The Numic Spread

Sign Describing Paiute Brush Shelters, Pipe Spring National Monument

The Great Basin and northern Colorado Plateau were occupied at the time of European Contact (generally between the late eighteenth and early nineteenth century for this region) by a variety of relatively small groups of hunter-gatherers, all of whom spoke closely related languages belonging to the Uto-Aztecan language family. By the early twentieth century these groups had become of considerable interest to anthropologists due to the harshness of their physical environment and the apparent simplicity of their social structure.

The most influential ethnographic studies of these groups were those conducted by Julian Steward among the Western Shoshonis in the 1930s. Steward developed a model of Western Shoshoni society in the Great Basin that emphasized the constraints imposed by this harsh environment and the consequent need for small group size and frequent movement in search of subsistence resources. Steward focused heavily on the political structure of Western Shoshoni society, which he divided into two fairly different forms of social organization: “village” and “band.” “Village” groups he characterized as small groups of people who generally aggregated in small villages during the winter at favored locations and dispersed into even smaller family groups in the other seasons to gather scattered resources. Each village group roamed through a territory which was not sharply defined but typically consisted roughly of a single valley, but the groups were fluid and people and families frequently moved from one to another for a variety of reasons such as resource availability and kin or trading ties.

Los Angeles Department of Water and Power Office, Lone Pine, California

“Band” groups, on the other hand, were larger and generally found in areas of greater concentration of resources than is typical in the arid Basin. The Owens Valley Paiutes were the preeminent example of this type of organization, as Owens Valley was teeming with resources and was divided into a number of band territories which were carefully guarded and defended by their resident bands. (This may seem incongruous to anyone who has been to Owens Valley recently and seen how dry and desolate it looks, but the story of how it got that way is of course a famous one.) Some of the Shoshoni groups in nearby valleys had similar social organization to the Owens Valley Paiute, but Steward considered band organization in most of the rest of the Basin to be a post-Contact phenomenon associated with the introduction of the horse and extensive cultural influence from the Plains buffalo-hunting cultures further east.

Steward considered all the cultures he documented in the Great Basin to be of relatively recent origin and the result of influences from many directions, some of them post-Contact (most obviously the case with horses). Archaeologists, however, were very impressed with the way the social models he outlined, especially the “village” one, evidenced a very close relationship between the resources of the land and the social structure of the people living in it. This seemed like a description of society at its most basic in a harsh environment, which might be a reasonable model for prehistoric societies who inhabited similarly harsh locations, including the Great Basin itself. As more discoveries came to be made extending knowledge of North American prehistory in what is now known as the Archaic period, especially in the deserts of the Southwest, archaeologists began to interpret them in light of Steward’s ethnographic data. In the 1950s this approach was codified by Jesse Jennings of the University of Utah into the concept of a “Desert Culture” that looked basically like Steward’s “village” societies and existed throughout the desert Southwest from the end of the Paleoindian period until the adoption of agriculture in some areas and until Contact in others (the latter being the societies Steward studied).

Paiute Brush Shelters, Pipe Spring National Monument

The idea of a relatively unchanging culture limited by ecological conditions and lasting for thousands of years was especially attractive to the “New Archaeologists” who emerged on the scene beginning in the 1960s and were very concerned with deriving general conclusions about social processes from the archaeological record using the scientific method (which led to them eventually being known as “Processualists”). Unlike in many areas where the cultural record was clearly very complicated and many different societies with different economies had followed each other over the millennia, here was a case where a single social model had endured for almost the entirety of the archaeological record, in some areas right down to the ethnographic present. Environmental and archaeological data could be gathered and compared to each other to test various hypotheses derived from explicit theories based on the baseline data established by Steward and Jennings. Most of the other variables that tended to confound such theory formation and hypothesis testing in other areas were held constant here by the harsh environment and the resultingly simple and stable societies that were adapted to it.

As a result, the Great Basin by the 1970s became the site of several major archaeological research projects sponsored by prominent institutions seeking to capitalize on this opportunity. They collected an enormous amount of useful data that shed important light on both general anthropological questions and the culture history of the Great Basin itself. The region was a triumph of processual archaeology and a showcase of its potential.

Datura Sign, Pipe Spring National Monument

There was a nagging problem with all this, however. The archaeology and ethnography (which were generally treated as basically the same) of the region painted a picture of remarkable cultural stability, but the linguistic evidence pointed in a very different direction. Explaining how requires some backtracking and explanation of the linguistic situation at Contact.

As I said earlier, it was apparent to anthropologists by the late nineteenth century that the languages spoken by the Basin groups were closely related to each other and more distantly related to many other languages of western North America (including Mexico) within what came to be known as the Uto-Aztecan language family. The internal structure of this language family was much harder to establish than its existence, however, and there’s actually still no consensus among linguists about how it should be divided. In 1925 Alfred Kroeber of the University of California proposed a division whereby most of the languages north of the US-Mexican border constituted a “Shoshonean” family that was subdivided into four branches. The branch consisted of the groups occupying the Great Basin and adjacent portions of the Colorado Plateau was known as “Plateau Shoshonean,” a label that stuck for decades afterward, and was divided into three branches itself. Steward in the 1930s kept Kroeber’s basic division of both the overall Uto-Aztecan family and the Plateau Shoshonean subfamily, although he gave the sub-branches different names. These three branches, with their general areas of occupation, are:

• Northern Paiute (Steward), Mono-Paviotso (Kroeber); from Owens Valley northward through western Nevada to southeastern Oregon
• Shoshoni (Steward), Shoshoni-Comanche (Kroeber); from Death Valley northeastward through eastern Nevada and northern Utah to western Wyoming, with the Comanche as an offshoot that migrated in the eighteenth century to the southern Plains
• Southern Paiute (Steward), Ute-Chemehuevi (Kroeber); from Panamint Valley and possibly southern Death Valley eastward through southern Nevada and southern Utah to western Colorado

The overall division of the linguistic groups and their general territories have not been controversial, and this three-part scheme continues to be the standard way to divide up these languages. Looking at a map of the territorial distributions, an interesting fan or wedge shape is very evident. Each of the three subdivisions extends from one or more isolated valleys in eastern California at the western edge of the Great Basin (the narrow end of the wedge) out across the Basin and, in some cases, beyond (the wide end of the wedge). Steward and Jennings didn’t have much to say about this distribution, but it would later become a crucial point of evidence in a very different interpretation of Basin prehistory that directly challenged the long-term and unchanging nature assumed in the Desert Culture framework.

Yucca Sign, Pipe Spring National Monument

In the 1950s a linguist named Sydney Lamb conducted extensive linguistic fieldwork among speakers of these languages and came up with much better data than Kroeber or Steward had been able to secure. He published an important paper in 1958 reporting on his resulting conclusions about the internal relationships of the languages and the implications for the prehistory of the region.

Lamb’s work confirmed Kroeber’s three-part division of Plateau Shoshonean, but undermined the notion of “Shoshonean” itself as a basic division of Uto-Aztecan. Instead he considered the “Shoshonean” subfamilies to be independent branches of Uto-Aztecan, and proposed new names for them to indicate this. “Plateau Shoshonean” thus became “Numic” after the word for “people” in the languages in question. In subsequent research the related term “Numa” has also become popular as a collective noun referring to speakers of these languages.

Entrance Sign, Death Valley National Park

Within Numic, Lamb found that each of the three subfamilies consisted of two languages, closely related to each other but not quite mutually intelligible and quite distinct from the languages in the other subfamilies. In each case one of these languages was spoken in the eastern California valleys at the southwestern end of the subfamily’s distribution and the other was spoken over the vast area to the north and/or east that made up the remainder of the distribution, with little variation over these huge areas.

Based on this distributional evidence, combined with some tentative glottochronological dates that Lamb prefaced with appropriate skepticism about the validity of glottochronology, Lamb concluded that the Numic languages had originated in the valleys of eastern California and had spread from there across the Great Basin quite recently, perhaps around 1000 years ago. Importantly, the subfamilies were apparently already distinct at this point, and their speakers seem to have moved in similar ways and directions but independently, which implied that there was some common force drawing them further into the Basin (or, perhaps, out of California). Lamb tentatively suggested that access to bison might have been part of the motivation for the migration, but without going into detail. The most important point, however, is that Lamb concluded that the linguistic uniformity of the Great Basin Numic languages suggests strongly that Numic speakers, including Steward’s famous Western Shoshoni whose culture was the basis for Jennings’s Desert Culture, were recent immigrants into most of the Basin, and not the surviving remnant of a widespread Desert Culture that had existed there for thousands of years. He acknowledged that this conclusion was in sharp contrast to the archaeological consensus, but put it out for discussion nevertheless.

Sign at Border of Ute Mountain Indian Reservation

Initially, at least, archaeologists didn’t buy it. They were quite confident of the validity of their Desert Culture model, and the subsequent rise of processual approaches only intensified the split between linguistic and archaeological interpretations of Great Basin prehistory. Not all linguists agreed with Lamb either, and various papers by both linguists and archaeologists in the succeeding decades proposed alternative explanations for the distribution of the Numic languages. Overall, though, most linguists came to be convinced by Lamb’s evidence that his interpretation was the most plausible, and by the 1980s even archaeologists began to be convinced.

Note that when I say “archaeologists” here I’m referring specifically to archaeologists who specialized in the Great Basin, especially those who focused on the western part of the Basin where Steward had done his work. Those archaeologists who studied the eastern Basin and the Colorado Plateau, many of whom were more associated with Southwestern archaeology, had much less trouble accepting the idea that the Numic-speakers were recent arrivals in the Basin, as they obviously were in the Plateau. The ethnographic literature on the Utes and Southern Paiutes contains various references to the remains of the Fremont associating them with the Hopis rather than with Numic-speakers, and Steward himself recorded a tradition among the Northern Paiutes that the area around Lovelock Cave had been inhabited by non-Paiutes fairly recently. Remember that Steward considered the cultures he studied to be relatively recent, which is consistent with a recent Numic spread and inconsistent with Jennings’s Desert Culture theory.

Owens Lake, California

The first major theory based on a recent Numic spread to be proposed by archaeologists was that of Robert Bettinger and Martin Baumhoff of UC Davis, who published an important paper in 1982 making their case. They argued that Lamb’s Numic spread could be explained through a processual model. Under this model the pre-Numic cultures of the Basin were said to be based heavily on the hunting of big game, especially bighorn sheep, while the Numic cultures were based on a more intense gathering of small seeds, a lower-ranked resource that was more effort to get and process but more reliable as a source of calories. Bettinger’s own fieldwork had been focused mostly on Owens Valley, which he concluded had been where the Numic speakers had developed this focus on seeds out of necessity given the density of resources and population (recall that Steward had also argued that this was an area of more elaborate cultures than most of the Basin, for the same reason). Bettinger and Baumhoff argued that population pressure stemming from the adoption of this strategy was the impetus for the Numic groups to begin to spread out into the rest of the Basin, where their more effective seed-based economic strategy allowed them to out-compete the pre-Numic groups, who were unable to adapt to a similar strategy fast enough to compete effectively because of societal inertia. Climatic changes that reduced the availability of game may have played a role as well. They supported this idea of a discontinuity by pointing to differences in rock art and artifacts between earlier and later periods in the Great Basin archaeological record, especially the increased presence of specialized seed-beating equipment in the later period, presumably Numic.

The Bettinger-Baumhoff hypothesis immediately aroused considerable controversy, and in the next few years many objections to it were raised, mostly by archaeologists but occasionally by linguists as well. Bettinger and Baumhoff responded to some of these objections in follow-up papers, and overall their arguments have sparked a serious and generally productive discourse on the prehistory of the Great Basin and how to reconcile the archaeological and linguistic evidence. Over time the general trend has been toward increasing evidence of a variety of types in favor of some sort of recent Numic spread, and more and more archaeologists have begun to accept the reality of it. DNA evidence demonstrating a major discontinuity between at least some pre-Numic human remains and modern Numic groups has added an important independent line of evidence for a Numic spread, and additional intensive research in Owens Valley has further clarified the archaeological picture there and given more context to cultural changes (such as the adoption of pottery) that may have played a role in the origins of the spread.

Kaibab Paiute Housing Development from Pipe Spring National Monument

So that’s the history of research into Numic prehistory in a nutshell. My take on it is that Lamb was clearly totally right that there was a Numic spread and that it was relatively recent (though his specific glottochronological dates are of course unreliable), and that Bettinger and Baumhoff may have been correct about its nature but that there remain some weak points in their theory. I think the archaeological reluctance to accept the idea of a Numic spread is due to a number of factors that have been problematic in the history of Americanist archaeology throughout the twentieth century but are particularly extreme in this case.

For one thing, there has long been a tradition of archaeologists projecting ethnographic data on post-Contact Native American groups uncritically back into the past. This was particularly common in the early twentieth century before it was widely accepted that the Americas had been occupied more than a few thousand years, and in that context it was at least understandable that Native cultures would have little time-depth. With the extension of the archaeological record further back in time and the development of more accurate and precise dating techniques, it became less justifiable to use ethnographic analogy and the Direct Historical Method so straightforwardly, but it has continued to some extent throughout the US, and the perceived harshness of the Great Basin environment and the relatively extensive ethnographic record there has made this tendency particularly pronounced there.

Badwater Basin, Death Valley National Park

The “New Archaeologists” of the 1960s and 1970s defined their approach explicitly in contrast to previous generations’ overreliance on specific ethnographic data and naive projection of it back into prehistory. In many parts of the US this meant a major shift, but again the specific characteristics of the Great Basin made the New Archaeological method look a lot like old-fashioned culture history. The apparent lack of change in the Basin’s archaeological record over millennia had meant that the culture history was interpreted as a story of stasis ending up with the ethnographic Numa, and this story of ahistoricality was easily translated into a story of consistent adaptations to a harsh and severely limiting environment. In both cases there was not actually any evidence strongly in favor of continuity of population (as opposed to adaptation), but that was a reasonable null hypothesis and, as often happens, over time it expanded from that to an unstated assumption. Bettinger and Baumhoff’s theory was presented very explicitly in the terms of processual archaeology but was nevertheless very controversial because of this assumption.

The generally ahistorical approach of the processualists is now less dominant in American archaeology than it was in 1982, and this is probably a factor in the increasing acceptance of a Numic spread among archaeologists. I find it a fascinating story both because it sheds light on the dynamic nature of prehistory and relationships between linguistic and cultural groups and because it illustrates important trends in the intellectual history of American archaeology in particularly vivid fashion. It’s also a story that seems to be more or less completely unknown among the general public, which is unfortunate, and I’d like to make more people aware of it. This post is a start.

Bettinger, R., & Baumhoff, M. (1982). The Numic Spread: Great Basin Cultures in Competition American Antiquity, 47 (3) DOI: 10.2307/280231

Jennings, J., & Norbeck, E. (1955). Great Basin Prehistory: A Review American Antiquity, 21 (1) DOI: 10.2307/276104

Kaestle, F., & Smith, D. (2001). Ancient mitochondrial DNA evidence for prehistoric population movement: The Numic expansion American Journal of Physical Anthropology, 115 (1), 1-12 DOI: 10.1002/ajpa.1051

Lamb, S. (1958). Linguistic Prehistory in the Great Basin International Journal of American Linguistics, 24 (2) DOI: 10.1086/464442

Steward, J. (1937). Linguistic Distributions and Political Groups of the Great Basin Shoshoneans American Anthropologist, 39 (4), 625-634 DOI: 10.1525/aa.1937.39.4.02a00070

Read Full Post »

Linguistics and Archaeology in North America

Entrance to Alaska Ferry Terminal, Bellingham, Washington

The same special issue of the journal World Archaeology that I was discussing in the previous post has an article looking specifically at the relationship between linguistic and archaeological evidence in the study of the prehistory of North America. It is by M. Dale Kinkade and J. V. Powell, two linguists who specialized in the languages of the Pacific Northwest, so while the paper consists in part of a survey of the distribution of language families across the continent and ideas about their prehistory, the discussion of the Northwestern families is more in-depth than the others, and the case study that forms another section is on a specific region on the Olympic Peninsula of Washington state. That bias aside, it’s a very good article summarizing the state of knowledge as of its publication in 1976, and there have been few major breakthroughs since then (with one major exception that I’ll discuss in a later post), so most of it still holds up quite well.

One of the best parts of the paper is its discussion of methodology, particularly the reasons that historical linguistics can’t provide actual dates for events in the past the way archaeology can. This is now widely accepted by almost all linguists, but this was not always the case, as Kinkade and Powell note:

The most notable attempt to devise a procedure for dating linguistic prehistory is lexicostatistics. Developed in the 1950s  by Swadesh and others, lexicostatistics is based on three assumptions, all of which are, unfortunately, invalid. These premisses [sic] are:

1. A basic core vocabulary of 200 (or 100) words are less subject to change than other parts of the language, these words including terms for the same items in all languages.
2. The rate of retention of vocabulary items in the basic core vocabulary is constant through time.
3. The rate of loss is the same in all languages.

Note the similarity between these assumptions and those underlying radiocarbon dating, which I don’t think is a coincidence. I suspect there’s an interesting project here in the history of science looking at the influences on Swadesh as he was developing his ideas as well as the factors that led his theories to become quite popular among both linguists and archaeologists in the 1950s. By the early 1960s an increasing number of critics were pointing out the flaws in Swadesh’s approach, although there were still a significant number of adherents of glottochronology well into the 1970s (including some of the contributors to this same special issue). In any case, since then linguistics as a discipline has soundly rejected glottochronology, so Kinkade and Powell’s discussion of it here feels prescient rather than dated.

Having rejected glottochronology, Kinkade and Powell stick to the more “traditional” ways of linking historical linguistics to the more general study prehistory, primarily by looking at the distributions of established language families and to some extent loanword patterns, placenames, and the distributions of certain important lexical items. They discuss each language family considered fairly well-established (although a few on their list were and still are controversial) and summarize the proposals for its Urheimat. Some of these are obvious and not at all controversial, such as the placement of Athapaskan in Alaska. Others, such as Uto-Aztecan, have been subject to such fierce debate that there is really no consensus about where the proto-language was spoken or even what evidence is most useful to decide that. Kinkade present a map with what appear to be their preferred guesses for where each protolanguage was spoken, but the number of question marks on it indicate how uncertain they are about many of those guesses.

The final section of the paper is an in-depth discussion of the linguistic evidence bearing on the question of the occupation of the Ozette Village site on the Olympic Peninsula. All evidence clearly indicates that in late prehistory this site was occupied by the Makah people who still inhabit the general area, but Kinkade and Powell point to several lines of evidence, especially placenames and oral traditions of both the Makah and the Quileute people who live to the south of them, suggesting that the general area was occupied in earlier times by speakers of a Chimakuan language related to Quileute (which is not related to Makah). Since there are deposits at Ozette going back very far, this suggests in turn that while the most recent inhabitants of the village spoke Makah, this may not have been true of their predecessors. Kinkade and Powell don’t go into much detail about how they think this linguistic change may have occurred, but what they do say suggests that they see migration of Makah speakers from the north having physically displaced Chimakuan-speakers. Given the evidence they present, however, to me it looks at least equally likely that the change was the result of assimilation of Chimakuan-speakers to the Makah language and culture. The Makah and Quileute are apparently very similar culturally, and there seems to be some evidence that it was the Makah who were considered the more prestigious group.

Overall, this paper is a very useful summary of the state of research into the prehistory of language families in North America as of the mid-1970s, and since this hasn’t been a very active field of study since then it’s still a pretty good guide to the issues. This subject could really use a lot more attention and especially cooperation between linguists and archaeologists, as it has quite a lot of overlooked potential to shed light on the past. There are some signs that this potential is beginning to be explored, but there’s still plenty more to be done with it.

Kinkade, M., & Powell, J. (1976). Language and the prehistory of North America World Archaeology, 8 (1), 83-100 DOI: 10.1080/00438243.1976.9979654

Read Full Post »

The Roles of Linguistics and Archaeology

Linguistics Building, Rutgers University, New Brunswick, New Jersey

Sorry for the extended hiatus; I’ve been busy with various things. I’ll have more on the Mississippians at some point, but for now I want to  discuss a more general issue: the relationship of historical linguistics to archaeology in attempting to reconstruct past events. Both disciplines provide ways to study past events beyond the reach of historical scholarship in the traditional sense, which is based on written documentation. (I would argue that in many cases oral tradition provides an additional line of evidence, similar to written history in many ways, useful for understanding the prehistoric past, but that is a controversial position and I’m not going to defend it in detail right now.) What makes linguistics and archaeology particularly powerful is that they are independent lines of evidence, which means that tentative conclusions drawn from one can be compared to the evidence from the other to see how well they coincide. This provides a much more robust and reliable reconstruction of past events than would be possible based on either line of evidence alone.

And yet, despite the potential interpretive power to be gained from using linguistics and archaeology in tandem, this integration is rare, and both linguists and archaeologists have a tendency to ignore each other’s work most of the time and, when they do acknowledge it, to use it in a very uncritical and superficial manner that doesn’t come close to unlocking the full power of real integration. Back in 1976 the journal World Archaeology published a theme issue with several papers exploring the potential for integrating linguistics and archaeology, and it’s a sign of how little this integration has progressed that some of these papers are still useful summaries of the issues and the state of research.

The introductory paper in the issue, and the one I will primarily focus on in this post, is by Christopher Ehret of UCLA, and it provides a general overview of the types of historical inferences that can come from historical linguistic research. He notes that most research up to that time had been oriented toward just one of these types: those inferences that can come from evidence of genetic relationship between languages, which is to say, the knowledge that a given group of languages descends from a single “proto-language” presumably spoken by a single socio-cultural group at some point in the past. These are certainly useful, in a general way, but there are real limits to how much can be learned just from knowing which languages are related in a given region. Looking at the reconstructible vocabulary of the proto-language can give some important information about the culture that spoke that language that can, at least in theory, be correlated with archaeological evidence to pinpoint which archaeological “culture” corresponds to the speakers of that proto-language. Two other papers in this issue address different aspects of this kind of research in different context and with different language families: Robert Blust’s paper on Proto-Austronesian shows how information gained in this way can supplement the archaeological record by providing evidence for the presence of certain items of material culture and social institutions that are not recoverable by archaeology because of their perishable or intangible nature, while J. P. Mallory’s survey of research along these lines on Proto-Indo-European mostly points out the difficulty of attributing cultural items for which there are reconstructible words to a single culture when the proto-language being reconstructed may not actually represent a single language or its speech community.

Moving beyond these issues, however, Ehret points out that there is more to historical linguistics than determining which languages are related and what words can be reconstructed for various proto-languages. A potentially much more productive line of evidence for culture history, and yet one that has seen remarkably little research, is loanword studies. Languages may adopt words from other languages for a variety of reasons, many of them quite important for understanding political and economic relationships between societies at various points in the past. Furthermore, loanwords are often (though not always) easy to identify either in currently spoken languages or in the reconstructed proto-languages from which they descend. Ehret gives various examples from his own research in Africa to illustrate how loanword studies can give substantial insights into cultural relationships and history.

It’s important to note, however, that while most of what Ehret says in this paper about general methodology is unexceptionable, his own conclusions about African prehistory are quite controversial and one of his faculty pages at UCLA even has this remarkable paragraph:

His linguistic works include A Comparative Reconstruction of Proto-Nilo-Saharan (2002), Reconstructing Proto-Afroasiatic (1995), and The Historical Reconstruction of Southern Cushitic Phonology and Vocabulary (1980). He has also written monographic articles on Bantu subclassification, on internal reconstruction in Semitic, on the reconstruction of proto-Cushitic and proto-Eastern Cushitic, and, with Mohamed Nuuh Ali, on the classification of the Soomaali languages. These reconstructions have not been well received, and are not followed by other linguists.

Evidence for the controversy engendered by Ehret’s interpretations comes in the very same special issue for which he wrote this introductory paper. One of the other papers, by the archaeologist D. W. Phillipson, addresses the Bantu language family and the potential for both linguistics and archaeology to shed light on the issue of when and how Bantu-speakers spread across much of southern and eastern Africa. Phillipson notes Ehret’s interpretations but disputes them in detail on various points. This is obviously a one-sided account, but Phillipson’s arguments seem pretty strong to me. I don’t know much about this issue, of course, and it’s quite likely that research has progressed a bit since 1976 in any case, so I’m not going to draw any conclusions about who was more right.

Africa is actually an interesting case here because it seems that historical linguistics has played a much bigger role here than elsewhere in developing hypotheses about prehistory (probably due at least in part to Ehret’s work). This is in contrast to the Americas, where linguistics and archaeology have mostly operated separately and the latter has been more dominant in developing historical hypotheses. I think the African model offers a potentially productive route for Americanists to take in trying to come up with more detailed reconstructions of culture history, although the many controversies over the proper interpretation of African prehistory show that this more integrated approach is by no means a cure-all.

I’ll have more on the potential implications of all this for North America later. For now I just want to introduce the topic.

Blust, R. (1976). Austronesian culture history: Some linguistic inferences and their relations to the archaeological record
World Archaeology, 8 (1), 19-43 DOI: 10.1080/00438243.1976.9979650

Ehret, C. (1976). Linguistic evidence and its correlation with archaeology World Archaeology, 8 (1), 5-18 DOI: 10.1080/00438243.1976.9979649

Mallory, J. (1976). Time perspective and proto‐indo‐European culture World Archaeology, 8 (1), 44-56 DOI: 10.1080/00438243.1976.9979651

Phillipson, D. (1976). Archaeology and Bantu linguistics World Archaeology, 8 (1), 65-82 DOI: 10.1080/00438243.1976.9979653

Read Full Post »

Does Language Density Correlate with Latitude?

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

Read Full Post »

Why Don’t Navajos Eat Fish?

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

Read Full Post »

Gendered Language among the Pueblos

 

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

Read Full Post »

The “Kiowa Apaches”: Neither Kiowa Nor Apache?

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

Read Full Post »

Athapaskan Influence on Tewa

Painted Wall, Santa Fe, New Mexico

There’s been quite a bit of research on relations between the Pueblo and Athapaskan peoples of the American Southwest, most of it falling within the broad domain of ethnography or sociocultural anthropology.  That is, there is quite a lot of evidence that some of the Athapaskan-speaking Apache groups, especially the Navajos, have been in close contact with the Pueblos and have adopted many Pueblo cultural practices.  It’s not clear when or how this happened, however; with regard to the Navajos specifically the Pueblo Revolt period (AD 1680 to 1692 or a bit later) has often been posited as a time when an influx of Pueblo refugees to the Navajo country led to the adoption of many Pueblo cultural practices by the Navajos, but recent archaeological research has cast doubt on this theory.  (More on that later.)  In general, there has been plenty of documentation of Pueblo influences on Apaches, but little progress on figuring when and how these influences occurred.

The issue of influences in the other direction, from Athapaskan-speaking groups to Pueblos, has received much less attention.  This is probably due mostly to the longstanding if mostly implicit idea among anthropologists that the Pueblos, as settled agriculturalists, were in some sense more developed culturally than the Athapaskans, who were certainly hunter-gatherers when they entered the Southwest although many of them adopted agriculture to varying degrees once they were there.  The idea seems to be that more “primitive” hunter-gatherers would of course have borrowed lots of cultural practices from the more “advanced” farmers they encountered (and, indeed, they did), but that the Pueblos wouldn’t have borrowed much, if anything, from the barbaric newcomers to the area they had inhabited for millennia.

Nevertheless, there is in fact some evidence for cultural influence flowing from Athapaskans to Pueblos in addition to vice versa.  Due to the abovementioned lack of research there aren’t very many specific instances of influence to point to, but I have found one clear-cut instance of linguistic influence in this direction.  This is explained in a short but very interesting article by Paul Kroskrity of UCLA published in 1985.  Kroskrity points out that Tewa, a Pueblo language spoken in northern New Mexico around Santa Fe as well as in one village in the Hopi area of Arizona (the latter resulting from a migration from the Galisteo Basin south of Santa Fe after the Spanish Reconquest of New Mexico in 1692), has a possessive construction very different from those found in the other languages of the Kiowa-Tanoan family to which it belongs.  The usual way of expressing possession in these languages is with a periphrastic construction roughly equivalent to “the x y has” in English.  Tewa, however, alone among Kiowa-Tanoan languages, has in addition to this construction another, simpler way to express possession with a single morpheme, a suffix -bí attached to the possessor.  This is equivalent to saying “y’s x” in English.  This suffix appears to be fixed in form, regardless of the number or gender of either the possessor or the possessed.

Since this suffix is found only in Tewa, and not in any other related languages, it is a good candidate for a borrowing from some other language.  And, as it happens, the Athapaskan languages have a very similar morpheme used to express a third-person possessor, a prefix bi- appended to the possessed.  Since both Tewa and Athapaskan have the usual word order “possessor-possessed” the fact that the morpheme is a prefix in Athapaskan but a suffix in Tewa is no problem; it seems Tewa just attached it to the previous word rather than the following one.

Linguistically the case is straightforward.  Culturally and historically, the implications are more complicated.  Generally linguistic borrowing goes from a language perceived to have more prestige to one perceived to have less, and while there are some exceptions this case can probably be best explained in that framework, which rather turns the idea of “advanced” Pueblos and “primitive” Apaches on its head.  At the very least it implies that the relationship was a two-way street.  This raises the interesting question of when and why Tewa-speakers would have perceived Athapaskans to have more status, to which I have no answer at this point.  An alternative explanation would be that some Athapaskan language was used at some point as a lingua franca for communication among the various Pueblos (which speak several languages, not all of them related), with the adoption of useful grammatical devices from this intercultural contact language being adopted into Tewa as more useful than the native constructions used for the same purpose.  This explanation however does not account for Tewa being the only one of the Kiowa-Tanoan languages to adopt this construction.  More extensive data on possible Athapaskan loanwords into Tewa and other Pueblo languages would be helpful in clarifying this.  Generally words are borrowed much more easily than grammatical structures like this, so any structural borrowing will usually only come after fairly extensive borrowing of words.

I think the most likely explanation for this borrowing is that there was some time in the past when Tewa-speakers were in close contact with speakers of one of the Apachean languages in a context where the Apachean-speakers had a relatively high level of perceived status.  This must have been before the early eighteenth century, when the Arizona Tewas migrated to the Hopi country from the Galisteo Basin, since Kroskrity’s data makes it quite clear that this construction is shared by both Tewa groups (which have not been in close contact since the migration).  The odd thing about this is that the Tewas are located near the center of the Rio Grande Pueblo culture area, which makes it look unlikely at first glance that they would have had more contact with outside groups than their linguistic relatives the Tiwa (to the north and south) and Towa (to the east and west).  While the Northern Tiwa and Eastern Towa are known to have had quite close connections to some of the Apache groups of the southern Plains (Jicarilla and Lipan, respectively), the Tewa were not in the same situation.

The only Athapaskan group that seems like a plausible candidate for loaning this construction to the Tewa is the Navajos, who in early historic times were located northwest of Tewa territory, not particularly close but without much in between.  Ethnographic and historic data tends to suggest that the Tewa and Navajos were generally adversaries in recent centuries, but this may not necessarily have always been the case.  They were certainly familiar with each other; indeed, the name “Navajo” comes from the Tewa word navahu, meaning “plowed fields” and borrowed in to Spanish as a way of distinguishing the Navajos, with their greater emphasis on agriculture, from other Apaches who were still primarily hunter-gatherers.  That very emphasis on agriculture, of course, may in and of itself indicate somewhat closer contact with the Pueblos.

This is a good example of the largely unrealized potential of linguistics to contribute data to the understanding of culture change and prehistory.  Linguists and archaeologist don’t tend to use each other’s data much (and when they do they all too often seize on the most superficial and/or tendentious interpretations available rather than the most reasonable), but there is an enormous amount of information available to contribute to a fuller picture of the past if people are willing to figure out how to use it judiciously.

Kroskrity, P. (1985). Areal-Historical Influences on Tewa Possession International Journal of American Linguistics, 51 (4) DOI: 10.1086/465943

Read Full Post »

Dena’ina and Ahtna

Sign at the Entrance to Ahtna Property, Mentasta, Alaska

In discussing Dena’ina, the most divergent of the Athapaskan languages, I mentioned that its Upper Inlet dialect (spoken in the area that is now Anchorage and the Matanuska-Susitna Valley, i.e., the most populous part of Alaska today) shows extensive influence from Ahtna, the language spoken to the east.  This probably accounts for most of the many differences between this dialect and the more conservative Dena’ina dialects further west in the lower Cook Inlet area.  James Kari of the Alaska Native Language Center, who did the research on Dena’ina dialects I discussed in the previous post, has also done a lot of research on this contact between Ahtna and Upper Inlet Dena’ina, and he published an article summarizing his findings in 1977.

Ahtna is also an Athapaskan language, and it may be more closely related to Dena’ina than either is to the rest of the family; Kari notes a few very deep-seated grammatical similarities between the two.  Be that as it may, however, Dena’ina has undergone unique changes that make it very different from every other Athapaskan language, including Ahtna, and the two are not mutually intelligible.  All the Dena’ina dialects show sufficient similarity to indicate that they arose relatively recently from a single protolanguage, as discussed by Kari in a more recent paper, and it seems that that language had been relatively isolated from most other languages in the family for a relatively long period of time before its recent expansion.  Kari favors a location in southwestern Alaska, probably northwest of what is now Lake Clark National Park, for the Dena’ina Urheimat, although he acknowledges that somewhere in the Upper Inlet is another possibility.  This seems implausible to me, however, since one of the major differences between Dena’ina and most other Athapaskan languages is its radically simplified vowel system, which seems to show evidence of influence from Yup’ik Eskimo, which is spoken along the west coast of Alaska and probably has been for a long time.  Kari notes that there are relatively few Eskimo loanwords in any Dena’ina dialect, however, which suggests to me that this vowel system may have spread to proto-Dena’ina from another Athapaskan language in western Alaska which had in turn gotten it from Yup’ik.  (Note that the Yup’ik are the Eskimos who are not Inuit; Inuit, which is spoken further north, has an even simpler vowel system.)

In any case, by the time the Dena’ina spread into the Cook Inlet basin and began to adopt, to varying degrees, a more maritime subsistence pattern they had been separated from the main mass of Athapaskan speakers, who were mostly concentrated along the Yukon and Tanana Rivers, for a very long time, and their language was very distinct.  This distinctiveness may in fact have been deliberate, at least in part; one of the most distinctive characteristics of Dena’ina is its use of words, often for very common concepts like “water” and “fire” that appear to be Athapaskan in origin but are not the standard Athapaskan roots used by most languages for those meanings.  Kari implies that this “tabooistic replacement” (a term I’m not crazy about) may have had something to do with maintaining cultural boundaries, probably an important consideration for a group expanding into new and relatively unfamiliar territory.

When the Dena’ina reached the upper end of Cook Inlet, they would have soon encountered the Ahtna.  While they were (and are) based primarily in the valley of the Copper River to the east, the Ahtna were a  large, powerful tribe that expanded as far west as the Denali area, and as Kari notes they appear to have had the superior position in their contacts with the Dena’ina.  There are many Ahtna loanwords into Upper Inlet Dena’ina, as well as a certain amount of phonological influence leading to simplifications in the consonant system that make Upper Inlet Dena’ina very hard for speakers of other Dena’ina dialects to understand.  There were some borrowings in the other direction too, but fewer.  Many of Kari’s informants even seem to have explicitly said that they considered the Ahtna culturally superior to themselves.  The historic and ethnographic records give plenty of supporting evidence of extensive contact, trade, and intermarriage between the Upper Inlet Dena’ina and the western Ahtna, which corroborates the very strong linguistic evidence Kari presents.

The upshot of all this for culture history is not totally clear, but it definitely seems that the Dena’ina most likely spread from west to east, and the Ahtna may have been spreading in the other direction at approximately the same time.  Archaeological evidence suggests that the Dena’ina spread into the Kenai peninsula (across the inlet from their apparent homeland) took place less than a thousand years ago.  Since it seems very clear that the Athapaskans had been in Alaska for a very long time before that, certainly long enough for Dena’ina in the southwest to diverge markedly from the various languages that form a large dialect continuum in the Tanana and Yukon valleys, this suggests that the story of Athapaskan prehistory is both very complicated and very long-term.  Specific inferences useful in understanding the migration of other Athapaskans to the American Southwest, a process which must have begun long before the Dena’ina entered the Kenai, are not apparent.  But then, this is the opposite end of the Athapaskan world, so that’s not all that surprising.

Read Full Post »

The Puzzling Dena’ina

Dena'ina Civic and Convention Center, Anchorage, Alaska

As I mentioned in the last post, it’s generally thought that the Athapaskan migrations which eventually led to the entrance of the Navajos and Apaches into the Southwest began in Alaska.  The northern Athapaskan languages are actually spoken over a very large area of northwestern Canada as well, but the linguistic evidence clearly points to Alaska as the original place where Proto-Athapaskan was spoken at the last point before it splintered into the various Athapaskan languages.  That is, the Urheimat of the Athapaskans seems to have been somewhere in Alaska.

There are two main pieces of evidence pointing to this conclusion.  One is the fact that it has been quite well established at this point that the Athapaskan language family as a whole is related to the recently-extinct language Eyak, which was spoken in south-central Alaska.  Eyak was clearly not an Athapaskan language itself, but it had sufficient similarities to reconstructed Proto-Athapaskan to establish a genetic relationship.  Since Eyak was spoken in Alaska, it therefore seems most probable that the most recent common ancestor of both Eyak and the Athapaskan languages (Proto-Athapaskan-Eyak) was also spoken in Alaska.  This is not very strong evidence on its own, however, since Eyak was only a single language spoken by a relatively small group of people, and there is some fairly strong evidence that they have not lived in their current location for a very long time (although they may well have moved there from elsewhere in Alaska).

A stronger piece of evidence is the internal diversity of the Athapaskan languages themselves.  A general principle in historical linguistics is that the Urheimat of a language family is likely to be found where there is maximal diversity among the languages in the family.  That is, since language families often spread through migration, areas with many languages from the family that are not particularly closely related to each other in relatively small area are more likely locations for the Urheimat than areas with fewer languages that are more closely related.  When it comes to Athapaskan, this condition obtains most strongly in Alaska.  The languages in Canada and the Lower 48 are all relatively closely related to each other within the family as compared to some of the languages in Alaska.  Although interior Alaska is overwhelmingly dominated by Athapaskan groups, the linguistic boundaries among these various groups, even those adjacent to each other, are often extremely sharp.

This is particularly the case when it comes to the most divergent of all the Athapaskan languages: Dena’ina.  (Note that in many publications this term is spelled “Tanaina,” and I have spelled it that way myself in the past.  ”Dena’ina” is the generally preferred form these days, being closer to the original pronunciation and consistent with the principles of the currently used orthography.)  This is the language traditionally spoken around Cook Inlet in south-central Alaska, including the Anchorage area.  While it’s clearly Athapaskan, it’s very weird as Athapaskan languages go.  It is not mutually intelligible with any other Athapaskan language, although it borders several of them, and it is in turn divided into several internal dialects that are strikingly diverse.  James Kari of the Alaska Native Language Center at the University of Alaska Fairbanks, published an article on Dena’ina dialects in 1975 based on original fieldwork.  The basic conclusions were that there are two main dialects, Upper Inlet and Lower Inlet, and that Lower Inlet is further subdivided into two or three subdialects: Outer Inlet, Inland, and Iliamna (which lies between the two other dialects and shows similarities to both, although Kari seems to consider it likely that it is ultimately more closely related to the Outer Inlet dialect with the similarities to the Inland dialect due to later influence).  In general, the Lower Inlet dialect is more conservative than the Upper Inlet one, which shows extensive influence from the neighboring Ahtna language, which is also Athapaskan but not very similar to Dena’ina.  Within the Lower Inlet dialect, the Inland dialect is the most conservative, as well as the one with by far the most speakers of any of the dialects as of 1975.  Both these conditions are presumably due to the relative isolation of this dialect, which is spoken in the Lake Clark area and further north in Lime Village.  Kari considers this the most likely homeland of Dena’ina speakers, whom he sees as having moved from the interior to the coast relatively recently, although he acknowledges that this cannot be established based on linguistic evidence alone.

Despite the relative conservatism of the Lower Inlet dialect, however, all its subdialects do show a  certain amount of influence from Yup’ik Eskimo (particularly in the development of the Proto-Athapaskan vowel system).  This is unsurprising, as these dialects lie on the boundary of the Yup’ik area to the west and south, and the Dena’ina groups in these areas show extensive Eskimo influence in many aspects of their traditional culture.  This is particularly the case for the Outer Inlet groups, especially those at the southern end of Cook Inlet, who adopted a much more maritime-oriented culture than is typical of Athapaskans.  These cultural distinctions within the overall Dena’ina society were documented by Cornelius Osgood of Yale in the 1930s.  In an article published in 1933 he noted that the main distinctions among the Dena’ina groups were economic, having to do with their subsistence systems, while other social systems were pretty similar across the various groups.  The Lower Inlet groups, especially those in the Seldovia area on Kachemak Bay near the outlet of Cook Inlet, showed a much heavier dependence on hunting sea mammals and a correspondingly heavier influence from nearby Yup’ik Eskimo groups with a similar adaptation than their compatriots further north who had a more typically Athapaskan lifestyle based on salmon fishing and hunting of terrestrial animals.

The fact that Dena’ina, the most divergent of the Athapaskan languages and therefore the one that most likely split earliest from Proto-Athapaskan, is spoken in Alaska makes it very likely that Proto-Athapaskan was spoken in Alaska as well.  Indeed, if Kari is right that the Lake Clark area was the original homeland of the Dena’ina, this potentially places Proto-Athapaskan quite far west within Alaska and quite close to areas traditionally occupied by Eskimo-speakers.  It should be noted, however, that it’s still very unclear when the breakup of Proto-Athapaskan occurred and who occupied which parts of Alaska at that time.  Nevertheless, the evidence from Dena’ina, which is still one of the least-understood Athapaskan languages despite its obvious importance for Athapaskan studies, seems to pretty clearly show that whenever the important early events in the history of the language family occurred, they almost certainly occurred somewhere in Alaska.

Osgood, C. (1933). Tanaina Culture American Anthropologist, 35 (4), 695-717 DOI: 10.1525/aa.1933.35.4.02a00070

Read Full Post »