Archive for the ‘Gathering’ Category


Trail in Brazos Bend State Park, Texas

The big story in the news these days is of course Hurricane Harvey, which has been battering the Gulf coast and adjacent areas of Texas and Louisiana for days now. While it has so far probably done the most damage in Houston, with record rainfall leading to massive flooding in one of the country’s biggest cities, Harvey first came ashore further south, near the small town of Rockport, Texas just north of Corpus Christi. Rockport was very severely damaged by the wind and rain, of course, and has gotten quite a bit of media attention for that.

Rockport has another claim to fame, however, at least for those of us interested in archaeology and prehistory: it is the namesake of the Rockport Phase, an archaeological complex that existed on the central part of the Texas coast in the late prehistoric period and is generally thought to be directly ancestral to the Karankawa people who occupied the same area at European contact. The Karankawa are among the better-documented of the many cultural groups that occupied the Gulf Coast, partly because of the detailed account of them left by Álvar Núñez Cabeza de Vaca, who was shipwrecked in this area in 1528 and spent several years living with the natives here and further west as he made his way back to his Spanish compatriots in Mexico. Archaeological research over the past few decades has both confirmed some aspects of this and other historic accounts and added additional information about the culture history of this area.

The Rockport Phase is characterized by a distinctive type of pottery, gray in color with thin, hard walls and a sandy paste. It can be plain (i.e., undecorated), incised, or, most distinctively, decorated with the black asphaltum found in the Gulf area and associated with its extensive petroleum deposits. The beginning date for the Rockport Phase varies in the literature but is in the range of AD 1000 to 1250; the variation is probably due to the fact that Rockport is clearly continuous with the previous Late Archaic culture of the same area. In general, however, the Late Prehistoric period on the coast is defined by the appearance of the bow and arrow and pottery, both of which seem to have reached the central coast around AD 1000 from the north. (Note that this makes at least the beginning of Rockport roughly contemporary with Chaco Canyon far to the west.) As noted above, Rockport is also clearly continuous with the historic Karankawa, and Rockport pottery has been found on some early historic sites.

While pottery is often associated with agricultural people, agriculture was never practiced on the prehistoric Texas coast or, indeed, most of the interior areas of prehistoric Texas. The Rockport people, like their neighbors in all directions, were hunter-gatherers, and they appear to have had a subsistence system based primarily on the rich aquatic resources of the coastal estuaries but with seasonal movements inland to hunt terrestrial game and gather plant resources including pecans and the fruit of the prickly pear cactus.


Warning Sign, Brazos Bend State Park, Texas

The stone tool assemblage of the Rockport Phase, at least from around AD 1250 on, was very similar to that of the inland groups in central and southern Texas, all of which were part of the Toyah Horizon distinguished by the use of Perdiz arrow points. This widespread lithic complex is generally thought to be associated with the hunting of bison, which appear to have rapidly spread south from the southern Great Plains into central and southern Texas during the thirteenth century AD, possibly in response to a drying trend beginning a couple centuries earlier that expanded the grasslands favored by bison. Despite Rockport use of this lithic complex and the presence of bison bone in some Rockport sites, however, stable isotope studies of human remains from cemetery sites on the coast that are contemporary with Rockport have not shown evidence that bison was a substantial part of the diet, which seems to have been heavily based on fish and other marine resources. More research may clarify this apparent clash of different types of evidence.

Speaking of those cemeteries, they area also unusual among hunter-gatherers but quite common in prehistoric Texas, in both coastal and interior areas. Cross-culturally, use of cemeteries rather than isolated burials by hunter-gatherers tends to be associated with “packing” into small territories due to high population densities, as well as with “intensification” of production of subsistence resources, especially aquatic ones. Some archaeologists have proposed theories linking intensification, which includes but is not limited to the development of agriculture, to increased population density due to highly productive resources in certain areas, which also leads to packing into smaller territories. Some of these theories further predict that this will mean less use of terrestrial hunting and increased use of aquatic resources where they are available, and plant resources where they are not.

This type of theory has been tested in Texas and found to largely but not completely explain the distribution of cemeteries and other signs of packing and intensification. In the Rockport area, which clearly had a relatively high population density and depended heavily on the aquatic resources of the estuaries, the theory seems to work. It also works for the Rio Grande Delta area to the south, where the populous Brownsville Complex had its own type of pottery as well as various cultural influences from and trade ties to the Huasteca region of northeastern Mexico to the south. It doesn’t really account for the presence of cemeteries and other signs of intensification in the more sparsely populated areas of central and western Texas, however, where hunter-gatherer populations are thought to have been much lower. Clearly more research on this issue is required. Many of these characteristics are associated with “complex” hunter-gatherers such as those of the Northwest Coast, but I doubt any anthropologist would describe even the higher-density groups on the Texas coast as complex in that sense.

It doesn’t get as much attention as some other areas, and it certainly isn’t as flashy as the ruins in the Four Corners region, but the archaeology of Texas is actually quite interesting. The University of Texas has a great website called Texas Beyond History that provides a lot of information in an easily accessible. It wasn’t a major source for this post, but it’s still definitely worth checking out. We’ve been seeing a lot about Texas in the news lately, but there’s much more to it if you dig a little deeper.


Texas Flag and Sundial, Brazos Bend State Park


Read Full Post »

Rio Grande from Coronado State Monument, Bernalillo, New Mexico

Rio Grande from Coronado State Monument, Bernalillo, New Mexico

Chapter seven of Crucible of Pueblos brings us to the final geographical region covered by the volume: the Rio Grande Valley, at the eastern edge of Pueblo settlement for the period in question. As it happens, I’m currently visiting my mom in Albuquerque, so I’m actually in this region as I write this. (Today also happens to be my birthday; I’m 31.) The chapter is by Steven Lakatos and C. Dean Wilson, and in a lot of ways it echoes an earlier paper by Lakatos about the Rio Grande Developmental Period that I have discussed before. This chapter, however, discusses only the Early Developmental Period, defined as AD 600 to 900, and primarily focuses on the part of the region that the authors called the Middle Rio Grande Valley, defined as lying between the Rio Puerco of the East on the west, the Sandia and Manzano Mountains on the east, the Isleta area on the south, and the La Bajada escarpment on the north. This is because agricultural populations only occupied this restricted area of the region during the Early Developmental, expanding north of La Bajada only after AD 900 when there was a huge increase in regional population at the beginning of the Late Developmental Period.

The key point Lakatos and Wilson make about the Rio Grande is that the Early Developmental period was a time of low population density and gradual growth, with little change in material culture over hundreds of years. This is in striking contrast to the “boom-and-bust” pattern now richly documented for the Mesa Verde region during the contemporaneous Pueblo I period there. The picture of continuity is reminiscent of that proposed by the authors of the previous chapter for the Little Colorado region, but it’s worth noting that the major data gaps that plague the study of that region are less of an issue for the Rio Grande, which has a long history of intensive archaeological research continuing to the present day. Furthermore, Lakatos and Wilson present several lines of evidence supporting their conclusions, which seem pretty solid to me. Based on this evidence, it really does seem like the Early Developmental was a time of low population, slow growth, and cultural continuity.

As Lakatos and Wilson note, this is actually a rather surprising conclusion in the context of many theories about early agricultural societies. Most strikingly, there is no evidence here for a “Neolithic Demographic Transition,” in which the increased productivity of agricultural societies compared to hunter-gatherers leads to massive growth among early agriculturalists, with all sorts of ecological and social consequences. Some have argued that the Mesa Verde boom-and-bust cycle is a result of this process. In the Rio Grande, however, the adoption of agriculture does not seem to have resulted in this sort of population growth. This is definitely not for lack of arable land, as the Rio Grande Valley is one of the richest agricultural areas in the northern Southwest, and it was intensively farmed later in prehistory and into historic times. Rather, Lakatos and Wilson argue that the richness of the Rio Grande environment allowed for a mixed farming-foraging economic pattern with high residential mobility, in contrast to the more agriculture-dependent societies further west. The greater importance of foraging versus farming is supported by evidence from faunal assemblages and wear patterns on human remains, and the mobility by the fact that residential pit structures were rarely remodeled.

In keeping with low density and high mobility, the settlement pattern consisted of scattered hamlets, with only occasional evidence for “communities” of hamlets loosely grouped together with possible communal architecture such as “protokivas” or oversized pit structures. Sites were mainly located along the major rivers of the region: the Rio Grande itself, the Rio Puerco of the East, the Jemez. Architecture consisted of residential pit structures and surrounding activity areas, generally oriented toward the east or southeast (perhaps oriented to the winter solstice).

Rio Grande people also appear to have been in closer contact with remaining hunter-gatherers than populations further west. It’s not clear if Early Developmental populations resulted from the adoption of agriculture by existing hunter-gatherers in the Middle Rio Grande Valley or if there was some migration of already agricultural populations involved, but in any case the areas north of La Bajada and east of the Sandias/Manzanos were definitely still occupied by hunter-gatherers during this period, and it’s clear that there was a lot of contact between the two groups. This may have contributed to the greater importance of foraging to Early Developmental people and their differences from other Pueblo populations.

Sandia Mountains from Tent Rocks National Monument

Sandia Mountains from Tent Rocks National Monument

All that said, the Early Developmental people definitely were part of the Pueblo cultural tradition, and their material culture shows a lot of connections to populations to both the west and south. This is particularly true of pottery, which was dominated by plain gray ware similar to that of late Basketmaker groups on the Colorado Plateau, but with small amounts of a decorated white ware, San Marcial Black-on-white, which shows stylistic influence from Mogollon populations to the south but with technological characteristics more like those of early white wares to the west. Lakatos and Wilson mention one model of Southwestern prehistory under which early “strong patterns” of material culture originated in the San Juan Basin (ancestral to the Chaco system) and in the river valleys of the Mogollon region, with the Middle Rio Grande forming a “weak pattern” with influences from both but in varying combinations.

The clear picture that emerges from this is of a small population of forager-farmers moving around within the Middle Rio Grande area but maintaining their basic cultural features with little to no change for about 300 years, from AD 600 to 900. Then, in a development that is likely very important but poorly understood, there was a massive increase in population at the same time that agricultural groups for the first time began to occupy the higher areas about La Bajada. Lakatos and Wilson note that the timing of this change, while not as precise as might be ideal, seems to correspond closely to the collapse of the late Pueblo I villages in the Mesa Verde region and the major population movements involved with the depopulation of that area, including the apparent influx of people into the Chaco Basin that likely laid the groundwork for the Chaco Phenomenon.

It seems very plausible that the increase in population in the Rio Grande was linked to these developments, though exactly how is unclear. Material culture actually remained fairly stable and consistent with Early Developmental patterns across this transition, although architecture did become more standardized and San Marcial Black-on-white was replaced by Red Mesa Black-on-white as the main decorated ceramic type. The latter change, especially, suggests influence from the west, as Red Mesa is the main decorated type in the Chaco area and other parts of the southern Colorado Plateau during this same period. It’s possible, as Lakatos and Wilson suggest, that the increased population in the Chaco Basin directly spurred Middle Rio Grande populations to move northward, although it’s not clear how exactly this would have worked. Other possibilities are that populations from the intermediate areas, such as the Puerco of the East, began to move eastward in the Rio Grande Valley as a result of the population movements immediately to the west of them, perhaps pushing existing Rio Grande populations north, or that western populations were moving directly to the Northern Rio Grande area above La Bajada, “leap-frogging” existing populations in the Middle Rio Grande.

The fact that material culture continued to show local Rio Grande features throughout the region, however, suggests that some level of assimilation or cultural accommodation between the locals and immigrants was involved, rather than a more directly confrontational situation. It’s noteworthy that Lakatos and Wilson don’t discuss evidence for warfare or defensive features at all, which of course doesn’t mean those things didn’t exist but does suggest that they may have been less prevalent than in some other regions.

Turquoise-Encrusted Cow Skull, Santa Fe, New Mexico

Turquoise-Encrusted Cow Skull, Santa Fe, New Mexico

Another thing Lakatos and Wilson don’t discuss, but which seems particularly important to understanding these relationships, is turquoise, specifically that from the well-known mines in the Cerrillos Hills east of the Sandias. Turquoise is of course strongly associated with Chaco, and while not all of the turquoise there has turned out to be from Cerrillos, a substantial portion of it definitely was. Evidence for increased connections between the San Juan Basin and the Rio Grande area at the same time as the rise of Chaco as a regional center is very intriguing in this light. Could increasing demand for turquoise at Chaco have led to the Cerrillos mines being a “pull” factor leading western groups into the Rio Grande Valley? Could the mines have even led local Rio Grande groups, or mixed groups of locals and immigrants, to move further east, across the mountains and even up over La Bajada into the Santa Fe area, which may have become more attractive as increased immigration reduced the supply of land in the Middle Rio Grande? And what about those remnant hunter-gatherer groups east of the Sandias and north of La Bajada? What happened to them? Were they attacked and defeated by the encroaching farmers? Pushed out into areas further north and east? Assimilated into agricultural society, which even in the Late Developmental period had a strong foraging component? There are a lot of questions about this period in this area, and very little evidence on which to base any answers. Lakatos and Wilson recognize this and suggest some research directions that would be helpful in answering the remaining questions, although they don’t point out as many as I have here.

Overall, this is a very informative chapter that brings into the discussion of Pueblo I societies an area that is often left out of these discussions. It’s an area of crucial importance for understanding regional dynamics throughout the northern Southwest, however, so I’m glad it was included in this volume. This chapter concludes the geographical summaries in the book; the remaining chapters cover various thematic topics of interest in understanding the Early Pueblo period as a whole.

Read Full Post »

Paiute Brush Shelters, Pipe Spring National Monument

Paiute Brush Shelters, Pipe Spring National Monument

As I mentioned in the previous post, the most mysterious thing about the Fremont is what happened to them. Unlike the Anasazi, who obviously became the modern Pueblos, the Fremont have no obvious connections to any modern groups. Fremont sites appear to disappear around AD 1300 in most areas, although there is some regional variation and in part defining an end date depends on how you define “Fremont.” Since the practice of agriculture is closely associated with the Fremont complex, the latest dates of sites with clear evidence for agriculture is one convenient way to date the end of the Fremont. In their important 1998 review essay, David Madsen and Steven Simms give the following dates for the end of agriculture in different Fremont regions:

  • Uinta Basin: AD 1000
  • Parowan Valley and Great Salt Lake wetlands: AD 1100 to 1150
  • “Much of the Fremont region”: AD 1250 to 1300
  • Northwestern Colorado: After AD 1450

(Note that Madsen and Simms annoyingly cite their dates as “Before Present” without specifying what date they are using for the “Present” or whether these are calendar or radiocarbon years; in calculating the above dates I have assumed a “Present” of AD 1950 as conventionally used in radiocarbon determinations.)

The Madsen and Simms date for much of the region is very close to the “Great Drought” of AD 1276 to 1299 known from Anasazi sites to the south (assuming of course that their dates are in calendar years). Given the low precision of the radiocarbon-based Fremont chronology compared to the tree-ring based Anasazi one, however, it is risky to make too much of coincidences like this, and the wide variation across different Fremont sub-regions suggests that something more complicated is going on here than a simple reaction to a single prolonged drought. The Great Drought may well have affected Fremont farmers, of course, but the Fremont data are not clear enough to establish a definitive association. In general a rough date of AD 1300 for the end of Fremont in most areas is widely used and probably close enough for most purposes. It does seem that some form of the Fremont lifestyle persisted significantly longer in northwestern Colorado, which could have served as a refuge for Fremont farmers displaced from other areas.

In keeping with their general interpretation of Fremont as involving a wide variety of adaptive strategies and frequent movements of people between farming and foraging, Madsen and Simms interpret the end of the phenomenon as consisting largely of farmers switching to foraging, along with possible immigration of foragers from outside the region. Basically they see this period as a time when the precarious balance between farming and foraging characteristic of the Fremont period tipped decisively in favor of foraging, perhaps in response to climatic changes that made foraging a more effective subsistence strategy.

Whatever the mechanism for the collapse of Fremont as an archaeological complex, the question of what became of the people remains. There are three main logical options:

  1. They died out entirely and left no descendants.
  2. They changed their culture and stayed in the same region.
  3. They left the region.

The first option is apparently attractive to a lot of people, judging by the popularity of descriptions of ancient peoples as “vanished” and so forth, but it’s actually quite rare for a group to literally die out entirely. It’s certainly possible that this is what happened to the Fremont, especially given the lack of continuity with later groups, but the number of people and large area involved make it implausible. That leaves us with either continuity between the Fremont and the ethnographic inhabitants of their region or a migration of the Fremont to somewhere else.

The idea that the Fremont might have developed into the hunter-gatherers known ethnographically in the eastern Great Basin and northern Colorado Plateau has a respectable history in the literature; as I noted in the previous post, James Gunnerson proposed just this back in the 1960s. The distinction in material culture between the Fremont and the Numic-speaking groups that followed them (Shoshone and Ute) is quite marked, however, as Albert Schroeder pointed out at the time. Furthermore, Madsen noted in 1975 that the distinctive Numic pottery is associated with the very different Fremont pottery at several well-dated sites in the region, suggesting that the two groups were distinct but contemporaneous. It is certainly possible that some of the Fremont assimilated into Numic society in some areas, or that the two merged in various combinations, and Madsen and Simms suggest that some such merging may have occurred in the Great Salt Lake area, though it’s not clear from their discussion whether they see the immigrant groups that merged with the Fremont as specifically Numic, as they propose a hiatus between this merged society and the ethnographically known culture of the region, perhaps due to the spread of European disease in the contact era. The whole issue of the Numic groups and how they got to where they are today is important in understanding the prehistory of these areas, but it is a big, complicated issue and I’ll address it more fully in a subsequent post. The material culture differences are significant enough that it seems unlikely that assimilation in place is the answer to the question of what happened to the Fremont in general.

As an alternative to seeing the Fremont as turning into the modern Numic groups that occupy the same areas, they may have migrated elsewhere. But where? Another theory noted in my previous post is that proposed by Melvin Aikens in the 1960s that the Fremont originally came from the Plains and ultimately migrated back there to become one or more of the ethnographically known Plains groups, probably Athabascan-speaking (i.e., Apache and/or Navajo). He based this theory on some suggestive parallels in material culture between the Fremont and Plains groups, especially the later Dismal River culture, generally thought to be associated with the Athabascan Na’isha. There certainly do seem to be some Plains-like traits in Fremont culture, including an emphasis on bison hunting, use of the shield-bearing warrior rock art motif, wearing of moccasins rather than sandals, etc. It’s not clear, however, whether these result from actual migrations of people from the Plains to the Fremont area or vice versa (and the two migrations Aikens posits would not be necessary in any case to explain the similarities). Aikens also used some physical anthropological evidence from skull morphology to support his theory, but the usefulness of the type of data he used was disputed even at the time, and it is not taken seriously at all now. Furthermore, more recent physical anthropological research using DNA analysis suggests strongly that there is no genetic connection between the Fremont and modern Athabascans or other Plains groups. In a sample of remains from the Great Salt Lake area the most common mitochondrial haplotype among Athabascans was not present at all, which is quite striking since it is quite common among Native American groups in general. It is of course possible, even likely, that this sample was not representative of Fremont groups in general, but of all the Fremont sub-areas the Great Salt Lake is the closest to the Plains both geographically and culturally, so if there’s no evidence of a genetic connection to the Plains from there it’s very unlikely that one will be found anywhere else. The same study found no clear evidence for a connection to the modern Numic groups either. This DNA stuff is another interesting, complicated issue that deserves its own post, but for now the upshot of this is that the Plains traits seen among the Fremont probably result from contact and cultural diffusion rather than migration in either direction, and the fate of the Fremont remains mysterious.

So if they didn’t go east onto the Plains, where did the Fremont go? The next obvious option is that they went south and joined the Pueblo groups with which they had many cultural similarities. This is another idea that has been proposed by some archaeologists, and it also appears to have support from oral traditions. David Pendergast and Clement Meighan published a paper in 1959 reporting that during their excavations of a site in southwestern Utah that would today be considered Parowan Fremont (though Pendergast and Meighan called them “Puebloid”) local Paiutes (a Numic group) told them some things about the people who had inhabited the site that they considered surprisingly accurate given the archaeological evidence. The Paiutes referred to the Fremont by the term Mukwitch, which is also the Paiute term for the Hopis, and reported that they had moved south and joined the Hopis when they left Utah. While their comments on the lifestyle of the Mukwitch and the reasons they had left were rather inconsistent, the consistency of the accounts of where they went is striking. The Paiutes also said that the Mukwitch were quite different from the Paiutes but had lived peacefully alongside them, which is noteworthy in light of the Numic pottery found in association with Fremont pottery mentioned above.

This paper has not been taken very seriously by archaeologists, and in fact I have not seen it cited at all in other Fremont literature. The only mentions of it I have seen, in fact, have been in cautions about the problems with taking oral traditions seriously, presumably because of the inconsistencies in the accounts. The accounts certainly are inconsistent on certain points, but consistent on others, and I think this paper deserves more attention from archaeologists wondering what happened to the Fremont. It’s certainly plausible that they moved south to join the Anasazi, and the Hopi are the most likely of the modern Pueblos for them to have ended up at for straightforward geographical reasons. A look at some of the recorded Hopi clan traditions with this in mind would likely be interesting. It’s unlikely that all of the Fremont moved south to join the Hopis, but it’s plausible that at least some did. Others may have stayed in place and been assimilated into the Numic groups spreading across the region, and still others might have died out entirely due to drought, warfare, or other factors.

So in some sense we’re back where we started, with no clear answer. But in other senses we do have some answers, at least in ruling out some options: The Fremont don’t seem to have either come from or gone to the Plains, and the Apaches are probably not their descendants. Wherever they did go (or stay), they changed their material culture rapidly and completely to assimilate into other groups, whether Numic or Hopi. This sort of rapid and complete assimilation is actually not as implausible as it seems; there are other examples of it in the prehistoric Southwest, and it must have happened quite a lot if the archaeological record is to be reconciled with the ethnographic one. All this suggests above all that the late prehistoric period, from AD 1300 on, was a time of immense change in the Greater Southwest, which makes it very difficult to figure out what was going on before that. Difficult, but not necessarily impossible. There are some ways to see through the haze.
Aikens, C. (1967). Plains Relationships of the Fremont Culture: A Hypothesis American Antiquity, 32 (2) DOI: 10.2307/277904

Armelagos, G. (1968). Aikens’ Fremont Hypothesis and Use of Skeletal Material in Archaeological Interpretation American Antiquity, 33 (3) DOI: 10.2307/278710

Gunnerson, J. (1962). Plateau Shoshonean Prehistory: A Suggested Reconstruction American Antiquity, 28 (1) DOI: 10.2307/278076

Madsen, D. (1975). Dating Paiute-Shoshoni Expansion in the Great Basin American Antiquity, 40 (1) DOI: 10.2307/279271

Madsen, D., & Simms, S. (1998). The Fremont Complex: A Behavioral Perspective Journal of World Prehistory, 12 (3), 255-336 DOI: 10.1023/A:1022322619699

Parr RL, Carlyle SW, & O’Rourke DH (1996). Ancient DNA analysis of Fremont Amerindians of the Great Salt Lake Wetlands. American journal of physical anthropology, 99 (4), 507-18 PMID: 8779335

Pendergast, D., & Meighan, C. (1959). Folk Traditions as Historical Fact: A Paiute Example The Journal of American Folklore, 72 (284) DOI: 10.2307/538475

Schroeder, A. (1963). Comment on Gunnerson’s “Plateau Shoshonean Prehistory” American Antiquity, 28 (4) DOI: 10.2307/278572

Read Full Post »

Basketmaker Pithouse, Mesa Verde

The Basketmaker III period (ca. AD 500 to 750) is a very important time for understanding the prehistoric Southwest.  Maize agriculture had been introduced earlier, although exactly how early is still a matter of debate, and it was definitely well-established by the immediately preceding Basketmaker II period, but Basketmaker III saw the introduction of beans, pottery, and the bow and arrow, all of which led to major changes in the lifestyles of local agriculturists.  Residence was in pithouses, which are clearly ancestral in form (and probably in function) to the “kivas” of later sites, and while these are usually found isolated or in very small groups, there are a few known examples of large “villages” containing dozens of pithouses.  The processes that led to the formation of these sites, as well as their relationships to the more common isolated sites, are very poorly understood, but it seems pretty clear that residential aggregation in certain locations during this period set the stage for the later formation of large villages during the succeeding Pueblo I period and afterward.

Two of the largest and best-known Basketmaker III villages are in Chaco Canyon.  The better-known of these, by far, is called Shabik’eschee Village, and it is located on the lowest terrace of a finger of Chacra Mesa at the east end of the current Chaco Culture National Historical Park.  Shabik’eschee was excavated by Frank H. H. Roberts in the 1920s as part of the Smithsonian/National Geographic project led by Neil Judd.  The main focus of the project was the excavation of Pueblo Bonito, but Judd had other members of the team, including Roberts, excavate several other sites in and around the canyon as well.  Roberts published his results in 1929, and this publication has been enormously influential in shaping subsequent interpretations of Basketmaker III villages and the period as a whole.

Looking South from Peñasco Blanco toward 29SJ423

The Chaco Project in the 1970s did some additional work at Shabik’eschee, as well as at the other Basketmaker Village in the canyon.  This site, known as 29SJ423, is just south of Peñasco Blanco at the far west end of the canyon, near the confluence of the Chaco and Escavada Washes.  It is situated in a similar location to Shabik’eschee, on a lower terrace of West Mesa (but above Peñasco Blanco, which is on the lowest terrace).  Tom Windes excavated a small portion of 29SJ423 in 1975, but he and other Chaco Project personnel soon came to the conclusion that additional excavation there would not be worth the considerable effort involved.  The collections from this excavation are important, however, since they were acquired using more careful, modern methods than Roberts’s.  Similarly, a very small amount of additional excavation at Shabik’eschee in 1973 has provided important supplemental information with which to evaluate Roberts’s interpretations.

Windes and Chip Wills published an article in 1989 looking back at Roberts’s interpretations at Shabik’eschee in the light of the additional knowledge gained by the Chaco Project excavations.  They concluded that some of Roberts’s ideas, such as his proposal that the site had two discrete periods of occupation separated by a hiatus during which it was abandoned, are likely untenable, and they also concluded that the site was considerably larger than Roberts thought.  They agreed with Roberts that some of the pithouses had been abandoned and their materials were used in subsequent construction, but they saw this as more of an ongoing process related to the short use-life of pithouses and the demands of demographic processes rather than a discrete series of two occupations.  They also saw more spatial patterning in the layout of pithouses within the site than Roberts did, suggesting that the pithouses grouped into what might be family residence units, although they were quite tentative in this finding and did not use these groups as units for any subsequent analysis.

Pinyon Trees, Pipe Spring National Monument

Wills and Windes also posited a novel interpretation for the site as a whole.  Rather than seeing it as a permanent agricultural village, they saw it as a site of occasional gatherings of more mobile families practicing a “mixed” subsistence strategy of small-scale agriculture along with hunting and gathering.  In their interpretation, a small number of families inhabited Shabik’eschee permanently, while others joined them periodically to take advantage of the site’s proximity to piñon woodlands in years with bountiful piñon-nut harvests.  They based this theory on the presence of two types of storage facilities at the site: household-level storage in the antechambers associated with some but not all of the pithouses (presumably the residences of permanent residents) and community-level storage bins scattered around the site.  The idea is that occasional surpluses of corn or whatever would be stored in the bins, and the people who lived at the site permanently watched over it and protected it.  Whenever there was a plentiful crop of piñon nuts, which happens at irregular intervals in the fall, people who lived the rest of the time in scattered locations throughout the area would congregate at Shabik’eschee to take advantage of this and stay for the winter.  If conditions in the spring were good for planting, people might stay longer and plant their crops in the area, but if not they would move on to more attractive planting locations.  Other pithouse villages, such as 29SJ423, would presumably have served similar purposes, allowing periodic aggregation to take advantage of various localized resources.

This is an interesting theory, but it’s based on exceptionally thin evidence.  Wills and Windes even concede that they are spinning this whole story purely from the nature of the storage facilities at the site, and they note that there are other ways to interpret the communal bins in particular.  Instead of protecting food stores during periods of reduced occupation, they may just have functioned to protect them in general.  The shape of the bins makes it more difficult to access their contents, which Wills and Windes interpret as evidence for a sort of semi-caching, but it would also just provide better protection from the elements, vermin, etc. for the contents.  Basically, there’s just no reason from the available evidence to buy the Wills and Windes theory.

"Pithouse Life" Sign at Mesa Verde

Indeed, the assumptions behind this theory seem problematic to me.  The ethnographic comparisons Wills and Windes use to support it are mostly from hunter-gatherer societies, and indeed their model seems to imply that the residents of Shabik’eschee were basically hunter-gatherers who did some farming on the side.  Such societies exist, and may well have existed at certain times in the ancient Southwest (such as the late Archaic), but recent studies have shown with increasing certainty that heavy dependence on agriculture was widespread already in the Basketmaker II period.  Wills and Windes seem to see the Basketmaker III inhabitants of the Chaco area as just beginning to experiment with adding agriculture to a hunter-gatherer lifeway, but it’s much more likely that they were full-time agriculturalists and had been for centuries.  They did of course still do some hunting and gathering, as their Pueblo descendants have continued to do up to the present day, but while this may in some sense qualify as a “mixed” economy that shouldn’t obscure the important fact that Pueblo societies have been overwhelmingly farming-based societies since well before the occupation of Shabik’eschee.

I think this interpretation, and others like it which were popular in Southwestern archaeology in the 1980s, results in part from the enormous influence of Lewis Binford on the development of processual archaeology.  Binford’s personal research and expertise were largely on hunter-gatherer societies, and the guidelines he set forth for “archaeology as anthropology” that were eagerly followed by young “New Archaeologists” were heavily influenced by that background.  Wills and Windes cite Binford several times in this article.

Excavating the Lift Station Site in the Chaco Maintenance Yard

Be that as it may, this is an important article just in providing an updated take on the facts about Shabik’eschee, which as Wills and Windes note has been very important in the interpretation of ancient societies generally.  It contains relatively little information about 29SJ423, but it does briefly discuss this site as a comparison.  It says even less about the much more numerous isolated Basketmaker III sites in the canyon, but it notes that Chaco Project surveys identified at least 163 pithouse sites from this period.  One that they didn’t find, because it was deeply buried under the ground, was later found by the park in the course of trying to build a lift station for the septic system.  This site, informally known as the Lift Station Site, is a Basketmaker III pithouse that was excavated while I was working at Chaco.  One of the more interesting things it revealed was an apparent location for pottery manufacture.

One of the major problems with trying to understand the Basketmaker III period at Chaco is precisely that the site are typically deeply buried, so it’s hard to even know how many of them there are.  It’s clear that this was a period of significant population in the canyon, but it’s hard to tell how many sites were occupied simultaneously.  This problem is exacerbated by the difficulty of dating many of the sites.  Tree-ring dates are often hard to obtain from the scarce wood found at excavated sites, and Shabik’eschee is particularly poorly dated.  The few tree-ring dates available seem to suggest it was occupied at some point after the mid-500s, but there are no cutting dates so any greater precision is impossible.  29SJ423 did produce two cutting dates, at 550 and 557, so it seems the two villages were most likely contemporaneous.  The isolated sites are even harder to date, of course, but the Lift Station Site produced corn that was radiocarbon dated.  I don’t know the dates that resulted, but I did hear that they were earlier than was expected based on the pottery types found.

Whole Pot from the Lift Station Site

The size of the Basketmaker III occupation at Chaco, and particularly the presence of the two large villages, has important implications for understanding the subsequent history of the canyon that I think are just beginning to be realized.  The local population seems to have declined during the subsequent Pueblo I period (ca. AD 750 to 900), when people seem to have begun to move in large numbers to higher elevations where they formed some really large villages.  However, it’s not clear that Chaco was completely abandoned during this period, and recent improvements in dating the early great houses in the canyon have shown that some of them, especially Pueblo Bonito, go back further than was once thought.  Pueblo Bonito is now known to have been begun no later than 860, and the earliest part of it may date much earlier, possibly to 800 or even before.  This means that the gap between the Basketmaker III villages and the earliest great houses suddenly looks a lot smaller, and may disappear entirely.  There are pithouses under the plaza at Pueblo Bonito that may date to very early Pueblo I or even Basketmaker III, and there is a small Pueblo I occupation at Shabik’eschee that dates as late as 750.  This suggests that these two iconic sites in Chacoan archaeology, generally interpreted in very different ways, may actually overlap in occupation.  This would require some serious modifications of the ways the origins of the Chaco system are often interpreted.

Chaco had been an important place for a very long time when it started to become a major regional center around AD 1040.  It’s looking increasingly plausible, though by no means certain, that it had been continuously occupied for 500 years at that point, and even if there was a brief gap between the Basketmaker III villages and the first Pueblo I great houses it is very unlikely that is was long enough for people to have forgotten about Chaco and what had happened there.  Even if many of the people who built and/or occupied the early great houses in the 800s hadn’t been born at Chaco, they probably knew it was there long before they made it their home.
Wills, W., & Windes, T. (1989). Evidence for Population Aggregation and Dispersal during the Basketmaker III Period in Chaco Canyon, New Mexico American Antiquity, 54 (2) DOI: 10.2307/281711

Read Full Post »

Petroglyphs of Quadrupeds at Atlatl Rock, Valley of Fire State Park, Nevada

In 1978 H. Martin Wobst of the University of Massachusetts at Amherst published a short article in American Antiquity entitled “The Archaeo-Ethnology of Hunter-Gatherers or the Tyranny of the Ethnographic Record in Archaeology.”  Despite the evocative title, the article itself is a highly theoretical argument about the proper relationship between archaeology and ethnography that is unlikely to be of much interest outside those fields.  Basically, Wobst argues that the archaeology of hunter-gatherer societies is overly dependent on concepts drawn from ethnographic study of modern hunter-gatherer societies, even though that ethnographic research has inherent limitations in what it can observe about those societies and is further limited by the specific priorities of the scholars who conduct it.  He therefore says that archaeologists should play a larger role in developing theoretical approaches to these societies based on archaeological data, which has its own limitations but is nevertheless better suited to studying certain topics, such as large-scale regional interaction, than is ethnography.  From the perspective of archaeology, Wobst’s article is clearly situated in the processualist tradition, with its emphasis on using archaeological evidence to reconstruct social behavior and contribute to general anthropological theory.

The most interesting part of Wobst’s article, however, is the acknowledgments at the end, which begin with this remarkable dedication:

I would like to dedicate this paper to Provost Dr. Paul Puryear, without whose failing support of Social Sciences at the University of Massachusetts, Amherst, I would have been done much earlier.

It seems that Paul Puryear was indeed in some sort of administrative position at UMass at the time, but beyond that I have no idea what Wobst is talking about here.  Still, it’s a welcome change from the anodyne expressions of gratitude that usually dominate these parts of papers.
Wobst, H. (1978). The Archaeo-Ethnology of Hunter-Gatherers or the Tyranny of the Ethnographic Record in Archaeology American Antiquity, 43 (2) DOI: 10.2307/279256

Read Full Post »

Novitski Hall, University of New Mexico School of Dentistry, Albuquerque, New Mexico

I’ve recently been discussing stable isotope analysis as a way to directly determine dietary practices from skeletal evidence, and that is certainly a powerful tool in learning about past societies, but there are some drawbacks to it.  Like all complicated laboratory procedures, it’s expensive, and it has the additional problem of being destructive.  If it’s done right, it only requires a small amount of bone, but it does involve destroying that bone in the course of analysis, which puts it in tension with recent political trends away from invasive and destructive types of research.  It is therefore good to have additional ways of evaluating dietary practices, despite the enormous potential of isotope studies.

One such line of evidence is much more low-tech, and quite simple, as well as being non-destructive.  It starts from the widespread recognition that different types of diets have different and readily detectable effects on teeth.  Specifically, diets high in carbohydrates tend to result in significantly more dental caries (cavities) than diets higher in proteins and fats.  A variety of factors are involved in determining the rate of caries in a given individual, including the form of the teeth and the presence of caries-resistant minerals such as fluoride in the environment, but diet is one important factor and the one that can most easily account for differences between populations and societies in the rate of caries.  The way this works is that diets rich in carbohydrates, especially refined carbohydrates such as ground meal or flour which stick to the teeth more easily, result in buildup of plaque that certain bacteria in the mouth feed on.  Those bacteria then release waste products including lactic acid, which eats away at tooth enamel can causes caries.  Fats and proteins also cause plaque buildup, but the bacteria don’t feed on this plaque, and it tends to be less acidic and therefore less conducive to caries formation.

The implication, then, is that societies that are dependent on agriculture, in which people eat large amounts of carbohydrates, will show much higher rates of caries than hunting and gathering societies in which people eat more fats and proteins.  This has indeed been confirmed by observation of caries rates in numerous contemporary and prehistoric populations.  There are some cases in which hunting and gathering populations can have relatively high rates of caries, such as a heavy dependence on gathered resources that are high in carbohydrates, such as acorns and pine nuts, but in general the difference between foragers and farmers is quite clear and can be used to determine to what extent a given prehistoric population depended on agriculture just by looking at their teeth.

Durango Herald Offices, Durango, Colorado

One relatively recent study looking at this issue in the context of the Southwest is by Karen Gust Schollmeyer and Christy Turner.  Turner is best known in the Southwest for his controversial ideas about cannibalism, but he has also had a longstanding interest in dental studies which I’ve discussed before.  This paper looks specifically at the dispute over the timing of agricultural dependence in the Southwest and whether it arose at the same time as the “Pithouse to Pueblo Transition” between the Basketmaker III and Pueblo I periods, generally dated to around 750 AD.  The hypothesis Schollmeyer and Turner tested is that the Basketmakers had a mixed farming and foraging economy with relatively little dependence on agriculture but that later Pueblo populations were heavily dependent on farming.  This predicts that Pueblo populations should have substantially more caries than Basketmakers.  The sample they used to test this was a large collection of human remains from various sites in southwestern Colorado, mostly in the Durango area and the La Plata Valley, in the collections of the Harvard Peabody Museum and the American Museum of Natural History.  Since these were mostly excavated before 1930 and information on their exact origin and time period is often vague, Schollmeyer and Turner grouped them into just two groups, Basketmaker and Post-Basketmaker, and ran a series of statistical tests on the number and placement of caries in each group.

They found that there was basically no difference.  Both groups had very high rates of caries, whether measured as total number of carious teeth, total number of carious teeth and teeth that fell out during life (which often results from caries), or total number of individuals with carious teeth.  Statistically most of these were indistinguishable from each other, although it’s important to note that this can’t be considered a truly random sample and statistics drawn from it shouldn’t be taken too literally.  Comparisons of caries on specific types of teeth and on different parts of teeth also showed high rates in both groups.  There were no significant differences regarding types of teeth.  There was a significant but weak difference in parts of teeth, with the later individuals having more caries on the parts of the teeth facing other teeth rather than facing outward.  There was also a similarly significant but weak difference in the number of individuals with caries, which was lower in the later group, implying more caries per person with caries than in the Basketmaker period, since the overall number of carious teeth was not different.  I don’t think too much should be made of either of these differences, given the nonrandom nature of the sample, and Schollmeyer and Turner are properly cautious in interpreting them.  They do make the interesting suggestion, however, that if there is anything to these differences they may imply differences in processing of maize over time, with later groups processing it more efficiently and intensively, using more effective tools, perhaps in response to increased populations.  They further suggest that this may be behind some of the dispute in the literature over the timing of dependence on agriculture, since some of the evidence put forth as showing a late date, coincident with the Pithouse to Pueblo Transition, is based on the use of larger and more efficient grinding tools.  This is all pretty speculative, and I don’t think it’s really any more likely than the alternative explanation that the significant differences are just statistical noise, but it’s an interesting thought.

Railroad Bridge, Durango, Colorado

Overall, this evidence supports the other recent studies showing that Southwestern populations seem to have been dependent on agriculture by at least as early as the Basketmaker II period.  It’s therefore not exactly groundbreaking, but it is useful to have as many lines of evidence as possible brought to bear on important questions like this, especially when most of them seem to point in the same direction.
Schollmeyer, K., & II, C. (2004). Dental Caries, Prehistoric Diet, and the Pithouse-to-Pueblo Transition in Southwestern Colorado American Antiquity, 69 (3) DOI: 10.2307/4128407

Read Full Post »

Looking East from the Pueblo Alto Trail

In my earlier post about Stephen Hall‘s recent paper reporting on maize pollen at Chaco Canyon dating as early as 2500 BC, I said briefly that this really shouldn’t be surprising to anyone who’s been following this kind of research closely, and also that I would discuss the context for it later.  Basically, the context is that there has been a considerable amount of evidence accumulating in the past twenty years or so firmly placing maize in the Southwest much earlier than most archaeologists had previously thought.  Much of this evidence has been gathered in a recent paper criticizing Jane Hill‘s arguments for the spread of maize having occurred as part of a migration of Uto-Aztecan speakers from Mesoamerica.  The earliest directly dated maize seems to be from the Los Pozos site in the Tucson Basin, with a date range of 2860 to 2470 BC (this represents the calibrated 95% confidence interval, as do all subsequent radiocarbon dates in this post unless noted otherwise).  According to the text of the paper, however, this date has been questioned, and the earliest uncontroversial direct date on maize comes from the aptly named Old Corn site near Zuni, with a range of 2460 to 2060 BC.  Note that the earliest date Hall found at Chaco (2567 to 2332 BC) has an upper bound slightly earlier than this, although the ranges overlap considerably.  Similarly early dates occur at some Tucson Basin sites and at Three Fir Shelter in northern Arizona, although the Three Fir Shelter date has a particularly large confidence interval.  The considerable geographic extent of these early maize dates implies that the initial introduction of maize to the Southwest may well have been quite a bit earlier than the earliest archaeological evidence for it.  It also suggests that at least in most areas, with some possible exceptions such as the Tucson Basin, maize was initially integrated into a hunter-gatherer lifestyle without causing major changes to either the subsistence system or the social structure of the groups adopting it.  This challenges many of the traditional assumptions about the effect of the introduction of agriculture on hunter-gatherer societies, including the idea that farming necessarily leads to an immediate shift to settled village life.  There is little to no evidence in most areas of any such shift for thousands of years after the earliest evidence for maize agriculture, suggesting that the impetus for such a transition (which did indeed happen in almost all parts of the Southwest eventually) must be sought elsewhere.

Una Vida from Petroglyph Area

Okay, so there’s plenty of evidence for agriculture as early as the pollen dates at Chaco in many parts of the Southwest, but Chaco ain’t Tucson.  The high, dry, harsh environment of the central San Juan Basin seems particularly unsuitable for agriculture compared to a lot of the other areas with early evidence for maize.  So surely this is a surprising finding at least for the local area, right?

Actually, no.  Evidence for the early use of maize in the general area of Chaco, though until now not within the canyon itself, has actually been out there for over twenty years, although it doesn’t seem to have gotten much attention in the subsequent literature.  It is described by Alan Simmons in an article from 1986, reporting on research done initially as a cultural resources management (CRM) project in connection with plans for coal mining in the area directly to the east of Chaco.  This project, known as ADAPT I, found many typical Archaic sites in the area, some of which were excavated as mitigation activities because of the expected impacts of mining.  (I believe this mining operation never actually happened, but I don’t know any of the details.)  These Archaic sites were typical for the San Juan Basin, which is to say that they were mostly lithic scatters with few diagnostic artifacts or datable materials, but a few of them ended up being very surprising in having more under the surface than was apparent at first, including possible hearths or ovens, some of which contained maize pollen associated with charcoal that was radiocarbon dated to the Archaic period and one of which even contained maize macrofossils that were directly dated to the late Archaic.  (The dates Simmons relates are apparently uncalibrated and are therefore not directly comparable to the calibrated date ranges given above, but they are comparable to the uncalibrated dates given by Hall for the pollen samples, including at the early end.)

Looking North from New Alto

This very unexpected result led to a new project, known as the Chaco Shelters Project, to try to get more data on the Archaic period in the Chaco area through intensive study of the type of site most likely to contain well-preserved material: rockshelters.  Two shelters were excavated, both outside of the park boundaries, one to the east (Sheep Camp Shelter) and one to the west (Ashislepah Shelter).  Neither provided much in the way of artifacts, which was something of a disappointment, but both showed evidence of Archaic use, and Sheep Camp Shelter contained macrobotanical remains of both maize and squash.  Neither contained maize (or squash) pollen.  Both maize and squash remains were directly dated to the late Archaic.  The squash results were more important, as the resulting dates were at the time the earliest evidence for the presence of squash in the Southwest.  Despite the presence of these direct dates, the earliest dates Simmons had were on charcoal associated with the maize pollen at the ADAPT I sites, and these could potentially be questioned given the known problems with this kind of indirect dating.  Now that Hall has come up with direct dates on the pollen itself which cover roughly the same timespan, however, Simmons’s indirect evidence looks more convincing than it may have at the time.

At the time Simmons was writing, there was much less direct evidence of Archaic agriculture in the Southwest than there is now, and the few reported early dates were problematic and disputed.  He therefore has a substantial and quite useful discussion of the general issue in addition to reporting his specific data.  He clearly comes down on the side advocating a gradual shift from hunting and gathering to agriculture, with maize being initially adopted by hunter-gatherers as part of a preexisting subsistence system and only becoming the primary means of subsistence considerably later.  This view was generally associated with arguments for in situ development of Anasazi culture out of the local Archaic tradition, and was opposed by those who preferred to see agriculture as being introduced later and all at once, perhaps as part of a migration of agriculturalists from the south (a position now maintained by Jane Hill, among some others).  With regard to the Chaco area specifically, Simmons accounts for the differences between the open ADAPT I sites (with maize pollen but little macrobotanical evidence) and Sheep Camp Shelter (maize and squash macrofossils but no pollen) by proposing that maize was initially integrated into a subsistence system involving seasonal mobility, with spring and summer being spent in the ADAPT I area, where the many sand dunes would have fostered the growth of wild plants, and winter being spent in the more protected rockshelters in the canyons to the west.  When maize and squash were added to the seasonal round, they would have been planted on or near the dunes in the spring, harvested in the fall, and taken back to the shelters to provide durable, storable food for the harsh and largely barren winter.  He admits that the shelter evidence, especially, is a bit weak, as it doesn’t really support the idea of extensive winter use of the shelters, but there were apparently many Archaic sites near the shelters that may provide more support for occupation of the general area during the winter.  He therefore concludes that the model he proposes is generally consistent with the data although more research is needed to confirm it.  He mentions Atlatl Cave, in the park and excavated by the Chaco Project, as one possible additional data point in favor of his model.

Pueblo Bonito from Peñasco Blanco

Hall’s new pollen data from within the canyon provides some support for Simmons’s model, especially in backing up the indirect dates with direct ones, but it also suggests some possible modifications to it.  It clearly seems to indicate that the canyon itself as well as the dunes to the east was used to grow corn during the Archaic.  This casts some doubt on Simmons’s proposal of seasonal mobility, at least in the way he frames it.  Growing corn in the canyon would imply occupation of it, perhaps in rockshelters such as Atlatl Cave, during the spring and summer in addition to (or instead of?) during the winter.  That is, Simmons’s seasonal mobility model could perhaps be “condensed” into a similar model of subsistence activities throughout the year, but with year-round residential occupation in the canyon.  People would presumably have still traveled around to gather resources in different areas at different times of year, but this would be “logistical” as opposed to “residential” mobility, with hunting and gathering parties setting out from a more permanent base camp solely to collect resources as opposed to moving the whole group to the location of the resources.  It is of course also possible that Simmons’s model should be modified in some other way, keeping the seasonal mobility but changing the role of the canyon in it, but it’s hard to see where else in the area would have been more suitable for winter occupation.

One of Simmons’s suggestions, that the advent of corn agriculture may have made the Chaco area an easier place to live by providing a fairly reliable food source in a place with limited and unreliable wild food resources, is particularly interesting.  The San Juan Basin is well-known for its considerable Archaic population, which has generally been interpreted as indicating seasonal mobility and specialized adaptations to the region’s sparse resources, but what if that population was actually supported in part by incipient maize (and squash) agriculture?  People often marvel at the apparent barrenness of the Chaco environment and the oddity of the idea that anyone would try to farm there, but this is another way to look at the harshness of the environment.  To put it differently, farming in Chaco is certainly difficult and risky, but can you imagine trying to live there without farming?
Merrill, W., Hard, R., Mabry, J., Fritz, G., Adams, K., Roney, J., & MacWilliams, A. (2009). The diffusion of maize to the southwestern United States and its impact Proceedings of the National Academy of Sciences, 106 (50), 21019-21026 DOI: 10.1073/pnas.0906075106

Simmons, A. (1986). New Evidence for the Early Use of Cultigens in the American Southwest American Antiquity, 51 (1) DOI: 10.2307/280395

Read Full Post »

Older Posts »