Go to Table of Contents.
List of Tables
List of Illustrations
Research Objectives and Methods
Architecture and Site Layout
Population Estimates
Faunal Remains
Archaeobotanical Remains
Human Skeletal Remains
Water Control and Subsistence
Abandonment and Emigration
Appendix A

Architecture and Site Layout

by Melissa J. Churchill

Buildings serve the needs and reflect the ideas and beliefs of the people who construct them (see Gilman 1987*1:538; Hegmon 1989*2:5; Kenzle 1997*1:197). It follows, then, that social organization should be reflected in the built environment, specifically in the architectural layout of a site (Lipe and Hegmon 1989*1:21). One of Crow Canyon's primary research objectives at Woods Canyon Pueblo was to characterize the internal organization of the village by examining its architectural layout. The discussion in this chapter focuses on the four main topographic areas of the site (referred to as "sections" to distinguish them from the seven numbered areas defined in "Chronology" on the basis of spatial associations between structures and nonstructures) and on changes that occurred in the configuration of the pueblo through time. I also describe the types of buildings at Woods Canyon Pueblo and discuss their possible uses.

Site Layout

Woods Canyon Pueblo was an aggregated, canyon-oriented village bisected by a drainage. The site has many of the characteristics that define late canyon-head and canyon-rim villages—for example, an informally bounded plaza, low walls that enclose part of the site, towers built on boulders, a D-shaped structure, and buildings located on the talus slope as well as on the canyon rim (Kelley 1996*1; Kenzle 1993*1, 1997*1; Lipe and Varien 1999*1:319). There are remains of buildings on the canyon rim, at the base of a 35- to 40-foot cliff that drops down from the rim, on a steep talus slope, and near the canyon bottom (Database Map 329 and Appendix A).

The basic architectural unit found across most of the site consists of a kiva and its associated surface rooms; such units often are referred to as "unit pueblos" (Prudden 1903*1), "Prudden units" (Lipe and Varien 1999*1:291), or "kiva suites" (Bradley 1992*2). A single unit pueblo is inferred to have been the residence of a household in the Pueblo II and III periods (Lipe and Varien 1999*1:291), and a household most likely consisted of "an extended family or other small co-residential group" (Lipe 1989*1:64). Household size is believed to have ranged from two to 12 individuals, with an average of five to eight (Lightfoot 1994*1:148).

The unit pueblos at Woods Canyon Pueblo are loosely scattered instead of being arranged in tightly clustered architectural blocks, as is the case at Sand Canyon Pueblo, a late Pueblo III village located about 10 km southeast of Woods Canyon Pueblo and investigated by Crow Canyon in the 1980s and 1990s (Bradley 1992*2). The configuration of Woods Canyon Pueblo may have been dictated largely by the steep terrain. On the basis of surface remains and the results of subsurface testing, we identified 50 kivas (10 definite, 24 probable, 16 possible). The number of surface rooms is difficult to estimate because of the effects of erosion; however, the presence of between 120 and 220 surface rooms is inferred from the number of kivas and the amount of wall rubble present (Lipe 1995*2:1). Other architectural remains at the site include 16 towers (14 probable, two possible); a D-shaped structure; several flat, open areas that might have been plazas; numerous checkdams; and extramural walls that might have served as terrace walls for gardens.

Architectural remains are clustered in four sections of the site that are defined roughly by topographic boundaries (see Database Map 334 and Table 4 in "Chronology" for an explanation of how these four sections relate to the seven numbered areas used by Churchill and Ortman in their evaluation of site chronology). The four sections are referred to throughout this report as (1) the lower west side (canyon bottom), (2) the upper west side, (3) the east side (east talus slope), and (4) the canyon rim (see Database Map 329).

Lower West Side (Canyon Bottom)

The lower west side is a relatively flat area at the base of the talus slope, near the canyon bottom and west of the drainage that bisects the site. In this publication, we often refer to this section as simply the "canyon bottom." The canyon bottom contains the highest concentration of kiva suites at the site, and the number of kiva depressions suggests that approximately 14 to 18 residences are present. Several towers are situated prominently on top of large boulders. Ashy midden deposits and two flat, open areas are also present. We tested four kivas (Structures 1-S, 2-S, 3-S, and 9-S) and two midden areas (Nonstructures 4-N and 5-N) in the canyon bottom. One of the open areas was also tested (Nonstructure 1-N). Low masonry walls in this area might have been built to create terraces for small gardens (see "Water Control and Subsistence"). The canyon bottom is believed to be the section of the site that was occupied earliest (see "Chronology").

Upper West Side

There are also clusters of buildings along the base of the cliff and on the steep talus slope west of the main drainage in a section of the site referred to as the "upper west side." Between seven and 12 residential units are present. Six of these residences were built on a relatively level bench at the base of the cliff. Also at the base of the cliff is a masonry room built under an overhang in the cliff face, as well as several wall remnants. On the talus slope, several flat areas upslope from natural drop-offs served as building sites for residences. Retaining walls were used to create additional flat areas where kivas were constructed. Two kivas (Structures 7-S and 8-S) and three midden areas (Nonstructures 3-N, 7-N, and 10-N) were tested in the upper west side. Occupation in this section of the site began later than in the canyon bottom (see "Chronology").

East Side (East Talus Slope)

The east side of the site consists of a steep talus slope east of the main drainage. Like their counterparts on the upper west side, kiva suites on the east talus slope were constructed in flat or nearly flat areas that had been manually leveled by adding dirt behind retaining walls. Approximately 11 to 18 residential units are present here. Several towers—including one that might have been square—were built near the base of the slope, adjacent to, and east of, the main drainage. We tested two kivas (Structures 4-S and 5-S) and two midden areas (Nonstructures 6-N and 11-N) in this section of the site. Like the upper west side, the east talus slope was occupied later than the canyon bottom (see "Chronology").

Canyon Rim

Most of the buildings and features on the canyon rim are clustered in an area called the "rim complex" (Database Map 263; also see Figure 1, an artist's reconstruction of the rim complex). The complex is located in the northeast portion of the site along a distinctive section of canyon rim and is bounded on the north by a masonry enclosing wall. This complex is the locus of public architecture at the site. In it we documented a D-shaped structure; several towers; a kiva; a flat, open area believed to have been a plaza; and several poorly defined masonry surface structures. A kiva (Structure 6-S) and the open area (Nonstructure 2-N) were tested in this section.

One of the towers in the rim complex spans a crevice in the cliff, which is believed to have provided access between the buildings of the rim complex and the structures at the base of the cliff; in addition, steps might have once connected the two (Database Photo 7482). For this reason, the buildings at the base of the cliff are considered part of the rim complex. These buildings are well preserved because a natural overhang in the cliff protects them from the elements—some walls stand two stories high, evidence of a degree of preservation typically associated with cliff dwellings. The buildings include single- and multiple-story rooms, a kiva, and what might be the lowermost story or level of the crevice-spanning tower mentioned above. No excavations were conducted in these buildings, but we did record the standing walls.

The buildings in the rim complex were constructed later than the buildings in the canyon bottom. The occupation of the rim complex was either contemporaneous with or slightly postdated the occupation of the upper west side and the east talus slope (see "Chronology"). If the latter is the case, the rim complex was the last part of the site to be occupied. There are also several probable towers on the canyon rim outside the rim complex near the west edge of the site. Finally, several checkdams are visible outside the rim complex in the main drainage above the canyon rim (Nonstructure 9-N).

Sequence of Construction

The architectural layout of the pueblo changed over time. The first construction, consisting primarily of a loose cluster of residences (kiva suites), took place in the canyon bottom in the mid–A.D. 1100s. In the early to middle 1200s, construction began on the upper west side, the east talus slope, and the canyon rim. Occupation in the canyon bottom appears to have been on the decline during this time. With the exception of the structures on the rim, almost all the buildings constructed in the mid-1200s were residences.

The rim complex proper might have been built slightly later than the buildings on the upper west side and east talus slope, and construction in this area apparently continued into the late A.D. 1270s. The architectural layout of the pueblo during the middle to late 1200s had changed from that of earlier years, with residences present along the slopes and at the base of the cliff, but not in the canyon bottom. Construction of the rim complex was probably the last major building effort at the pueblo.


Studies of architecture are important because architectural variation—that is, different kinds of buildings, different styles of construction, and differences in feature assemblages—has the potential to shed light on a variety of issues of broad anthropological concern, including social differentiation, rank, and status (Kuckelman 2000*2:par. 1). In this section, I discuss kivas, surface rooms, towers, and public architecture: where and how they were constructed, the types of features and artifacts they contain, and how they might have been used. The greater part of this chapter is given to the discussion of kivas because kivas were the only buildings that were actually excavated (tested) at Woods Canyon Pueblo. Our knowledge of rooms, towers, and public architecture at the site is based solely on surface indicators and spatial relationships.


The estimated number of kivas at Woods Canyon Pueblo is 50, nine of which were tested (Structures 1-S through 9-S). We also recorded observations about one unexcavated but partly exposed kiva (Structure 13-S). The following summary of previous research on kivas in the Mesa Verde region provides a general context for the evaluation of the kivas documented at Woods Canyon Pueblo.

Kivas in the Mesa Verde Region: Previous Research

Kivas in the Mesa Verde region are believed to have been used by households or other small, coresidential groups. Two arguments support this inference. First, if we assume that there is a regular relationship between the size of a kiva and the size of the group that used it, we can argue that Pueblo II and Pueblo III kivas, which are relatively small, could have accommodated only small groups of people (Lipe 1989*1:54). Second, the ratio of rooms to kivas serves as a proxy for the number of people that used a kiva suite: the higher the ratio, the greater the number of people. The average ratio of six to nine rooms per pit structure in the Pueblo I–Pueblo III periods (A.D. 850–1300) is significantly lower than the ratio of 26.7 rooms per kiva in the Pueblo IV period (A.D. 1300–1600) (Lipe 1989*1:Table 1). These ratios suggest that kiva suites were used by much smaller groups before A.D. 1300 (Lipe 1989*1:56).

In recent years, archaeologists have reevaluated the long-held assumption that prehistoric kivas were ceremonial chambers analogous to historic and modern-day Pueblo kivas. Archaeological evidence indicates that the kivas associated with Pueblo III unit pueblos were used primarily for domestic activities and some small-scale ritual activities (Adler 1989*1; Cater and Chenault 1988*1; Lekson 1988*1; Lightfoot 1992*1; Lipe 1989*1; Lipe and Varien 1999*1:291; Varien 1999*1). There is, however, evidence that some kivas may have had more specialized ritual use (Bradley 1993*1; Wilshusen 1989*2) or perhaps served as "elite," or "elaborated," residences (Lipe and Varien 1999*1:336; Ortman 1998*2).

Using cross-cultural data from nonranked societies, Adler (1989*1) distinguishes between low-level integrative facilities and high-level integrative facilities. Low-level integrative facilities serve small groups of people and are used in a variety of ways for both domestic and ritual purposes. High-level integrative facilities, on the other hand, serve larger groups of people and are specialized for ritual use. Adler argues that kivas in the Mesa Verde region were not specialized for ritual use. The average size of pre–A.D. 1300 kivas, and the estimated populations they served, fall within the size and population ranges that correlate with low-level facilities in his study.

Mealing bins, metates, hearths, domestic pottery, loom anchors, and lithic debris have been found in kivas in the Mesa Verde region, suggesting that domestic activities took place in them (Cater and Chenault 1988*1). Lekson (1988*1:213) argues that a small standard kiva is a domestic pit structure, comparable to earlier pithouses, and that there is architectural continuity between the pithouse and the kiva until A.D. 1300. He believes that not until the Pueblo IV period, when the number of kivas declined dramatically, did kivas change from domestic structures to community-level facilities (Lekson 1988*1).

Bradley (1993*1) infers that several kiva suites at Sand Canyon Pueblo were special nondomestic facilities. This inference is based on the presence of one or more of the following characteristics: lack of domestic features, limited storage space, a high labor investment in construction, limited or restricted access, and poor integration among structures or between structures and the adjacent open space. Wilshusen (1989*2) believes that the presence of certain features in pithouses and kivas—namely, altars, prayer-stick impressions, and floor vaults (roofed sipapus)—may signal more specialized ritual use in pit structures.

Features associated with specialized ritual activity have been documented in a kiva (Structure 1501) inside the D-shaped structure at Sand Canyon Pueblo. The kiva contained a floor vault, one of the most conspicuous ritual features in Mesa Verde kivas (Bradley and Churchill 1995*1; Ortman 1998*2:181; Wilshusen 1989*2). After the vault was intentionally filled, two pits were dug into the fill and lined with bottomless pottery mugs that may have been supports for portable altars (Wilshusen 1989*2). A mealing bin was also present in Structure 1501. The location of this kiva inside a public building and the combination of elaborate ritual features and domestic features suggest that this structure may have been an elaborated residence (Lipe and Varien 1999*1:336; Ortman 1998*2:181). The D-shaped structure contains many rooms, some of which are inferred to have been used for storage. Perhaps these facilities were used to store surplus goods, with the people living in the kivas inside these facilities controlling the goods (Lipe and Varien 1999*1:335–336).

Architectural Block 300 at Sand Canyon Pueblo also might have been a storage facility, considering that an estimated 30 rooms and only one kiva are present. The kiva in this room-dominated block may have been an elaborated residence. If people lived in the D-shaped structure and Block 300, and if they were able to control surplus supplies, they may have had higher rank or status than other residents of the pueblo. In a related study, Huber (1993*1) argues that a kiva suite in Block 100 at Sand Canyon Pueblo possesses the characteristics of a household of an influential leader. Namely, it is larger than average, it exhibits greater formality and reflects a greater investment of labor in terms of its construction, and it contained large pottery serving vessels.

Kivas at Woods Canyon Pueblo

All of the documented kivas at Woods Canyon Pueblo are associated with contiguous, aboveground rooms that form "roomblocks"—that is, they are part of the residential architecture at the site. Table 1 compares the location, architectural characteristics, floor features, and floor artifacts of the nine tested kivas and the one kiva whose wall was partly exposed on the modern ground surface (Structure 13-S). Unfortunately, the sizes of the kivas could not be compared, because the area of each kiva exposed during testing was too small to allow us to estimate total size. Furthermore, we could compute the ratio of rooms to kivas for only one kiva suite. Counting the number of rooms associated with the remainder of the tested kivas was not possible, because room outlines could not be distinguished in the roomblock rubble. On the basis of the available data, it appears that the majority of the tested kivas (Structures 1-S, 2-S, 3-S, 4-S, 5-S, 7-S, 8-S, and 9-S) were used for residential purposes and household-level rituals. Limited evidence hints at the possibility that two of the documented kivas (Structures 6-S and 13-S) might have functioned differently and may have been used as elaborated residences or for specialized rituals.

Kiva Location

The majority of kivas at Woods Canyon Pueblo are located in clusters of residential structures. However, two kivas (Structures 6-S and 13-S) are located in, or are connected to, the rim complex, the only known public area at the site. Structure 6-S is located in the canyon rim portion of the rim complex, and Structure 13-S is located at the base of the cliff, in the lower portion of the rim complex. Their distinctive locations in the most public part of the village may indicate that they were used differently than other kivas at the site.

Structure 6-S is associated with a large tower (Structure 28-S) and several possible rooms, including Structure 29-S. Secondary refuse (Nonstructure 2.4-N) identified adjacent to this architectural unit might have been discarded during activities that took place there. The layout of these buildings differs slightly from the layout of a typical kiva suite: the tower is immediately north of the kiva, and the possible rooms are northwest of the kiva and west of the tower. However, the configuration of these buildings may have been partly dictated by the presence of a large boulder west of the kiva.

The kiva at the base of the cliff (Structure 13-S) is spatially associated with approximately 13 rooms and a possible tower. Structure 13-S and its associated rooms and tower constitute a kiva suite. The ratio of rooms to kivas is slightly higher than the typical room-to-kiva ratio in the Mesa Verde region, which is approximately 10 or 12 to one (Bradley 1992*2). Nonetheless, this kiva suite does not qualify as a room-dominated block—that is, a block with more than 20 rooms per kiva—as defined by Bradley (1992*2:81). Access to this kiva suite is limited. It was built on a ledge at the base of a cliff that is surrounded by a steep and treacherous slope. No excavations were conducted in this building, but we did record standing walls and visible features.

Kiva Architecture

Little architectural variation was observed in the 10 documented kivas at Woods Canyon Pueblo (Table 1), although it is important to note that relatively little architecture was revealed during Crow Canyon's investigations. Nonetheless, in Structure 13-S, a short segment of wall visible on the modern ground surface provided evidence that this kiva was constructed differently than the two other Woods Canyon kivas whose upper lining walls were exposed. The masonry upper lining wall of Structure 13-S rests on top of pilasters and beams that span pilaster to pilaster. This method of construction produced large roofed niches or shelves below the wall and between the pilasters. The upper lining wall appears to have supported the roof. This kiva is similar to Kiva E at Mug House at Mesa Verde National Park (Rohn 1971*1:78–79), but is unlike most other documented kivas in the Mesa Verde region.

The other two kivas whose upper lining walls were exposed at Woods Canyon Pueblo (Structures 1-S and 9-S) have earthen, rather than masonry, walls (though their bench faces are masonry). The differences in the construction of upper lining walls observed at Woods Canyon probably relate to chronology: earthen walls are believed to have been an earlier construction style (Brew 1946*1; Lipe and Varien 1999*2:262; Smith 1998*1; Varien 1999*2), and the two kivas at Woods Canyon with this type of upper wall are located in the canyon bottom, which was the earliest-occupied part of the site. Structure 13-S, on the other hand, has a masonry upper lining wall and is located at the base of the cliff in the rim complex, an area of the site that was constructed and occupied later.

Of the masonry architecture exposed during excavation, the amount of surface treatment of stone was almost identical except for the bench face in Structure 1-S, which exhibited less pecking. The size and shape of stones used for benches and pilasters was uniform from kiva to kiva. Therefore, it does not appear that the construction of any one kiva at Woods Canyon required the expenditure of more effort than was afforded the others, at least in the sample of 10 structures considered here.

Kiva Floor Features and Floor Artifacts

A comparison of Woods Canyon kivas does not reveal distinct differences in their assemblages of floor features and floor artifacts. The most common floor features in the kivas that were tested were small pits of unknown function. Hearths were found in four kivas (Structures 2-S, 3-S, 6-S, and 8-S), and indirect evidence of hearths (fire-reddening or ash accumulations on the floor) was found in two more (Structures 5-S and 7-S). Hearths were used for heating, lighting, and cooking, and are standard features of kivas.

Evidence of ritual activity includes the presence of a sipapu with possible prayer-stick impressions (paho marks) in Structure 2-S. There are also possible sipapus in two other kivas, Structures 3-S and 7-S. These features are classified as simple or complex sipapus, neither of which is as elaborate as a roofed sipapu or roofed floor vault. The presence of simple or complex sipapus is believed to indicate structure use by a lineage, clan, or kiva group, rather than use by a community (Wilshusen 1989*2). Hence, the presence of these sipapus argues for small-scale ritual use of the Woods Canyon kivas in which they were found.

Limited amounts of de facto refuse were left on the floors of six of the nine tested kivas (Structures 1-S, 3-S, 4-S, 5-S, 6-S, and 7-S). These assemblages were dominated by items—such as grinding implements, pottery vessels, bone tools, and chipped-stone tools—used for everyday domestic activities. The presence of these items suggests that domestic activities took place inside these six kivas during their final days of use. Because the same types of artifacts are found in the associated trash deposits, I infer that domestic activities routinely occurred in and around these kivas. The final use of one kiva (Structure 5-S) was as a burial chamber in which at least 10 individuals were interred (see "Human Skeletal Remains").

In conclusion, I infer that most of the tested kivas at Woods Canyon Pueblo were used for domestic activities and household-level rituals. The locations of Structures 6-S and 13-S in a public area, and perhaps the unusual construction of Structure 13-S, suggest special use. These kivas might have been residences within public areas, or they might have been used for ritual activities. Their association with surface rooms, their small size, and the presence of domestic features and domestic floor artifacts suggest that residential activities occurred in them, and it is likely that they served as elaborated residences rather than as ritual facilities.

Surface Rooms

We recorded rooms with walls visible above modern ground surface in the lower portion of the rim complex (Table 2). There are 13 rooms in this area, and they are spatially associated with one kiva (Structure 13-S). The room-to-kiva ratio is at the upper limit of the average range but does not fall within the range required for interpretation as a room-dominated block as defined by Bradley (1992*2).

It has been hypothesized that two architectural complexes containing many rooms and few kivas at Sand Canyon Pueblo (Block 300 and the D-shaped structure) were centralized storage facilities controlled by the residents living in the kivas (Lipe and Varien 1999*1:335–336). At Woods Canyon Pueblo, we evaluated the rooms associated with Structure 13-S for size, presence or absence of sooting, and presence or absence of doorways in an effort to determine whether the rooms in this kiva suite might have been storage facilities (Table 2). We assumed that storage rooms would be smaller than living rooms. Sooting on walls or ceilings suggests that a hearth was present; the presence of a hearth, which would have been used for cooking, heating, and lighting, is probably an indicator of living space. Finally, the presence of doorways suggests regular traffic in and out of a room—a condition that is more likely to exist in living rooms than in storage rooms.

The size of the rooms in the lower portion of the rim complex at Woods Canyon Pueblo ranges from 3.24 to 8.17 m2, with the average room size being 4.76 m2. This average is larger than the average size of completely excavated rooms at Sand Canyon Pueblo, which is 4.25 m2 (Bradley 1992*2:81). If we omit from calculations the largest room recorded in the lower rim complex at Woods Canyon—a long, narrow alcove room—then the average area is 4.45 m2, which is still slightly larger than the average size of rooms at Sand Canyon Pueblo. Thus, it can be concluded that size alone does not support the hypothesis that these are storage rooms.

The presence of sooting on the ceiling/overhang of one room (Structures 19-S), on a wall in another room (Structure 25-S), and on the ceiling and one wall of a third room (Structure 22-S) suggests that these three rooms contained hearths and therefore were used as living spaces. However, one of the rooms with a hearth, Structure 25-S, is very small and has a low ceiling, both characteristics typically associated with storage rooms rather than with living quarters. This makes me less certain of the use of Structure 25-S. The absence of sooting in the exposed portions of the nine remaining rooms suggests that they were more likely used for storage than for living space.

Four doorways were preserved in the rooms in the lower rim complex. In one case, the doorway provided passage between two ground-story rooms (Structures 23-S and 25-S). Doorways are also present in an alcove room (Structure 22-S) and in a second-story room (Structure 21-S). In both cases, the doorways provided access between rooms and a probable courtyard on top of the kiva roof (Structure 13-S). The fourth doorway is located in a ground-story room (Structure 11-S), and it provided access either to the outside or to another room that has since fallen off the cliff edge.

Following Bradley's (1993*1) approach to spatial analysis within kiva suites at Sand Canyon Pueblo, I suspect that access via the doorways at Woods Canyon Pueblo was neither completely unrestricted nor completely restricted. A more precise assessment is impossible because, inevitably, some doorways and roof hatchways were destroyed when the roofs and parts of the walls collapsed. The presence of a doorway does not preclude use as a storage facility, especially if the room was built in a small alcove where entry from above would have been impossible. The doorway data for Woods Canyon Pueblo are difficult to interpret. There does appear to be a correlation between sooting and doorways, with two of the three rooms in which sooting was observed also having doorways (Structures 22-S and 25-S).

In summary, there are at least two (Structures 19-S and 22-S), and possibly three (Structure 25-S), living rooms in the lower portion of the rim complex. The presence of sooting, more than any other characteristic, was used to make this inference. The number of storage rooms is more difficult to estimate. Structure 12-S clearly is a storage room because it is quite small and is located on a ledge above the other rooms. Access to this room was restricted because of its location. Structure 26-S is also inferred to be a storage room because of its small size and the absence of sooting. The probable use of the remaining eight rooms is unknown. The data—in particular the lack of sooting—suggest, but do not prove, that most of the rooms were used for storage and that therefore this kiva suite could have been a centralized storage area. The presence of living rooms indicates that people also inhabited this part of the village and might have controlled supplies in the storage rooms.


Because actual tower walls are not visible on the modern ground surface at Woods Canyon Pueblo and no towers were excavated, our discussion of this type of architecture is limited to other evidence observed on the ground surface. Four towers, all defined by mounds of rubble, are thought to be contained within the upper portion of the rim complex, that is, the area bounded by the enclosing wall. There is also evidence of additional towers on the canyon rim west of the rim complex, near the base of the east talus slope, and in the canyon bottom. Kivas and rubble mounds associated with roomblocks are present close to the towers. It is not clear whether these towers were associated with residences or functioned as public architecture. Many of the towers were built on large boulders. Boulders are a common location for towers at Hovenweep National Monument (Thompson 1993*1) and in other parts of the Mesa Verde region (Kelley 1996*1). Kelley (1996*1:109) states that boulder towers are associated with late Pueblo III villages. If the boulder towers at Woods Canyon Pueblo were built in the A.D. 1200s, then they might have been the last buildings constructed in the canyon bottom, in which case they were contemporaneous with the rim complex instead of with the residences in the canyon bottom. Alternatively, the towers may have been contemporaneous with the residences in the canyon bottom, which would mean they were present earlier in the site's history.

Public Architecture

"Public architecture" is defined as structures and features constructed and used by more than one household (Churchill et al. 1998*1). Archaeologists often use size and form to classify a building or other space as a public facility. The vagueness of the term—which sometimes is used as a catch-all category for any building or space that does not "look" residential—and the fact that there have been only limited excavations of public architecture in the Mesa Verde region limit our understanding of these facilities. Like other researchers (e.g., Adler and Wilshusen 1990*1), I believe that the activities that occurred within public facilities helped to integrate people at some level. It is not clear, however, whether all inhabitants had access to all public buildings and spaces (Lipe and Ortman 2000*1:95).

The public architecture at Woods Canyon Pueblo is clustered in the rim complex and includes a D-shaped structure, four towers, and a probable plaza (Database Map 263). This space is delineated by an enclosing wall. The clustering of public structures and the presence of enclosing walls are characteristic of other canyon-rim villages that date from the late Pueblo III period (Lipe and Ortman 2000*1:107–108).

The wall that encloses the rim complex is made of large, tabular or blocky pieces of unshaped sandstone. The wall is uncoursed to semicoursed and two stones ("double-stone") wide. Its existing height is 84 cm. Using the rubble height recorded by Kenzle (1993*1:97), I calculate that this wall probably originally stood about 1.5 m tall. In her study of enclosing walls in the Mesa Verde region, Kenzle (1997*1:207) postulates that such walls "functioned as sociophysical boundaries erected for the purposes of spatial demarcation and social regulation." She further argues that most walls were not formidable obstacles to opponents, but would have provided cover for defenders (Kenzle 1993*1:96). The wall at Woods Canyon Pueblo encloses only part of the village; however, the space and buildings that it encloses constitute the most prominent and distinctive architecture at the site. For this reason, I believe that the wall probably served as an important sociophysical boundary. It is unclear whether it was built for defensive purposes as well. If it was, perhaps the rim complex was seen as a part of the village that warranted some protection.

Most of the area inside the enclosing wall is flat, open space. It is likely that this area was used as a plaza for public gatherings such as dances, feasts, and ceremonies. We were unable to identify a plaza surface; if one existed originally, it is likely that it has since eroded away. If this space was used as a plaza, I speculate that the activities that took place there were probably more inclusive than exclusive in nature because they occurred in an open, highly visible area. If this was so, the plaza probably served as an integrative facility for the inhabitants of the pueblo. The plaza is smaller than the plaza at Sand Canyon Pueblo (Bradley 1992*2) but larger than the plaza at Castle Rock (Kuckelman 2000*1). It may not have been large enough to accommodate all the inhabitants of the pueblo; on the other hand, the village population may have been smaller when the plaza was in use in the late A.D. 1200s than it was earlier in the occupation of the site.

The building known as the D-shaped structure (Structure 17-S) is located in the rim complex. The curved wall of this building is visible, but what is assumed to have been the straight south wall is missing and probably fell over the edge of the cliff (thus, the description of this structure as D-shaped is largely inferential). D-shaped structures were built during the late Pueblo III period in the Mesa Verde region (Churchill et al. 1998*1). The D-shaped structure at Woods Canyon is smaller than some other similar structures documented in the Mesa Verde region, including Sun Temple at Mesa Verde National Park (Fewkes 1916*1) and the D-shaped structure at Sand Canyon Pueblo (Bradley and Churchill 1994*1, 1995*1). It is comparable in size to the D-shaped structure at Cannonball Ruins (Lipe 1996*1) and to the bi-wall structure at Horseshoe Ruin at Hovenweep National Monument (Winter 1975*1).

The internal space of the D-shaped structure at Woods Canyon Pueblo is divided by an east-west wall. The area south of this dividing wall is further subdivided into at least two spaces, including one known room (Structure 16-S). This room was added after the dividing wall was constructed, and it is of particular interest because it has characteristics that are atypical of masonry rooms. For example, the interior faces of the stones in the walls are pecked, and three wall niches are present. Two doorways, one of which is T-shaped, provided access to the room. At some point, the north doorway was plugged and a niche was created. This is similar to the remodeling of an exterior doorway in the D-shaped structure at Sand Canyon Pueblo (Bradley and Churchill 1994*1:3–6). The location of Structure 16-S inside a D-shaped structure, the presence of niches, and the surface treatment of the interior wall faces all suggest that this room might have been used for some special purpose.

Four towers were identified in the rim complex, two of which are of particular interest. The largest tower (Structure 29-S) is spatially associated with other buildings, including a kiva (Structure 6-S) and several possible surface rooms (see paragraph 25). Structure 18-S, the other noteworthy tower, is adjacent to the D-shaped structure. It spans a crevice in the cliff face, and the walls probably extended down into the crevice. It most likely provided access between the rim area and the structures at the base of the cliff. A segment of a curved wall found at the base of the cliff (Structure 14-S) might be the remains of the bottom story of this tower. Similar circular structures built around the tops of cracks in cliff faces have been recorded at other aggregated sites (Varien et al. 1996*1:90) and are referred to by Thompson et al. (1997*1:154) as "architecture with unknown function."

Nonarchitectural data also provide some insight into the possible uses of the rim complex. Although intrasite analyses did not reveal significant differences among the four main sections of the site in terms of their faunal assemblages, there are some differences in the archaeobotanical and artifact assemblages that suggest that the rim complex might have been used differently than other sections of the site.

In "Archaeobotanical Remains," Rainey and Jezik note that, although the two most common food plants at the site (corn and cheno-ams) were found in the rim complex, few other food plant remains were recovered and none were found in the hearth of Structure 6-S, the tested kiva. On the other hand, they document a greater variety of fuelwood in samples from this part of the site than might be expected by chance. Although sample size and differential preservation cannot be entirely dismissed as potential contributing factors, one possible explanation of the observed pattern is that the rim complex was used for public functions involving (1) the communal consumption of food prepared elsewhere at the village and (2) the burning of a wider variety of woods than in other sections of the site. Interestingly, despite the suggestion that foods were not routinely cooked in the rim complex (or at least not in the tested kiva), the relative abundance of ground-stone tools and peckingstones in this part of the site may indicate that corn was processed into meal there, perhaps for ceremonial feasting (see "Artifacts"). In any case, the archaeobotanical data suggest that the rim complex might have been used differently than the other three sections of the site, which lends support to, but does not prove, the hypothesis that the rim complex was a special-use public area.

In "Artifacts," Ortman compares pottery sherd and rim-arc data for the early Pueblo III (canyon bottom) and late Pueblo III (rim complex, upper west side, and east talus slope) components at Woods Canyon Pueblo to see whether there were differences in the types and/or sizes of vessels used through time. He found that there is a higher percentage of corrugated gray jar sherds in assemblages associated with the late Pueblo III component, which might reflect an increased use of corrugated jars for cooking. Furthermore, the corrugated jar rim sherds from contexts dating from this time were larger than those dating from other times, suggesting that more large jars were used during the late Pueblo III occupation. Ortman concludes that both patterns suggest that food preparation on a larger scale intensified during the late Pueblo III occupation, which might reflect an increase in communal feasting activities.

A related study by Ortman (see "Artifacts") of white ware bowls from Woods Canyon Pueblo found that the exterior surfaces of large serving bowls were decorated more often than were the exteriors of small bowls. Furthermore, the exterior surfaces of bowls dating from the late Pueblo III occupation were more likely to be decorated than were the exterior surfaces of bowls dating from the early Pueblo III occupation. These data are consistent with findings at other late Pueblo III sites in the central Mesa Verde region. Ortman argues that decorating the exteriors of large bowls may correlate with increased communal feasting in public spaces.

It is interesting that there does not appear to have been any public architecture at Woods Canyon Pueblo until the rim complex was built. The rim complex was constructed during the late Pueblo III occupation of the site. Some structures in the lower portion of the rim complex were built as late as the A.D. 1270s. Although the types of white ware pottery found do not preclude construction in the rim complex as early as A.D. 1225, white ware attribute data and tree-ring dates suggest that construction probably did not start until slightly later, sometime in the A.D. 1250s (see "Chronology"). If occupation of the rim complex began in the 1250s and ended in the 1270s, then this part of the site was occupied for no more than 30 years.

This raises the question of whether there were enough people living at the pueblo during the A.D. 1250–1280 period to warrant the construction of a public facility. There are not sufficient data for us to determine whether the occupation on the upper west side and east talus slope was contemporaneous with, or slightly earlier than, the occupation of the rim complex. Basically, if there were only a few people living at the pueblo between A.D. 1250 and 1280, there was no need for a public facility. It is possible that the rim complex was less of a public facility then we have assumed; perhaps it was simply the final residential area at the site and was occupied by two households living in the upper and lower kiva suites.

I would argue, however, that the construction of an enclosing wall, a plaza, several towers, and a D-shaped building would have been very labor intensive and would have been unnecessary if people were using the rim complex simply as a final residential area. Perhaps the nonresidential architecture was constructed earlier in the village's history, with the remaining households moving into the rim complex later, in the late A.D. 1200s. If so, it suggests that the public elements of the rim complex were built at a time when the population of the village was large enough to support a public facility.

We know for certain that the rim complex was not in use during the early Pueblo III occupation of the site because it had not yet been built. We also know that the occupation of the rim complex was short and occurred late in the history of the site. What we do not know is the number of people living at the pueblo when the rim complex was in use. I argue that people were residing in the upper west side and east talus slope areas in the A.D. 1250–1280 period and probably were using the rim complex. Our limited testing did not provide sufficient data for us to evaluate in detail how the rim complex was used. However, there is architectural evidence of both nonresidential and residential buildings in the rim complex. I conclude that the rim complex was not exclusively a public facility, nor was it solely a small residential area. The current evidence suggests that it was used for both public and private purposes. It seems plausible that the people living in the two rim-complex households oversaw or controlled the public facility in some way. Further testing at the rim complex and at similar canyon-rim complexes at other villages in the Mesa Verde region is needed to address the important questions posed by Lipe on the Woods Canyon Pueblo National Register form:

Do these represent the residential compounds of the community leadership, as well as places for community assembly and ceremony? . . . Or are the late 13th century complexes of towers and other features truly community property, i.e., non-residential structures that were erected cooperatively to impress outsiders, to provide facilities for community ceremonies, and perhaps to serve as defensive retreats? [Lipe 1995*2:13–14]


The configuration of Woods Canyon Pueblo changed throughout its history. The earliest construction was centered in the canyon bottom and consisted of a dense concentration of residences (kiva suites). Later, residences were constructed on the east and west talus slopes, the base of the cliff, and the canyon rim, which shifted the core area of occupation away from the canyon bottom.

The architecture below the canyon rim is dominated by household residences, but the architecture on the canyon rim is primarily "public." The rim complex was constructed during the late Pueblo III occupation of the village, probably starting in the A.D. 1250s. Two kiva suites in the rim complex may have had special use, either as elaborated residences or as places for specialized rituals. If these two kiva suites represent elaborated residences, the inhabitants may have had special rank or status.

References cited | To borrow, cite, or request permission

Copyright © 2002 by Crow Canyon Archaeological Center. All rights reserved.