Variability and Uniformity in Ute Domestic Architecture

By Rand A. Greubel

(Alpine Archaeological Consultants, Inc.)

Presented at the Colorado Council of Professional Archaeologists Annual Meeting,

March 3, 2001, La Junta, Colorado

The title for this paper was inspired by an idea I had about so-called brush structure technology; more specifically, Ute wickiups, which grew out of my experiences looking for ¾ and occasionally finding ¾ wickiups on survey and excavating several such structures during the recent TransColorado Natural Gas Pipeline data recovery project. I, like other archaeologists who have dealt with these types of structures, have grappled with what they mean, how to categorize and compare them, and in many cases how to even recognize their existence when nearly all physical trace of the structure is gone. This paper addresses what I feel are some important and worthwhile issues related to interpreting wickiups.

Ute architecture of the Protohistoric and early Historic periods has, in general, been sparsely described in the archaeological literature, primarily due to a dearth of data from excavated contexts. Tipis, which were adopted by the Eastern Ute during the historic period, are well understood from a technological standpoint, in large part because they survived into historic times and are still widely used. Wickiups, often derided as mere ephemeral brush structures, have not fared so well. Most of what is known, or what is believed to be known, about Ute (and Paiute) wickiups has been gleaned from ethnography (e.g., Kelly 1964; Smith 1974; Stewart 1942) and a limited number of archaeological studies based on primary and secondary data sources (e.g., Baker 1996; Buckles 1971; Conner 1988; Huscher and Huscher 1939; Sanfilippo 1998; Scott 1988).

Unlike architecture that is regarded as “substantial,” such as pithouses and above-ground masonry structures, archaeological wickiups, for the most part, lack well-defined architectural traits that provide bases for comparison and permit the construction of typologies that allow structures to be defined and traced across time and space. In the absence of features such as intact walls and well-defined floors that can be consistently defined archaeologically, we are forced to discuss the architectural aspects of wickiups from the perspective of a few surviving examples that have been adequately described during the last few decades, to rely on historical or ethnographic descriptions of wickiups, or to abandon talking about architecture ¾ in the sense of built superstructures ¾ altogether, and instead focus on site structure or the layout of features and activity areas in relation to the dwelling. One thing that is clear is that Ute domestic architecture was highly variable. But why were wickiups so variable and just how variable were they?

It is appropriate to begin a discussion of Ute wickiups by addressing this variability. Thirteen years ago the CCPA hosted a symposium entitled Archaeology of the Eastern Ute, which resulted in the publication of an important volume devoted to this subject. One of the papers in that volume, by Doug Scott, reviewed the then-current state of knowledge about Ute wickiups. Scott wisely referred to wickiups as either “conical timbered lodges” or as “wickiups,” avoiding that unfortunate term, “brush structures,” which actually describes only a subset of wickiups, albeit a common and important subset. In fact, wickiups may have also been bark structures, hide structures, and, in recent history, canvas structures. Aside from covering, Scott noted that wickiups exhibit variability in many other ways as well: lean-to versus free standing, pole length, pole diameter, the use of stones around the base, the presence or absence of interior hearths, and unprepared dirt floors versus juniper bark floor coverings. To these might be added the use of either or both small tree trunks and branches as poles, the use of forked poles, the use of expedient poles formed from tree branches pulled down and broken, presence or absence of associated artifacts, which direction the entryway faces, and many other considerations. Aside from merely illustrating the obvious fact that conical pole structures exhibit considerable variability, the point is that archaeologists who have worked with such architecture have either tended to downplay the variability by considering all wickiups to be ephemeral in nature, or have attempted to come to grips with it by constructing typologies. It is my contention that neither route offers the most satisfactory approach to interpreting this type of architecture.

One problem with the first approach ¾ that is, de-emphasizing variability in favor of the view that wickiups are crude, simple structures that represent ephemeral or very brief occupations, and that any differences they exhibit have little real significance ¾ is that it ignores the increasing body of data that indicate that wickiups can be quite sophisticated dwellings that sometimes were occupied for rather long periods of time. The tendency to ignore or de-emphasize wickiup variability is dangerous because it fails to investigate the meaning of the variability.

The second approach tries to capture and address variability by creating wickiup types based upon characteristics that tend to vary from structure to structure. Typologies have their advantages and this approach can potentially be useful by allowing us to formalize the changeable aspects of this type of architecture and to explore their interplay with other variables such as environment, subsistence strategy, duration of occupation, site function, and others. However, typologies can also obscure variability by creating the impression that our types represent real phenomena, and thereby drawing our attention away from the truly important issues and creating a dangerous situation wherein we will begin to simply hang convenient labels on these structures and cease to talk about why they are variable.

The exercise of constructing a wickiup typology misses what is arguably the most important aspect of this technology — its plasticity and adaptability. It should be expected that every wickiup, although perhaps sharing certain important characteristics with other wickiups, was unique to a certain degree, and that this uniqueness can potentially reveal information about site function, seasonality, social organization, subsistence, and other avenues of inquiry. With this in mind, it seems appropriate to discuss this technology in terms of the range of options available to the wickiup-builder in a particular environment, because the options not chosen for a particular structure may be as informative as those that were. For example, unlike tipis, wickiups were not limited by standardized construction methods and curated construction materials, but could be adapted to suit the environment, season, current weather conditions, local building materials, soil type, site function, duration of occupation, and other variables. Thus, personal preference coupled with unique circumstances may have played major roles in wickiup construction. I feel that a strongly contextual approach is useful for addressing the flexibility of this technology and does not necessarily preclude the simultaneous use of a simple typology that might allow intrasite and intersite comparisons.

To illustrate this approach, I would like to briefly discuss two wickiups that I recently excavated during the TransColorado Natural Gas Pipeline data recovery project. Structure 2 at the Simpson Wickiup site (5SM2425), eight miles west of Norwood, Colorado, was a relatively large wickiup constructed of juniper poles, all of which were collapsed except one. It had been built into a large living juniper tree that formed the northwestern boundary of the structure. The poles consisted of both the trimmed trunks of small juniper trees as well as large branches. As dendrochronology failed to provide dates from the structure poles, it is not known whether the poles were acquired fresh or as dead wood. Considering that as many as six of the poles were juniper trunks ¾ not branches ¾ it seems probable that at least some were obtained from living trees, as it is unlikely that an ample supply of dead junipers of precisely the right size and shape were found standing on the site, unless they had been killed through bark stripping during earlier occupations, which as we know from John Cater’s paper is a real possibility. Aboriginal tree-felling at the Bustos Wickiup site, a Protohistoric Shoshone encampment in Nevada, has been described by Simms (1989); similar methods may have been employed here, although no direct evidence for them was found. However, even if the trees were dead and procured by toppling them, there would still have been a large amount of labor involved in trimming off the branches. The use of trunks, which probably served as foundation poles, allowed the structure to be built larger and sturdier than it could have if only branches were used. The extra work involved in acquiring trunks suggests that structure size, sturdiness, and a long use-life were critically important to the builders. Ethnographic research suggests that structure size is linked to anticipated mobility; that is, the longer an occupation is anticipated to be, or if reoccupation is planned, the larger and more substantial the dwelling will be (Kent 1991). Structure coverings can perhaps be viewed along this same line of reasoning — more care will be taken in selecting closing material for a dwelling that is intended for a lengthy stay than one that will be occupied only briefly. A brush covering, for example, is quick and expedient but will not keep out moisture and wind as effectively as a juniper bark covering. There was little direct evidence of the closing material used on most of Structure 2. However, the presence of stripped juniper trees nearby, including one tree-ring dated to the period when the structure was likely inhabited, suggests that the wickiup may have been covered with strips of bark. Season of occupation likely also played a crucial role in choosing certain options over others. In general, these considerations are consonant with those linked to anticipated duration of occupation; that is, in addition to being better suited for a lengthy occupation, a larger, more substantial structure is also more appropriate for cool-weather habitation. Ute architectural technology, therefore, reflected planning depth (Binford 1987) in the sense that decisions regarding building materials and construction techniques were integrally linked to future plans (anticipated mobility), tempered by seasonal and weather considerations. In this particular case, additional evidence that a rather lengthy, cool-season occupation either took place or was planned was the presence of an interior hearth, intramural activity areas, and patterned disposal of refuse in extramural areas.

My second example is taken from the Schmidt site, several miles from the Simpson Wickiup site, near Redvale, Colorado. The Schmidt site (5MN4253) is a large site with multiple components including several Ute camps, three of which were excavated to one degree or another. Structure 2 in Locus 2 illustrates a common problem in the investigation of so-called ephemeral architecture, a problem that will only increase with time. It was so highly deteriorated that no structure was apparent based on attributes normally used to define the presence of wickiups. Two heavily decayed poles ¾ one of which was forked ¾ lay near a very old, mature juniper. Many of the lower branches on the northeastern side of the juniper appear to have been trimmed and possibly broken off, presumably to allow enough headroom for the structure. The lowest of these branches was approximately 1.4 m above the modern ground surface. The two weathered poles and the trimmed branches of an old tree were the only surviving architectural remains of the wickiup.

Excavation of this area revealed little. No floor could be defined and no features were found that could be construed as internal to the structure. The conjectural interior of the structure was notably sparse in terms of debitage, faunal remains, and rock of any type. In contrast, a cluster of five features was situated immediately to the east, and a dense concentration of lithic debitage and tools lay 2-3 m to the southeast. This patterning—a relatively empty space directly beneath a large juniper, surrounded by cultural features and artifact concentrations—is probably the best evidence that a dwelling existed at this location. The hearth and ash dump cluster may have been an extramural work area, and the artifact scatter to the south of the features may reflect a combination of primary and secondary deposition of flaking refuse and discarded tools. The lack of a central hearth suggests a warm season occupation. It is equally plausible that a brush shade or windbreak ¾ structure types that might not be normally characterized as wickiups ¾ was present. This would be an even stronger indication that the occupation represented by the feature cluster described above took place during a warm season. Structure 2 reflects two important principles, one being the Ute penchant for the ingenious and opportunistic modification of the natural environment to provide shelter, and the other being that spatial patterning of features and materials can be used to infer the former presence of a habitation structure that is no longer extant. I realize that we are drifting away from our original consideration which is architecture, but my point is that interpretations of architecture must be contextual, and that in cases like this must include, or even rely on, other lines of evidence such as spatial patterning.

So far I have intentionally avoided very much discussion of spatial patterning, or site structure. In a paper that dealt with Ute wickiup site structure that I presented at the last Rocky Mountain conference (Greubel 1999), I argued that many Ute wickiups may exhibit a fairly consistent suite of spatial characteristics that will serve to distinguish them in the archaeological record from the habitation structures of other, adjacent cultural groups such as the Navajo. These characteristics include things like entryway orientation, interior hearth placement, hearth furniture, arrangement of intramural activity areas, and the locations of extramural hearths, activity areas, and refuse deposits. It may be that I have only identified local patterns that will not hold true on a regional scale, but the point is that wickiup spatial patterning generally appears to be less variable than the architecture itself. In other words, whereas the construction of wickiups was conducted within broad parameters, allowing considerable latitude with construction materials and structure morphology, the underlying spatial principles of wickiups may have been much more consistent and uniform from structure to structure. This is different from the architecture of other groups such as the Navajo. Although Navajo hogans also exhibit variability, overall they seem to have conformed to more narrowly defined structural parameters than did wickiups, and they also had very predictable layouts. Navajo hogans and the tipis of many plains groups are uniform in construction and layout because they are invested with a great deal of symbolism linked to social values, cosmology, and religious beliefs (Oetelaar 2000; Sanfilippo 1998). Based on ethnographic observations (e.g., Smith 1974) and information given by native informants (e.g., Sanfilippo 1998), it seems unlikely that Ute wickiups were subject to equally stringent rules and codes governing their construction and layout. More likely, construction was dictated by necessity, available materials, and setting; whereas layout may have been ruled by custom combined with very practical considerations such as anticipated duration of occupation, prevailing breezes, and refuse disposal practices. The archaeological data seem to support this interpretation.

If it is true that Ute wickiups were less restricted by architectural conventions related to social, religious, or symbolic issues than were, for example, Navajo hogans, Lakota tipis, and Anasazi kivas, then they may have been more responsive to purely environmental, economic, or ergonomic factors. Wickiup sites, therefore, may provide excellent laboratories for examining issues related to use of the environment, seasonality, subsistence strategies, and technology, because they may lack, or exhibit less of, the religious or symbolic aspect in their construction.

The above discussion suggests that we should expect Ute wickiups to be highly variable architecturally, and should, perhaps, not attach too much significance to wickiup form. Instead, each aspect of an individual structure — such as pole type, closing material, entryway orientation, setting, and associated features and activity areas — should be examined with the goal of determining why that particular option might have been selected over any other possible option. Because of its premise that structure form is closely linked to site function, anticipated and actual duration of occupation, subsistence, and many other considerations, such an approach will necessarily situate the structure within the larger context of the site and the settlement system, as opposed to the alternative approach of categorizing it according to a simple typology based on a limited number of variables. Of course, the contextual approach should also be flexible enough to recognize the existence of legitimately different structure types where they exist, so as to distinguish habitation structures from similar constructions such as menstrual huts, hunting blinds, windbreaks, shades, or even meat-drying racks.

In conclusion, although I don’t think that I have defined any new tools or methodologies for analyzing Ute wickiups or other types of conical pole structures, I hope that I have brought some of the problems associated with such tasks into sharper focus by pointing out some of the important interpretive issues surrounding this type of architecture.

References Cited

Baker, Steven G.

1996 Numic Archaeology on the Douglas Creek Arch, Rio Blanco County, Colorado: Ute Rancherias and the Broken Blade Wickiup Village (5RB3182). Centuries Research, Inc. Prepared for the Bureau of Land Management, Meeker, Colorado.

Binford, Lewis R.

1987 Researching Ambiguity: Frames of Reference and Site Structure. In Method and Theory for Activity Area Research: an Ethnological Approach. Edited by Susan Kent, pp.449-512. Columbia University Press, New York.

Buckles, William G.

1971 The Uncompahgre Complex: Historic Ute Archaeology and Prehistoric Archaeology of the Uncompahgre Plateau, West Central Colorado. Ph.D Dissertation, Department of Anthropology, University of Colorado. University Microfilms, Ann Arbor.

Conner, Carl E.

1988 Archaeological Investigations at 5EA433. In Archaeology of the Eastern Ute: A Symposium, edited by P.R. Nickens, pp. 190-205. Occasional Papers No. 1, Colorado Council of Professional Archaeologists, Denver.

Greubel, Rand A.

1999 The Spatial Organization of Activities at Two Ute Sites in the Southern Rocky Mountains. Paper presented at the 4^th Rocky Mountain Anthropological Conference, Glenwood Springs, Colorado.

Huscher, Betty H. and Harold A. Huscher

1939 Field Notes for 1939. Unpublished ms. on file in the Department of Anthropology, Denver Museum of Natural History.

Kelly, Isabel T.

1964 Southern Paiute Ethnography. University of Utah Archaeological Papers No. 69. Salt Lake City.

Kent, Susan

1991 The Relationship Between Mobility Strategies and Site Structure. In The Interpretation of Archaeological Spatial Patterning, edited by Ellen M. Kroll and T. Douglas Price, pp.33-59. Plenum Press, New York, New York.

Oetelaar, Gerald A.

2000 Beyond Activity Areas: Structure and Symbolism in the Organization and Use of Space Inside Tipis. Plains Anthropologist 45(171):35-61.

Sanfilippo, Joanne M.

1998 Ute Wickiups or Navajo Forked-Stick Hogans: Determining Ethnicity through Architecture in the Archaeological Record. Thesis, submitted to the Department of Anthropology, Northern Arizona University, Flagstaff.

Scott, Douglas D.

1988 Conical Timbered Lodges in Colorado or Wickiups in the Woods. In Archaeology of the Eastern Ute: A Symposium, edited by Paul R. Nickens, pp. 45-53. Colorado Council of Professional Archaeologists Occasional Papers No. 1. Denver, Colorado.

Simms, Steven R.

1989 The Structure of the Bustos Wickiup Site, Eastern Nevada. Journal of California and Great Basin Anthropology 11: 2-34.

Smith, Anne M.

1974 Ethnography of the Northern Utes. Papers in Anthropology, No. 17. Museum of New Mexico Press, Santa Fe.

Stewart, Omer C.

1942 Culture Element Distributions: XVIII, Ute-Southern Paiute. University of California Anthropological Records 6(4):231-354. Berkeley.