remains of executions of suspected witches.  His evidence for this draws mostly from modern Pueblo ethnography, especially records of witchcraft trials at Zuni in the late nineteenth and early twentieth centuries as well as ceremonies conducted at several Pueblos to cure ailments attributed to witchcraft.  He notes, importantly, that this does not mean he is merely projecting the ethnographic present into the past; indeed, the modern activities he describes would likely not account for anything like the prehistoric assemblages.  The Zuni trials, many of which were subject to intensive interference by Anglo anthropologists and government agents, provide interesting data on Zuni attitudes toward witches, but they resulted in very few executions, and there is little to no data on the way those few executions were carried out.

For data that could possibly account for the intensive processing of the remains of executed witches, Darling points to the ceremonies, which in modern times largely consist of the symbolic hunting, capture, dismemberment, and burning of the witches in the form of effigy dolls.  Combining this with the evidence from the Zuni trials, the idea seems to be that at some point in the past similar “trials” often ended up with executions carried out similarly to the ritual destruction of the dolls.  This is a bit speculative, but it’s not a hugely implausible theory.

It’s not totally clear, however, that this is actually an alternative to the cannibalism theory.  I think it could just as easily be seen as a supplement to it, with the intensive processing of executed witches at some times in the past also involving consumption of their flesh. Darling doesn’t address this objection specifically, but it seems like it has occurred to him, as he makes a big deal about how cannibalism itself is strongly associated with witchcraft in Pueblo thought.  Indeed, one of the reasons he argues the cannibalism theory is wrong is that this association is so strong that Pueblos would never engage in cannibalism, even under conditions of severe subsistence stress, for fear of either being thought of as witches or even actually becoming them.  This is fair enough, as far as it goes, but Darling doesn’t seem to consider the possibility that some of these attitudes toward witchcraft and cannibalism may have developed relatively recently, perhaps even in reaction to prehistoric incidents of cannibalism.  Whether or not the people who dismembered and perhaps cannibalized others in the twelfth century interpreted what they were doing as punishing witches, it seems pretty plausible to me that later Puebloans reinterpreted those events as the actions of witches themselves, and incorporated cannibalism into their cultural ideas of witchcraft.

Overall, Darling’s article raises some important points, and presents some interesting data on Pueblo witchcraft ideas, but it is highly unconvincing in presenting an alternative to the cannibalism theory.  As noted above, it is totally inadequate as a response to cannibalism evidence that comes directly from coprolites rather than indirectly from bones, and the alternative theory it proposes is actually not necessarily incompatible with cannibalism at least in some instances.  The fact that some archaeologists continue to rely on it as the basis for arguing against cannibalism as an explanation for certain assemblages just goes to show how few other interpretations are really plausible at this point.

Darling, J. Andrew (1998). Mass Inhumation and the Execution of Witches in the American Southwest American Anthropologist, 100 (3), 732-752 DOI: 10.1525/aa.1998.100.3.732

Geronimo Birthplace Marker, Gila National Forest, New Mexico

The term “Apache” is one of the most widely known names for Native American groups, but it’s actually quite problematic.  There is, I think, a general perception that it refers to a specific “tribe,” but it doesn’t.  What it really is, at least as it’s used today, is a designation for all the Southern Athapaskan groups except the Navajos.  These groups include the Mescalero and Jicarilla Apaches in New Mexico and the Western Apaches (several closely related contiguous groups) in Arizona.  There are also some smaller groups that have largely merged with the others, most notably the Chiricahuas, who famously fought the US Army for many years under Geronimo and other famous chiefs but have since mostly merged with the Mescaleros.  These various Apaches, however, don’t form a single group linguistically.  Some of the languages, especially Western Apache, are closer to Navajo than they are to some other “Apache” languages.  This makes “Apache” polyphyletic in classificatory terms.  In other words, there is no group of “Apache” languages the members of which are more closely related to each other than to other Athapaskan languages.

This fact has been recognized for a long time.  Harry Hoijer published an initial attempt at classifying the Southern Athapaskan (or “Apachean”) languages in 1938.  In it he posited that all the Southern Athapaskan languages form a single genetic unit divided into Eastern and Western branches, with Navajo, Western Apache, Mescalero, and Chiricahua all belonging to the Western branch, with Jicarilla and the apparently closely related Lipan belonging to the Eastern branch.

Also belonging to the Eastern Apachean branch in this classification were the so-called “Kiowa Apaches,” although Hoijer recognized that their language diverged from all the other Apachean languages in important ways.  This group has long been a puzzle for ethnographers.  They have been called both “Kiowa Apaches” and “Plains Apaches,” but some anthropologists have preferred to call them “Na’isha” after their name for themselves, and I will follow their example for reasons that will soon become apparent.  Culturally and politically, they are very closely connected to the Kiowas, who speak a totally unrelated language (which has its own puzzles, but that’s a separate matter) and as of the nineteenth century lived way out on the Southern Plains, far removed from any other Athapaskans groups.  The Na’isha, however, do not speak Kiowa but their own language, which is clearly Athapaskan.  Despite the clear divergences between it and the other Southern Athapaskan languages, it does also show some similarities to the Eastern Apachean languages specifically, and Hoijer therefore classified it in 1938 as an Eastern Apachean language with some major divergences from the others, in the same way that he considered Navajo a Western Apachean language with some divergences.  Later in life, however, after fuller data became available on a wider variety of Athapaskan languages outside the Southwest, he seems to have changed his mind and reclassified Na’isha as an Athapaskan language closely related to the Apachean languages but not more closely related to them than to some other Athapaskan languages found further north.  In 1985 Martin Huld published an article pointing to some additional differences between Na’isha and the Apachean languages that further support the idea that Na’isha is not Apachean.

The picture emerging from this research is that both Na’isha and Proto-Apachean seem to have separated from other Athapaskan languages somewhere on the far northern Plains, perhaps in southern Alberta.  The Athapaskan language still spoken in the Calgary area, Sarcee, also shows some similarities to both Na’isha and Apachean.  The implication of these conclusions is that all of these languages were once spoken in the same area by groups in close contact with each other, resulting in many similarities in the languages.  Crucially for understanding the prehistory of the Plains, however, the Na’isha seem to have split off and headed south separately from the Apachean speakers.  This model is in strong contrast to other models which have the Na’isha splitting from the other Apacheans after the latter had moved to the Southwest, and it is more consistent with Na’isha and Kiowa oral tradition, which holds that they had been associated with each other for a very long time and came to the southern Plains from the north.

So, it seems that the “Kiowa Apaches” are neither Kiowa, despite close association with the Kiowas, nor Apache, despite their Athapaskan language.  Instead, they most likely represent an additional Athapaskan migration to the south, separate from the migration(s) that brought the Navajos and the various other Apaches to the Southwest.  This is why both “Kiowa Apache” and “Plains Apache” are very problematic as terms; the Na’isha are probably not “Apaches” in the linguistic sense, nor are they Kiowas, and they are not the only Athapaskans to occupy the Plains either.  The Eastern Apachean Jicarilla and Lipan Apaches also lived on the Plains, to the west of the Kiowas and Na’isha, closer to their Western Apachean brethren.  These groups are, indeed, often considered intermediate culturally between the Southwest and the Plains.  They were in contact with the Pueblos, and the Jicarillas even farmed to some extent, but they also hunted bison and had various Plains cultural traits, although to a lesser extent than the Na’isha, Kiowas, and Comanches, the preeminent bison-hunting cultures of the southern Plains.

The importance of this conclusion for archaeology is that the “Athapaskan migration” that has been so elusive in the archaeological record likely consisted not of a single large group but of multiple smaller groups which would be very hard to find evidence for, some of which later coalesced into the Apache groups known from the ethnohistoric record.  This is another example of the huge and largely untapped potential for linguistics to contribute to archaeological problems.

Hoijer, H. (1938). The Southern Athapaskan Languages American Anthropologist, 40 (1), 75-87 DOI: 10.1525/aa.1938.40.1.02a00080

Huld, M. (1985). Regressive Apicalization in Na’isha International Journal of American Linguistics, 51 (4) DOI: 10.1086/465932