Human Skeletal Remains

by Cynthia S. Bradley


In keeping with the Crow Canyon Archaeological Center's policy with regard to the treatment of human remains (see the on-line field manual), no deliberate attempt was made to locate human bone at Woods Canyon Pueblo. Even so, 24 isolated bones and the remains of 11 discrete individuals were found during test excavations. The latter were assigned "Human Remains Occurrence" (HRO) numbers, in accordance with the protocol outlined in Crow Canyon's field manual. Excavation of human remains was limited to those bones revealed in the original excavation units; no units were expanded or added for the purpose of uncovering additional skeletal material. Of the individuals (HROs 1–11) found, six were almost completely exposed and five were partly exposed. Their bones were analyzed in situ to minimize disturbance. Artificial cranial deformation, associated artifacts, and mortuary locations indicate a prehistoric Pueblo cultural affiliation.

Ten individuals (HROs 1–10) were on the floor of a kiva (Structure 5-S), within a 2-x-1-m test trench (see Database Map 318 and Table 1). Each skeleton in this multiple interment was in articulation and, although lying close to others, was generally discrete. All were in very good condition except for postmortem compression fracturing of the crania. One individual (HRO 11) was in a shallow pit-grave cut into a talus slope; this child was discovered in the corner of a 1-x-1-m test unit located southwest of Structure 5-S (Database Map 294). This incomplete skeleton was articulated but in very poor condition; the bones were either significantly damaged or destroyed by alkaline soil conditions and root growth.

Despite the small total number and limited exposure of the remains found at Woods Canyon Pueblo, the mass burial in Structure 5-S presents a circumstance seldom seen in archaeological sites. Often, as with midden burials, the relationship between burials is unclear, and it is impossible to determine whether the individuals were alive at the same time. In contrast, the shared mortuary circumstance of the 10 individuals in Structure 5-S indicates that these people were contemporaries whose deaths were fairly concurrent. This tight spatial and chronological control provides us an unusual opportunity to address the following questions:

  • Did these individuals experience simultaneous physical stress?

  • Can the cause or causes of death be determined? Can all the deaths be attributed to the same cause, or did different people die of different causes?

  • Does the skeletal evidence add insight into the circumstances of burial?

  • Finally, could these individuals have been members of the same family?

Commonly, the purpose of human skeletal analysis is to gain information from a representative sample of individuals in order to shed light on how the larger group experienced and responded to environmental and social circumstances. However, I did not use this approach here because there is no way of knowing whether the Woods Canyon remains are representative of the health and/or mortuary circumstances of all the people who lived at the village. The sample of individuals is small, and the data are further limited by the generally incomplete excavation of the skeletal remains. For these reasons, the primary purpose of this chapter is to describe the health status and unusual mortuary circumstances of these particular people. Some basic comparisons with other ancient Pueblo groups are made in order to place the Woods Canyon skeletal material in a broader regional context.

Twenty-four isolated bones were also found in locations across the village (Table 2). These remains, not recognized as human in the field, were analyzed in the Crow Canyon laboratory. The condition of the bones ranged from poor to very good, with much of the damage resulting from root growth and erosion. These bones showed no evidence of antemortem (before death) or perimortem (at or about the time of death) fractures, nor any evidence of anemia or infection. Age was estimated on the basis of developmental characteristics, but it was not possible to determine the sex of any of the individuals represented. Conservatively, a minimum of one infant and two adults are represented by these isolated bones.

The isolated bones were found in the postoccupational fill of kivas or in nonstructural areas such as middens. More than half of the isolated bones were in a dense midden (Nonstructure 10-N) in the upper west side of the site and are probably from the same individual, an infant. These remains were not articulated, nor did they come from one discrete layer of fill. The isolated bones are not addressed any further in this report other than to be included in estimates of the minimum number of individuals found during test excavations.

In addition, the superior cranium of an adult, probably male, was found just beyond the limits of the test trench in Structure 5-S. This skull was exposed when a metate within the test trench was removed from the south profile; because it was outside the test unit, it was not excavated further. However, the presence of the cranium shows that at least one more person was part of this multiple interment, and this individual is included in the estimate of minimum number of individuals.


Standard osteological data were collected according to the criteria set forth in Standards for Data Collection from Human Skeletal Remains (Buikstra and Ubelaker 1994*1). The information recorded included skeletal measurements and inventories, as well as assessments of age, sex, dentition, pathologies, trauma, and nonmetric traits. Additional sources are cited throughout this chapter. Bureau of Land Management (BLM) guidelines, derived from the Native American Graves Protection and Repatriation Act, require an assessment of cultural affiliation. Also, when the evidence allows, the analysis of human remains attempts to address possible causes of death.

I conducted the skeletal inventory, metric analysis, and taphonomic analyses; Sali Underwood, a consulting bioarchaeologist, analyzed dentition and nonmetric traits. Together, we assessed age, sex, and pathological conditions. The skeletons were excavated with bamboo tools and brushes; dental picks were reserved for cleaning teeth. In accordance with BLM permit stipulations and Crow Canyon's human remains policy, all observations were made in the field with minimal lifting and washing of the bones; however, we did lift and/or clean bones, such as mandibles, that required closer inspection for reliable observation. For the most part, the bones were not cleaned with water, although surfaces where pathological conditions commonly occur—for example, teeth, cranial vaults, and eye orbits—were washed thoroughly. I am confident that we identified most osteological conditions; any that were overlooked would have been mild and insignificant. All skeletons were mapped, photographed, and recorded on Crow Canyon's human remains occurrence forms as well as on forms provided in Buikstra and Ubelaker's (1994*1) Standards; both sets of forms are archived at the Anasazi Heritage Center in Dolores, Colorado, and at Crow Canyon.

Results and Discussion

In this chapter, baseline mortuary and biological data are presented and interpreted. This information is summarized in the appropriate data tables, which provide information on the skeletal assessments by individual. In cases in which attributes examined may be unfamiliar to the reader, brief explanations of the characteristics precede the presentation of the Woods Canyon Pueblo data. Skeletal evidence of age, sex, and dental and pathological conditions follows the criteria given in Buikstra and Ubelaker (1994*1).

Although information on each skeletal indicator was garnered to the fullest extent possible, not all possible analyses were included in this study. For example, dental defects were evaluated as indicators of health, not in terms of their frequency per tooth type. Likewise, skeletal measurements were taken according to the guidelines in Buikstra and Ubelaker (1994*1), but they are not used in this report except as needed for health assessments. Assessments of health often could not be made, however, usually because the skeleton was poorly preserved or only partly excavated. This circumstance, combined with the small number of individuals in the sample, precluded statistical analyses of the skeletal remains. Comparisons are limited to frequency analyses and are intended to provide a general evaluation of a condition or circumstance.

Age, Sex, and Demography

Among all 11 Woods Canyon individuals, there were seven subadults (birth to 20 years of age) and four adults (20 years and older). In Table 3, these remains are further categorized according to age and sex categories as defined in Buikstra and Ubelaker (1994*1). Subadult age estimates are based on Ubelaker's (1989*1) sequence for dental development of American Indians. We used skeletal development criteria to refine dental age estimates and to estimate age in cases in which the dentition was unobservable (Bass 1987*1; Buikstra and Ubelaker 1994*1; Krogman and Iscan 1986*1). For example, the age at death for one child whose dentition suggested 11 years ± 30 months could be refined on the basis of pelvic (acetabulum) development (Bass 1987*1:187) to 11 to 12 years.

Adult age estimates were based on skeletal development; multivariate analysis could be applied only in the case of the adult female. For three individuals (HROs 4, 6, and 7), only the foot bones were exposed. One of these (HRO 7) was represented by only a few foot phalanges and, beyond the determination of adult status, was not further evaluated for age. The other two individuals (HROs 4 and 6) are assigned to the "young adult" age category on the basis of degree of epiphyseal union of the phalanges, which in both individuals is complete; the absence of arthritic growths and the generally youthful appearance—as seen in rounded calcaneus and talus facet rims—also support this age estimate. Epiphyseal union of foot phalanges occurs at about 16 years of age in females. The calcaneus of HRO 6 is small and gracile, suggesting that this person might have been female. So, although this individual is categorized as a young adult, the size and youthful appearance of the foot bones could indicate a slightly younger age, perhaps late teens.

Gender was assessed only for those individuals classified as adults; no attempt was made to determined the gender of subadults, because the skeletal remains of people in this age group do not have distinctive sex characteristics. The adults include a female (HRO 3), a possible female (HRO 6), and two individuals for whom sex could not be determined (HROs 4 and 7). The person on the floor just beyond the limits of the test trench is most likely male, as determined on the basis of brow ridge development. This indicates that at least one mature male could be in the unexcavated part of the kiva.

Mortality is the best indicator of an individual's inability to overcome physiological stress. Under normal conditions, nonantibiotic societies have a predictable mortality pattern, as exemplified in the human remains from Grasshopper Pueblo (Hinkes 1983*1). Grasshopper Pueblo is a large Pueblo IV village in east-central Arizona that was occupied from about A.D. 1275 to 1400. The site has a large, well-documented skeletal population, and its inhabitants might have faced stressors similar to those faced by the residents of Woods Canyon Pueblo. The human remains data for Grasshopper Pueblo are compared with our limited data for Woods Canyon in Figure 1. The mortality pattern documented for Grasshopper is characterized by a high death rate for infants, declining rates during childhood and adolescence (with the rate reaching its lowest point in the mid-teen years), then a slowly rising rate throughout adulthood.

In contrast, Figure 1 shows that the small sample of individuals from Woods Canyon Pueblo is characterized by an unusually low infant mortality rate. This may be partly the result of sampling bias; no effort was made to obtain a representative sample of human remains from the site. Figure 2 illustrates the mortality pattern when only those individuals from Structure 5-S are included, and again, the infant death rate is very low. On the other hand, the 40 percent child death rate is somewhat high when compared with Grasshopper Pueblo, although such death rates do occur in modern nonantibiotic societies. The affected children in such groups generally are young, however—about three to five years of age. Once they reach the age of five to seven, children in nonantibiotic societies tend to survive until adulthood, accounting for the fact that the remains of older children and adolescents are found relatively infrequently in archaeological sites.

At Woods Canyon Pueblo, three of the five individuals classified as "children" are older than seven years, and at least one could be an adolescent. The adult skeletons are notable in that three of them are most likely to have been young adults or—if they were females—possibly even older adolescents. This age group tends to be resilient except for mortality associated with complications of childbirth. Although at least two individuals are most likely female, it is unlikely that they both died during childbirth at nearly the same time.

The incomplete excavation of skeletal remains complicated our osteological assessments. In the case of Structure 5-S, it affected our understanding of demography as well. Several circumstances strongly suggest that there are more burials in this kiva, but outside the area of excavation; these include the tightly packed and sometimes overlapping placement of remains, as well as the discovery of the cranium of at least one more person just beyond the limits of the test trench. Whether the remains of many additional people are present or not, the possibility that there are certainly reduces my confidence in the demographic assessments for Structure 5-S.

The total maximum population of the village—that is, the total number of people who lived at the site over the entire occupation—is estimated to have been about 200 to 350 people (see "Population Estimates"). The minimum number of individuals represented in the skeletal remains found during test excavations is 15, which is 4 to 7 percent of this total (the minimum number of individuals was estimated on the basis of number of discrete burials, as well as numbers of people represented by isolated bones, exclusive of teeth and phalanges, which can be lost during life). It should be remembered, however, that the momentary population of the village—that is, the actual population at any given point in time—would have been lower, probably closer to 70 to 250 (see "Population Estimates").

Mortuary Context

Body Positioning and Grave Goods

The positioning and location of human remains and their associated grave offerings constitute a "mortuary context," and this context provides valuable insights into the burial practices of a group (Bradley 1998*1). According to Turner and Turner (1999*1:44), conventional, or formal, burials in ancient Pueblo sites consist of remains that "are almost always fully articulated and were placed in regardful locations and positions with mortuary offerings and other indications of consideration and care for the deceased." Although even formal burials show some variation, an individual's positioning has an arranged, deliberate appearance, indicating an attempt to conform to a cultural standard.

During the Pueblo III period, individuals in the Mesa Verde region usually were placed in semiflexed, flexed, or, less commonly, extended positions (Turner and Turner 1999*1). A flexed position resembles a fetal position, whereas a semiflexed position consists of legs together, knees drawn up, and arms extended next to, or folded across, the trunk. Persons may be placed either on their sides or their backs. In an extended position, both legs are straight, possibly crossed at the ankles, and the arms either are folded across, or extend alongside, the trunk. See Ubelaker (1989*1) for further discussion of these positions and their normal variations.

Also during this period, people in the Mesa Verde region were buried in a variety of locations. Cemeteries in the Western sense—that is, numerous individuals interred some distance from a habitation area—are virtually absent. Instead, burials are typically found at habitation sites in places that offer protection from disturbance by carnivores and weathering (Lightfoot and Kuckelman 1994*2). The most common places are middens, cliff crevices, and under the floors or in fill deposits of abandoned ground-level rooms (Turner and Turner 1999*1). In both ancient and historic times, the bodies of Pueblo children were sometimes buried in pits beneath the floors of rooms that continued to be used for other activities; they were also placed on floors in rooms that were then abandoned (Hinkes 1983*1). Sepulcher burial, in which individuals were placed on the floors of intact rooms which were then sealed (doorways were plugged and/or hatch covers put in place), have been discovered at Pueblo III sites in the northern San Juan region (Bradley 1988*1; Morris 1924*1; Nordenskiöld 1979*1). Formal burials have also been found in kiva fill (Cattanach 1980*1:143, Table 6; Martin et al. 1995*1). Burial in trashy fill indicates that an abandoned kiva was being used as a midden; interment in natural fill indicates that the kiva served as a mortuary location following its active use.

Kiva floors, however, are very unusual burial locations (Bradley 1998*1; Turner and Turner 1999*1). Human remains on kiva floors often show signs of perimortem trauma, and their positioning is informal, usually sprawled and often face down (see Turner and Turner 1999*1). Human remains found in these circumstances have been associated with the abandonment of the structure (Larralde 1998*1) and/or emigration from the village (Bradley 1998*1; Lipe et al. 1999*1). Although such human remains are often referred to as "burials," they generally are not believed to represent common mortuary practice.

Mortuary offerings, such as pottery vessels, are often found in formal ancient Pueblo burials, and they are more often associated with adults than with children (Akins 1986*1:126; Rohn 1971*1:92; Whittlesey 2001*1). Mortuary items are usually located near the head. Individuals have been found wrapped in turkey feather blankets and cotton cloth; others have been found covered with wood planks (Bradley 1988*1:Table 6; Cattanach 1980*1:141). Many individuals, however, were not buried with associated grave goods; this is especially true of children. Thus, the absence of such items does not by itself suggest an atypical mortuary circumstance.

At Woods Canyon Pueblo, one individual (HRO 11) was in a conventional mortuary context (Database Map 294). This child had been placed in a prepared pit on a talus slope and was in a semiflexed position. Two vessels had been left as grave goods. In contrast, the remaining individuals (HROs 1–10) were found on the floor of Structure 5-S, a kiva, and were apparently part of a mass burial (Database Map 318). Their mortuary context was unusual in terms of the number of individuals and the positioning and location of remains. Skeletal and stratigraphic evidence indicates placement within a limited period of time after what was probably a catastrophic event.

As shown on Database Map 318, all of the more-completely exposed individuals in Structure 5-S were on their backs (supine). An adult female (HRO 3) was the only person in the more-typical semiflexed position. Three children (HROs 1, 2, and 5) were in generally extended positions. The legs of an infant (HRO 8) and a child (HRO 9) were sprawled. The left leg of HRO 9 was akimbo at placement, but his or her right leg might have collapsed into its current location from an original raised-knee position. We were unable to observe the position of the three remaining adults (HROs 4, 6, and 7) and an adolescent (HRO 10), although HROs 6 and 7 appear to have been very close to each other. Even though the positions of many of the individuals were unusual, their configurations still suggest that they neither fell nor were dropped through the hatchway but, rather, were deliberately placed.

The child buried in the talus slope (HRO 11) was accompanied by grave goods. It is difficult, however, to determine whether the artifacts found near the individuals buried in Structure 5-S are mortuary offerings. There are a few complete or reconstructible artifacts on the kiva floor (see "Abandonment and Emigration") that are not associated with the human remains, which indicates that the kiva had been or was in active use up to about the time of the mass burial. Although a few artifacts were found near skeletons, their types and locations are more suggestive of circumstantial rather than intentional association. Of the items found, a pottery disc resting under a child's pelvis (HRO 5) is the most likely to be a grave offering. It is not clear whether the axe under an adult (HRO 7), the quartz stone near a child (HRO 5), and the partial reconstructible vessels near two adults (HROs 3 and 7) are grave goods. These items may have been overlooked, discarded as refuse, or deliberately placed as mortuary offerings. For example, the foot of HRO 7 was found resting on an axe, which could have been left as a funerary object but also could have simply been left in place when the kiva was appropriated for a burial chamber. Because axes are not common grave goods and funerary objects are not usually found next to feet, the latter interpretation might be the more plausible.

The use of burial shrouds has been documented at some ancient Pueblo sites (Bradley 1988*1; Rohn 1971*1; Whittlesey 2001*1). It is possible that now-decomposed shrouds were used for some of the individuals at Woods Canyon Pueblo. It seems unlikely, however, that those individuals with sprawled positioning had been wrapped in such cloths.

The mortuary circumstance of the 10 individuals found in the kiva is quite unusual. Unlike most atypical mortuary contexts, however, this one appears to have involved some formality; that is, the positioning of these individuals has a generally composed look, suggesting care in placement. It appears that someone spent at least some time attending to the burials, as evidenced by the sandstone slabs directly on top of a child's (HRO 2) cranium and the slab between an infant (HRO 8) and a child (HRO 9). A slab is also present on another child's (HRO 5) torso, and although this could be a rock from the collapsed structure, the underlying bones have only minor dry-bone compression damage. The locations of the slabs on these skeletons, as well as the slight postmortem compression damage to the bones beneath them, indicate that the rocks had not fallen from the walls but were deliberately placed as part of the interment. Most likely, the individuals in Structure 5-S are part of a mass burial involving care and consideration in the treatment of their remains.

Sequence of Body Placement in Structure 5-S

Placement Relative to Other Bodies

The 10 individuals exposed on the floor of Structure 5-S were placed close to one another and, for the most part, in a single layer. Many were placed with their heads oriented to the south. The absence of sediment between overlapping bones, as well as the close placement and undisturbed condition of each skeleton (including their easily displaced hand and foot bones), strongly suggests that this was a single burial episode.

The six subadults were clustered in the northern two-thirds of the trench; the four adults were found in the southern one-third of the trench. If intentional, this grouping may indicate that burial practices involved age grading, or it could mean that individuals were placed according to familial relationship. However, because there might be additional people buried in this kiva, the observed pattern may be specious.

The order in which some of the bodies were placed is reconstructed on the basis of how I believe they overlapped before postmortem slumping (Database Map 318 and Table 4). HRO 2 was placed before HRO 3; the latter is believed to be resting on, or just above, the kiva hearth (see paragraph 34). The placement of HROs 5 and 8 (in unknown order) was followed by the placement of HRO 9 and then HRO 10, but the timing of this sequence in relation to the placement of HROs 2 and 3 is unclear. The haphazard positioning of HROs 8 and 9 suggests that placements may have become even more hurried as the episode unfolded. Reconstructing the entire burial sequence is difficult because it is not always clear whether a bone's position is the result of intentional placement or postmortem slumping. For example, if the arm of HRO 5 was originally covered by the ribs of HRO 2, then HRO 5 would have been placed before HRO 2 and thus HRO 3. Also, if the overlap of HRO 2's lower right arm by HRO 1's posterior cranium was intentional, then HRO 2 would have been placed before HRO 1; this configuration could have either preceded or followed the placement of HRO 3. The tight spacing makes it unlikely that both HROs 1 and 5 were placed before HRO 2. The placement sequence of HROs 4, 6, and 7 cannot be determined.

Placement Relative to Structure Use and Abandonment

Skeletal and stratigraphic data provide the most reliable indicators of the kiva's use and abandonment history, although the two data sets do not always support identical interpretations. The mass burial marked the final activity in the kiva. For two reasons, I believe that the kiva was used until the time that the bodies were interred. First, no fill from structural collapse or natural deposition had accumulated on the floor before the individuals were placed. If fill had been found between the floor and the remains, it would indicate that there was a hiatus between the time the kiva was in active use and the time of the burials. We would not expect much fill to accumulate on a kiva floor while the roof was still intact, although natural deposits can enter a kiva through its open hatchway.

The second reason for inferring that the kiva was used up until the time of burial is that localized, low-intensity burning on the underside of the adult female's (HRO 3) shoulders suggests that embers were on the floor and/or in the hearth when she was placed in the kiva. Apparently these embers burned through her skin and thin muscle layers, causing heat damage to both scapular spines. The posterior head of her right humerus has an even, reddish rim circumscribing an area of destroyed bone. Her torso was lying on a circular depressed area that was roughly where the hearth should have been (this area was unexcavated, but ashy fill was observed). Her spine's alignment became curved during decomposition, as it conformed to the uneven underlying fill. Similar low-intensity burning affected the left wrist of the nearby child, HRO 2. No burning indicative of damage from burning roof materials is present on the superior bone surfaces of either person. Rather, the color, intensity, and pattern of burning suggest that these individuals were placed near or over a thermal feature that had been recently used (the degree of damage does not suggest placement over an actively burning fire, however). A fire may have been smothered just before HRO 3 was placed, but stratigraphic evidence is unavailable.

Interment in sediment tends to hold bones in place during decomposition. Although the positioning of the individuals in the kiva varied, multiple lines of evidence indicate that they shared an open-air treatment; that is, they were not covered with dirt at the time of burial. Almost every individual showed signs of bones having shifted, which is indicative of slumping during soft-tissue decomposition. Entire anatomic units (such as the lower leg) of two adults and two children were displaced. The long-bone epiphyses of some children were in correct anatomic position, but their adjoining diaphyses (shafts) had rotated by as much as 180 degrees. The adult female's lower left leg bones were similarly rotated. In addition, the unfused epiphyses of some children were found several centimeters below their associated diaphyses. As sediment built up following the removal of the roof, soil compression could have caused some displacement, such as the upside-down position of HRO 4's metatarsals. Cumulatively, however, the biological evidence suggests that most displacement occurred in an open-air environment. This conclusion is further supported by stratigraphic evidence: cranial vaults that are covered with dirt during interment tend to fill with sediment fairly quickly, but the observable crania of the people in Structure 5-S contained relatively small amounts of dirt. In addition, there was no stratigraphic evidence in the surrounding and overlying sediments to suggest that dirt was intentionally placed on top of the remains. Instead, roof-fall deposits rested directly on the skeletons.

Open-air interment, however, does not mean that these individuals were unprotected. Several factors indicate that they were placed within a roofed structure. The bones were in generally excellent condition, and no carnivore or rodent damage was observed. There were no indications of displacement from rodents, carnivores, people, or flooding of the structure. Also, the bones were unweathered, with only minor postmortem erosion. The skeletal and stratigraphic evidence suggests that the kiva was used as a sepulcher, or tomb, although this behavior most likely was a response to a particular circumstance rather than representing a typical burial custom.

Churchill (see "Abandonment and Emigration") infers from the archaeological evidence that the large roof timbers in this kiva were salvaged and the smaller timbers were burned. The question is, How soon after the placement of the bodies in the kiva did this occur? Skeletal evidence suggests that the kiva was still roofed at least several months or perhaps years after the burials; the lack of animal disturbance and damage to the bones also suggests that the hatchway and ventilator system were sealed. In their analysis of crime scene evidence, Krogman and Iscan (1986*1:28) find that ligaments, which connect the joints as well as the bones of the hands and feet, are the last soft tissues to decompose. They are more likely to loosen first in infants, then in children, and eventually in adults. As discussed above, several individuals in Structure 5-S have bones that apparently were displaced during soft-tissue decomposition. One infant (HRO 8), two children (HROs 2 and 9), and two adults (HROs 3 and 4) have rotated bones. The rotation of the adults' bones in particular suggests that there was an extended period of time when their remains had to have been uncovered by sediment but protected by an intact roof. The adolescent's (HRO 10) completely excavated right ulna was not associated with other right arm or hand bones, suggesting that it was no longer enfleshed or held in place by ligaments when it became displaced.

Although the time required for bone rotation following ligament decomposition is uncertain, several factors suggest that more than a brief time elapsed in the case of the individuals in Structure 5-S. Krogman and Iscan (1986*1:23) note that the decomposition process involves both external and internal sources, such as soft-tissue invasion by insects, putrefaction from intestinal bacterial growth, release of body fluids and gases, and the decay of soft tissue, organs, muscles, ligaments, and bony structures. The final stage of decomposition involves the progressive loss of bone moisture, transforming elastic "green bone" into brittle "dry bone." Cool environmental conditions, such as would have existed in a roofed kiva, decelerate decomposition, as would a central location in a mass burial. Exposure to air accelerates decay, although this knowledge apparently is derived from the observation of human remains left on the surface of the ground. If, as I believe, the entry to the Woods Canyon kiva was sealed, the availability of oxygen may have been diminished. Soil conditions, which play a significant role in decomposition rates, would not have affected the burials in the kiva, because the bodies were not covered with sediment.

Roof-fall deposits rested directly on the remains, including a burned beam on the crania of HROs 3 and 5. Neither skull, however, showed burning or heat damage. Each cranium had only minor postmortem fractures indicative of slow compression forces. Although a burned beam was found directly on the adult female's face, her fragile nasal and lacrimal bones were articulated and essentially intact, albeit displaced within the cranial cavity. Both the lack of burning and the minimal postmortem fracturing suggest that a significant amount of time elapsed before the beams slowly collapsed onto these crania. A very few bones, such as the pelvis of HRO 3 and the radius of HRO 10, have perimortem/early postmortem fractures that suggest damage from falling rocks, possibly occurring during the removal of the roof. Overall though, the minimal postmortem damage is more indicative of slow compressive forces than of a sudden impact from roof collapse. This interpretation, however, appears to be contradicted by the stratigraphic evidence. Churchill (personal communication 2001) maintains that the absence of natural laminae in the roof fall of Structure 5-S indicates that the roof did not collapse gradually over a long period but, rather, collapsed more rapidly than may be indicated by the skeletal evidence. For a detailed description about the stratigraphy, see The Woods Canyon Pueblo Database.

To summarize, the skeletal evidence suggests the following sequence of events: (1) the kiva was actively used up to the time of burial; (2) the individuals were placed in the kiva over a short period of time; (3) the remains were not covered with sediment, but all chamber entryways were probably sealed; (4) the roof was partly salvaged and burned at a later, perhaps much later, date; and (5) the remaining roofing materials and/or fill deposits collapsed relatively slowly onto the remains. Both the skeletal and stratigraphic evidence suggests that the burning of the kiva roof was not directly related to the mass burial, because of the time that elapsed between these two events. Nor did the interment of these individuals coincide with the end of the site occupation; the burial occurred long before the roof beams were salvaged, and the beams are believed to have been salvaged for use at the pueblo before the end of occupation.

Skeletal Evidence of Health

Various skeletal indicators provide evidence of the bony reaction to infection and inadequate nutrition. For the most part, only chronic and/or serious episodes affect the skeleton. Throughout human history, most people who have died from acute disease processes, such as respiratory or gastrointestinal infections, do not show a skeletal response (Ortner and Putschar 1981*1:105). The Woods Canyon Pueblo skeletal remains were evaluated for evidence of dental enamel hypoplasia, infectious caries, porotic hyperostosis, periostitis, nonspecific lesions, and stunting (Table 5).

Dental Enamel Hypoplasia

Dental enamel hypoplasia (DEH) is a developmental defect created by the disruption of tooth-crown, or dental enamel, formation (Goodman and Rose 1991*1) in utero and during the first seven or so years after birth. It provides a permanent marker of physiological stress since, unlike bone, enamel is not remodeled as an individual ages (Hillson 1986*1:136). DEH provides the only reliable determination of an individual's age at the time a defect develops—referred to as "age-at-stress"—because enamel formation occurs in a predictable, or chronometric, sequence. Goodman and Rose (1991*1:288) developed the chart used in this study, applicable to permanent teeth only. Most commonly, DEH results from the body's systemic response to infection, malnutrition, or a combination of these factors. Systemic reactions involve more than one tooth simultaneously, in more than one quadrant of the mouth. DEH can also develop from trauma, such as a blow to the mouth, in which case the defect is unilateral and localized—that is, it affects only one or a very few adjacent teeth. Assessments of an individual's degree of DEH (mild, moderate, or severe) are based on the proportion of teeth affected and the number of episodes represented.

The occurrence of DEH has different implications in terms of the health of the individual, depending on whether the condition is observed in deciduous or permanent teeth. Because deciduous teeth develop in utero, deciduous-tooth DEH is an indirect indicator of maternal stress, such as the mother's infection or malnutrition; deciduous DEH rarely occurs in individuals whose mother's are healthy and well fed. In contrast, permanent teeth develop between birth and about seven years of age, so DEH in permanent teeth reflects the relationship between a child's health and the environmental conditions to which he or she is exposed. Deciduous-tooth and permanent-tooth DEH are discussed separately below.

As shown in Table 5, six of the 11 Woods Canyon Pueblo individuals (55 percent) could be assessed for dental enamel hypoplasia, and each had at least one tooth showing evidence of DEH. The infant has an almost complete set of deciduous teeth. Two children have at least seven deciduous teeth, but only one and three permanent teeth, respectively, that could be evaluated. One adult and two children have dental arches that have at least 12 visible permanent teeth, with a minimum of seven anterior teeth (incisors and canines) per individual.

Three children were evaluated for deciduous DEH. The teeth of one child (HRO 1) are unaffected, but another child (HRO 5) and an infant (HRO 8) have a moderate degree of dental hypoplasia. Deciduous DEH occurs in 47 percent of the observable teeth, with an average 1.08 defects per affected tooth. The Grasshopper Pueblo data are not comparable, but the frequency of deciduous DEH at Sand Canyon Pueblo, a late Pueblo III site not far from Woods Canyon Pueblo, is 5 percent (Bradley 1998*1). Because of methodological limitations, it is unknown whether these children at Woods Canyon Pueblo experienced a single event or multiple episodes of stress. However, the defects observed in HROs 5 and 8 do suggest maternal stress.

Five individuals have observable permanent teeth. Each has DEH, but two (HRO 1 and HRO 5) have too few observable teeth for reliable interpretation, and data for these individuals are not included in the following analyses. Permanent-tooth DEH occurs in 96 percent of the observable anterior teeth of the remaining three individuals (HROs 2, 3, and 11), with an average of 1.46 defects per affected tooth. Both HRO 2 and HRO 11 have mild cases of DEH, whereas HRO 3 is moderately affected. Each of these three individuals experienced at least one episode of systemic stress. Their average peak (most common) age-at-stress occurred at about 3 years +/- 6 months, which is in keeping with that seen in other ancient Pueblo populations (Bradley 1998*1; Malville 1997*1). The evidence indicates that DEH was pervasive, yet mild to moderate in expression.

Infectious Caries

Caries that invade the pulp cavity offer a significant risk of infection from plaque bacteria such as Streptococcus mutans, Actinomyces israelii, and Gram negative bacteria. None of the five individuals (HROs 1, 2, 3, 5, and 11) with assessable dentition had infectious caries. In fact, none had any carious lesions, and the few dental calculus deposits present were insignificant in size.

Porotic Hyperostosis

Porotic hyperostosis is the skeletal reaction to iron-deficiency anemia that affects the eye orbits and cranial vault bones. Porotic hyperostosis occurs when the blood-forming marrow bone between the inner and outer vault surfaces expands as the body responds to the need for increased hemoglobin production. This skeletal response can create surface lesions that have a porous, spongy appearance. In New World populations, anemia results from nutritional deficiency, parasitism, or infection. Maize-reliant diets produce especially high rates of anemia. Such diets are unbalanced to begin with, and corn is low in iron and has phytic acid, which greatly inhibits intestinal absorption of iron.

Five individuals from Woods Canyon Pueblo (46 percent) could be assessed for porotic hyperostosis (HROs 1, 2, 3, 5, and 8). Of these, the adult female (HRO 3) has two dime-sized vault lesions, and the infant (HRO 8) has orbital lesions only. All the lesions are small and mild in expression. Both individuals had active lesions, but the insignificant appearance of the lesions suggests that anemia was minimally, if at all, related to their cause of death. The 40 percent frequency of porotic hyperostosis in these five individuals is similar to that of other ancient Pueblo groups (Bradley 1998*1; Stodder and Martin 1992*1). Although the sample size is small, the data indicate that anemia was not a significant or chronic problem in terms of frequency or severity.


Periostitis is an inflammation of the periosteum, the membrane of connective tissue covering bone shafts. It is the most common bone inflammation found in skeletal remains. Localized periosteal lesions can develop in response to bacterial infections of the overlying tissues such as those that occur with cuts, bruises, or fractures. Primary periostitis can also occur as a skeletal response to a systemic bacterial infection, such as streptococcus or staphylococcus. In such cases, the lesions can be extensive and circumferential, and they can affect several bones simultaneously. The tibia is the most frequently affected long bone, and bilateral lesions (lesions on both tibias) are a common, highly reliable indicator of systemic infection.

At Woods Canyon Pueblo, nine (82 percent) of the individuals were sufficiently exposed (that is, at least one tibia could be observed) to allow for assessment of this condition. None had tibial lesions, including bilateral lesions in those persons for whom both tibiae were observable. Only one child (HRO 11) might have had a localized periosteal lesion. This child exhibits what may be a small area of sclerotic (healed) bone on the distal (elbow) end of the left humerus; the right humerus is unaffected. Alternatively, the observed irregularity may have been caused by erosion; the individual had been placed in a shallow pit-grave on the talus slope, and the bone was damaged postmortem. Assuming that periostitis is present, however, its restricted appearance indicates a localized infection or trauma. It is interesting that the (possible) lesion is in the area of the capitulum, which shows initial fusion to the diaphysis (shaft) but not to the trochlea. The capitulum is an articular structure that normally begins fusion with the trochlea—another articular structure—at about 13 years of age, followed by fusion to the shaft of the humerus in the late teens. There is no indication of trauma, but my colleague, Sali Underwood, noted that the premature fusion of this capitulum to the diaphysis could be a consequence of infection. The possibility of a congenital condition is examined in the discussion of nonmetric traits.

Regardless of whether HRO 11 had a localized inflammatory response, the nine Woods Canyon Pueblo inhabitants whose tibiae could be evaluated did not experience significant or chronic levels of infection. The absence of either systemic responses or unambiguous localized lesions is comparable to that seen in other ancient Pueblo groups (Akins 1986*1; Bradley 1998*1; Hinkes 1983*1; Karhu 1996*1).

Nonspecific Lesions

Both the left and right pubic symphyses of the adult female (HRO 3) had defined, discrete lesions. The left symphysis had a small pit, containing sclerotic (healed) bone. The two pits on the right symphysis were also small and contained both healed and unhealed bone, showing a reactive process at the time of death. The bone around the lesions was normal, and it is unclear whether the lesions developed from infection or were a product of trauma. Pulling of the cartilage covering the pubic symphysis, such as might occur during childbirth, could cause such lesions to develop (Krogman and Iscan 1986*1).


Stunting, or growth retardation, is defined as stature that is significantly below average. Stunting is assessed by comparing the maximum length of a long bone, preferably the femur, to the average length of the same bone in same-aged and—in the case of adults and older adolescents—same-sexed individuals from a genetically similar population. Long-bone growth occurs in a genetically controlled sequence but is greatly affected by an individual's health and nutritional status. Persistent adolescent anxiety has been tied to blunted human growth hormone response, resulting in stunted final stature in females (Pine et al. 1996*1:860). An acute physiological stress can halt growth temporarily, although children usually experience catch-up growth. When stress is chronic or of long duration, this rebound effect is less likely to occur. Thus, since linear growth takes place from the fetal period until the end of puberty, long-bone length is considered to be an effective, long-term indicator of an individual's health status.

There are no established osteological standards for determining stunting in skeletal remains. Hinkes (1983*1) considers stunting to be present when long-bone length is greater than one standard deviation below the mean height for age. On the basis of research data on living children (Khan et al. 1996*1), I consider stunting to be more accurately identified when it is defined as exceeding two standard deviations below the mean (Bradley 1998*1). To minimize the issue of genetic variability, the reference population used in this study is from Grasshopper Pueblo.

Five of the 11 individuals from Woods Canyon Pueblo (46 percent) could be assessed for stunting (HROs 1, 2, 3, 8, and 9); that is, their ages could be estimated and the lengths of their long bones could be measured. None of the individuals can be considered stunted, regardless of whether Hinkes's or Bradley's criteria are applied, which indicates that these people had been adequately nourished throughout their lives. Moreover, their average to slightly-above-average heights may reflect low levels of infection and anemia.

The long-bone lengths of two of the subadults were similar to the average documented for Grasshopper Pueblo children of the same ages; the measurements for two other children were slightly above average. Although the Woods Canyon children were perhaps not quite as tall as their counterparts at Sand Canyon Pueblo (Bradley 1998*1), the evaluation of stunting suggests overall similarities in the two populations. The height of the adult female, who achieved a final stature of 153.60 +/- 3.816 cm, was only slightly below the average female height of 155.6 cm in the Mesa Verde region (Martin et al. 1995*1).

Concurrent Indicators of Infectious and Nutritional Stress

Bioarchaeologists recommend the use of several, rather than single, skeletal indicators to make more accurate assessments of physiological stress, especially in regard to nutritional status (Armelagos and Goodman 1991*1:51). The multiple-indicator approach is considered to be more revealing of patterns of chronic, acute, or subgroup physiological stress (Martin et al. 1995*1). This technique also addresses the bioarchaeological reality of incomplete remains, in which the same indicator cannot always be assessed in all skeletons found at a site. In addition to providing diverse data on an individual's physiological history, multiple-indicator analysis allows a more accurate evaluation of the intensity of skeletal stress in a specific stress episode (Cook and Buikstra 1979*1:656). That is, the severity of physiological stress increases as the number of affected bones and tissue types increases.

None of the seven more-completely exposed individuals from Woods Canyon Pueblo has concurrent markers of infectious or nutritional stress. Although comparative information is generally not available for Mesa Verde–region populations, this frequency is similar to the low rate (5 percent) documented for Sand Canyon Pueblo (Bradley 1998*1).

Skeletal Evidence of Injury and Activity

Skeletal injury includes burning, fracture, and osteoarthritis, the last indicating a response to joint trauma. In addition, habitual and prolonged activities can place stress on bony muscle attachments, producing such reactions as lipping, spurring, and elevated tubercles (Kennedy 1989*1:134). As detailed in Table 6, every individual in the Woods Canyon Pueblo assemblage was evaluated for possible burning and fracture; however, only the adults were evaluated for osteoarthritis and occupational markers, because these are age-related conditions.


Bone is burned most often as the result of human actions, but it can also occur under natural circumstances. Specific changes in bone color and surface texture indicate the degree of thermal alteration (White 1992*1). The extent of burning is affected by such factors as the duration and intensity of heat exposure, the thickness of the overlying soft tissues, and bone moisture content (Turner and Turner 1999*1).

Two Woods Canyon Pueblo individuals (HROs 2 and 3) have very localized burning. The affected left wrist of a child (HRO 2) and shoulder areas of the adult female (HRO 3) discussed above apparently resulted from accidental exposure to a thermal feature at the time of burial (see paragraphs 33–40).

In addition to having a perimortem fracture, the adolescent's (HRO 10) right ulna has a small, localized black area superior to the medial notch. The constraints imposed by in-field analysis and the size of the discoloration made it difficult to determine whether the blackening was caused by low-intensity burning or postmortem staining. The collapse of a burning or burned beam onto the ulna could have created such a discoloration, as well as caused a fracture and/or dislodged the bone. Alternatively, the decomposition of soft tissue can stain bones, especially in open-air, mass burials. Several bones from nearby individuals show similar discolorations, but their affected areas are large enough that it could be determined that they were stained rather than burned. Even if the discoloration on the ulna of HRO 10 is the result of burning, the combination of perimortem fracture, localized burning, and displacement of this bone can still be reasonably explained as the consequence of natural depositional forces.

In sum, the coloration of each affected bone is indicative of exposure to a low-intensity heat source. These very few cases of burned bone seem to represent accidental or depositional circumstances rather than intentional human behavior.


Skeletal fracture provides direct evidence of both accidental and intentional human actions, as well as a better understanding of depositional events. For example, the head and face are common targets of intentional violence, and "parrying" breaks to the midshaft of the lower arm can be the result of defensive reactions. In contrast, the locations of fractures from accidental injuries are more random, although fractures might occur in patterns that suggest an environmental condition—for example, a high frequency of broken wrists in populations living in icy climates.

Evaluations of bone fractures include assessment of when the fractures occurred. Antemortem trauma is a skeletal fracture that occurs before death. Such fractures are clearly recognizable because they show evidence of healing, with a greater degree of healing associated with a longer time since injury (Krogman and Iscan 1986*1). Perimortem trauma refers to skeletal damage occurring at or about the time of death, when the bone is still moist, elastic, and "green." Postmortem trauma occurs after death, when the bone has lost its organic, collagen component and has become dry and brittle. Green long-bone fractures have sharp edges and are often longitudinal or spiral, and small flakes of bone can be held in place by periosteal tissue. On the other hand, dry long-bone fracture edges are usually rough and jagged (White and Folkens 1991*1:358).

Postmortem breaks are characterized by irregular fracture lines, and there are no adhering bone flakes because the periosteum is no longer present. Long-bone postmortem fractures are usually recognized as short, straight splinters at transverse angles to the diaphysis (Villa and Mahieu 1991*1:34). A blow that merely creates a depression fracture in a living or recently deceased person will shatter the same bone in a long-dead individual. Neither perimortem nor postmortem fractures will show evidence of healing. However, because bone can remain green for several years after death (depending on the depositional environment and the age of the individual), it often is difficult to differentiate perimortem from ancient postmortem breaks. Thus, fracture assessments must take into account mortuary context.

All human bones in the Woods Canyon Pueblo assemblage were analyzed for evidence of skeletal trauma. Antemortem and perimortem fractures are rare. Postmortem fractures are widespread but generally insignificant. Only one (11 percent) of the Woods Canyon Pueblo skeletons has an antemortem fracture. This adult (HRO 4) has a well-healed fracture of the left fifth metatarsal, which is a common accidental injury in all populations. The adjoining phalanx is normal, as are the remaining metatarsals and phalanges.

Of the hundreds of bones observed in Structure 5-S, only four to five postcranial bones (less than 1 percent) show perimortem fracture; these incidences involve two (18 percent) of the 11 individuals (HROs 3 and 10). In both cases, the damage almost certainly resulted from natural, depositional events rather than from intentional human activity. The adult female (HRO 3) has minor cracking of the ventral surface of her fifth sacral vertebra, a complete fracture of the ischiopubic ramus of her right innominate, and a depression fracture in the iliac blade of her left innominate. The instrument of damage was a sandstone block lying almost horizontally in the pelvic cavity, with one of its sharp corners next to the depression fracture. In adult females, this region is well protected by muscle, internal organs, and flesh. Consequently, this injury probably took place after enough soft tissue had decomposed that the cavity was no longer protected, but while the bone was still green. Considering the environmental circumstances of this mass burial, it is unlikely that the damage occurred at or very soon after the time of death. The depth of the fracture suggests that the bone was damaged when the block fell from a height (perhaps as the roof collapsed) rather than as the result of slow compression during the filling of the kiva.

The right femur of HRO 3 has a perimortem or ancient postmortem break across the upper third of its shaft. The fracture is intermediate between transverse and spiral, and the edges are sharp and squared, whereas the exposed bone surface is slightly rough. The corner of a sandstone slab was under the femur at the location of the break. The break apparently was caused by slow compression of the femur onto the edge of the slab, most likely sometime after death when the bone still retained some moisture.

The adolescent (HRO 10) has a green fracture of the disarticulated right ulna; there may also be a small area of localized burning, discussed above (paragraph 63). This bone was out of anatomic position, lying anterior side up and across the tibia and fibula; although it is possible that it belongs to another individual, its development and size are consistent with the development and size of HRO 10's tibia and fibula. The ulna was completely excavated, but only the proximal two-thirds is present. The distal end of the humerus was not seen beyond the profile of the test unit, and no radius or hand bones were found; these bones would have been discovered if this were an articulated lower arm. The ulna has a complete fracture at midshaft. The posterior surface has a spiral fracture, which resulted in the removal of a small bone flake. The fracture edges and bone flake scar are sharp. The break on the anterior surface is jagged and had caused the removal of a small, sharp flake of bone. The medial side of the shaft from the metaphysis to the break is undamaged, but the shaft's lateral surface is sharp and jagged in the area of the break. No adhering bone flakes are present.

There is no artifact evidence to indicate what might have caused the ulna to break, but the biological evidence does not suggest intentional human activity. This carefully examined bone has no cut marks, chop marks, or abrasions, nor does the break look like a parrying fracture. The lack of such damage on any of the other bones in Structure 5-S further argues against fracture as the result of violence. The ulna is fractured on three surfaces, which could suggest hammer-and-anvil damage; such trauma can occur when a bone lying on a hard surface is struck by another hard object. Humans can produce this type of damage, and the intentional fracture of this bone cannot be ruled out (White 1992*1). However, the occurrence of stone slabs underneath other individuals in this kiva could mean that a slab served as an accidental anvil, perhaps during the collapse of the roof. Possibly this break occurred when the bone was still "green" but no longer enfleshed or held in place by ligaments. If so, it may have been lying on a slab, struck by a falling object, and dislodged to its current position. The correct anatomic position of this bone would have been beyond the limits of the test trench, so it was not possible to determine whether the ulna was originally lying on a slab, as in the case of the right arm of nearby HRO 9. Alternatively, this could be a perimortem break with subsequent displacement of the bone by roots or rodents, although no root damage or tooth marks were observed.

All the skeletons except one adult (HRO 7) have postmortem damage typical of dry-bone breakage. Although pervasive, the fracturing is at a low, insignificant level. Most of the affected bones have irregular, straight edges, as observed in one child's (HRO 5) two transverse fractures of the left femur; one of these fractures, visible on Database Map 318, occurred during excavation. Other bones, such as the infant's (HRO 8) right radius and ulna, have short, irregular, longitudinal splinters with straight edges. None of these fractures have adhering flakes. Most long bones were sufficiently intact for accurate measurement. Unfortunately, the crania of three children (HROs 1, 2, and 5) and one adult (HRO 3) suffered extensive dry-bone damage either from slabs that were deliberately placed as part of the interment or from the weight of collapsed roof beams. The cranial bones are essentially in place, suggesting that compression, rather than a sudden impact, caused the observed damage. Dirt inside the cranium can provide compression resistance for vault and facial bones, but most of the cranial vaults of the individuals in the kiva were only partly filled with sediment.

Overall, the frequency and degree of skeletal fracture is insignificant. There is no indication that either the one antemortem fracture or the two perimortem fractures represent purposeful, violent intent. No bones have cut marks, chop marks, or anvil abrasions, and there are no cranial depression fractures. The lack of antemortem or perimortem cranial trauma in the Woods Canyon Pueblo assemblage is comparable to what has been observed at other ancient Pueblo sites (Akins 1986*1; Hinkes 1983*1; Debra Martin, personal communication 1997). Significantly more perimortem cranial trauma has been documented in the human remains from Sand Canyon Pueblo (Bradley 1998*1) and Castle Rock Pueblo (Kuckelman et al. 2002*1).

Occupational Markers in Skeletal Remains

Repetitive motions can create specific bony reactions, such as squatting facets. Because the more-completely revealed skeletons at Woods Canyon Pueblo were remains of younger people, identification of occupational markers was limited. However, the adult female (HRO 3) has a well-developed foramen magnum facet, associated with a flattened and lipped axis dens (second cervical vertebra). Merbs (1983*1) notes that this can result from hyperextension of the neck, as may occur with tump-line use.


Osteoarthritis is a degenerative disease of the joints resulting in inflammation and destruction of both bone and cartilage. This age-related condition is often associated with abrupt repetitive motions or skeletal trauma. The remains of only one Woods Canyon Pueblo adult show signs of osteoarthritis. The left elbow of the adult female (HRO 3) has barely discernible osteoarthritis; her right elbow is unaffected. Also, her axis dens (second cervical vertebra) has minor lipping.

Skeletal Evidence of Relatedness

In this section, I discuss skeletal characteristics that provide insights into possible cultural and biological relationships. These characteristics include artificial cranial deformation, congenital traits, and nonmetric traits.

Artificial Cranial Deformation

Artificial cranial deformation occurs when an infant's pliable cranial bones are flattened or shaped by cultural practices. As seen in ancient Pueblo skeletons, unintentional, often asymmetrical, deformation can be caused by the use of cradleboards. A specific head shape can be produced by wrapping the head in a deliberate, prescribed pattern. Among prehistoric Southwestern populations, artificial cranial deformation is indicative of Pueblo cultural affiliation. Each of the three Woods Canyon Pueblo individuals who could be assessed for this condition (HROs 1, 3, and 5) had lambdoidal cranial deformation and therefore a Pueblo cultural affiliation.

Congenital Defects

As noted by Barnes (1994*1:xxx), "every population has its own genetic pattern of developmental tendencies for producing particular defects." The Woods Canyon Pueblo individuals exhibit several anomalies (see Table 7) that are more indicative of familial ties than of traits, such as monozygotic twinning, that work at the population level (David Hildreth, personal communication 2000). The incidences of these anomalies do not firmly demonstrate a familial relationship, but it is interesting that they are present in almost every individual who was more than slightly exposed.

Three individuals (HROs 1, 3, and 9), all found in Structure 5-S, have premature fusion of an epiphysis; the number increases to four if the fused capitulum of HRO 11 is of genetic rather than infectious origin. Two individuals (HROs 1 and 9) have fusion anomalies involving digits. The adult female has both early and late fusion anomalies that affect the cranium and several long bones.

Although less clear, it is possible that other individuals have fusion anomalies as well. A child (HRO 2) has a bilateral sacral hiatus (nonfusion) of sacral vertebrae 4 and 5. This can be a normal condition (Barnes 1994*1) or a mild form of spina bifida (Steele and Bramblett 1988*1). Another child (HRO 5) has only four of five sacral vertebrae, and the first sacral vertebra is abnormally large. Although this child lacks either sacral vertebrae 3, 4, or 5, one coccygeal bone is definitely present. Postmortem loss is very unlikely because these bones were articulated and in excellent condition. HRO 5 also exhibits a symmetrical hiatus of the third and fourth sacral vertebrae (by sequence rather than anatomic number). These sacral anomalies, which are neural tube defects, have a strong familial tendency; however, external factors like maternal mineral deficiency are needed for expression of the trait (Barnes 1994*1:41). In addition, the infant (HRO 8), who was 12 to 24 months old (18 months ± 6 months), has an incompletely fused mandibular symphysis. However, this could be a developmental condition since the symphysis normally fuses at about one year of age. Finally, three of the six people with observable dentition (HROs 3, 5, and 8) have bilaterally rotated teeth in uncrowded arches, with different teeth affected in each case. This, too, could be evidence of a familial connection among these individuals.

None of the above defects would have had a significant, or perhaps even noticeable, effect on health. Premature fusion of a phalange or metatarsal results in a somewhat shorter but still usable digit. HRO 5's congenitally absent sacral segment and possible sacral hiatus could have led to bladder incontinence; however, this child's bones do not otherwise suggest significant physiological stress. Finally, the early fusion of the adult female's (HRO 3) cranial sutures only affected the shape of her head.

Nonmetric Traits

Nonmetric traits are skeletal features that can be observed but not measured, such as the presence of a septal aperture of the humerus. These traits are considered to show familial inheritance and are used to evaluate biological distance (Buikstra and Ubelaker 1994*1). We recorded standard nonmetric trait data on the Woods Canyon Pueblo remains. Most of the traits were scored as unobservable. Of the few that were observable, most involved only four individuals (HROs 1, 2, 3, and 5). The sample size is simply too small for a reliable assessment of kinship based on nonmetric traits.


Archaeologists discovered 24 isolated human bones and the articulated skeletal remains of 11 individuals during fieldwork at Woods Canyon Pueblo. One child was in a shallow burial pit on the talus slope, near Structure 5-S. Ten individuals, infant through adult, were found on the floor of Structure 5-S, a kiva, within a 2-x-1-m test trench. In addition, the cranium of an adult, probably male, was present at floor level beyond the limits of the test trench. It is possible that more individuals are present in areas of the kiva that were not excavated.

Although human remains on kiva floors are generally found in sprawled, face-down positions—indicating that they were not formally buried—the individuals in Structure 5-S appear to have been part of a mass burial involving some degree of care and consideration. Together, their generally extended, supine positioning and the sequence in which they were placed indicate that they had been placed carefully, although probably hurriedly, in the chamber, which was then sealed. The kiva roof was at least partly burned after some unknown, but not short, period of time. The skeletal evidence suggests that as the roof deteriorated, burned beams slowly collapsed onto the crania of two of the individuals, causing postmortem compression damage (the stratigraphic evidence, however, suggests a more rapid collapse of the roof).

Human remains were not found in any of the other eight tested kivas at Woods Canyon Pueblo, which further suggests that this particular mortuary location represents the response to a catastrophe. Their mortuary circumstance indicates that these people died within a short period of time, although it is unclear whether the time interval was hours or days. Given the unpleasant effects of decomposition, especially when uncovered remains are in a small, poorly ventilated space, it is unlikely that people were placed beside each other in the kiva over a much longer period of time. The lack of dirt between overlapping individuals further suggests a single burial episode.

The condition of the skeletons, along with the prevalence of bone displacement and rotation, indicates that these individuals were not immediately covered with dirt. In my opinion, it is unlikely that the roof was dismantled and burned at the time of burial, which would have resulted in falling beams, stones, and sediments damaging the bodies of the individuals. If that had happened, resilient soft tissues would have protected their bones, and any fractures would have occurred as green-bone breaks. Instead, almost all breakage involves dry-bone fractures indicative of slow compression. There are only two or three perimortem fractures, but biological evidence indicates that they did not occur at the time of burial. The collapse of the roof in the very early postmortem period might have caused some of the bones to move into their current rotated positions; however, other rotations could have occurred only after the decomposition of ligaments. This interval could have been prolonged, considering the numerous factors that could have delayed decomposition.

The virtually simultaneous deaths of the people in the kiva allow the assessment of their health at a given, if unknown, point in time. And, since it takes time for the skeleton to develop and remodel stress reactions, we could evaluate their health within the year or so before death. On a cautionary note, inferring the health status of the inhabitants of Woods Canyon Pueblo on the basis of a small sample is risky; for example, the frequency of anemia might have been higher if more crania could have been evaluated. Essentially, however, the skeletal data indicate good health throughout the months preceding death.

The skeletal remains found in Structure 5-S showed little evidence of disease. Every individual who could be assessed had dental enamel hypoplasia, but the manifestations were generally mild and most likely related to the developmental vulnerability of young children. The frequency of this condition at Woods Canyon Pueblo, as well as the average age-at-stress, is similar to the patterns documented for other ancient Pueblo groups. There is no sign of serious, chronic, or systemic infection. Although several crania were unobservable, there is little evidence of anemia in those that could be examined. The average to above-average stature of the Woods Canyon inhabitants further suggests that, for the most part, they were adequately nourished. No individual has skeletal stress markers that developed concurrently, as tends to happen in serious or prolonged episodes of stress.

In the Mesa Verde region, multiple burials have been reported at Pueblo sites dating from the Basketmaker through Pueblo III periods (Nordenskiöld 1979*1; Turner and Turner 1999*1). Yet, such mortuary circumstances are uncommon. When discovered, they tend to be interpreted as being the result of epidemics, especially when children are involved (Morris 1933*1:200). On the other hand, human remains found on kiva floors often show evidence of violent death.

The foregoing discussion brings us to the question, What caused the deaths of the individuals whose remains were found in Structure 5-S at Woods Canyon Pueblo? Here, I consider the likelihood of death from illness, poisoning, and violence. All in all, the biological data suggest death from an acute condition. It is not possible to identify each person's specific cause of death; however, given their apparently synchronous burial, it is more likely that all or most of the people whose remains were exposed in Structure 5-S experienced a single, common lethal stressor rather than multiple (different) problems.

It is possible that these people succumbed to a virulent, infectious disease that killed before a bony response could take place, that spread quickly, and that affected people of different ages (at least infants through young adults). It is difficult to identify specific pathological infections in past peoples. A potential cause of death is plague, a bacterial infection that is endemic in the northern San Juan region today. Individuals faced with an endemic disease often develop resistance as they age, which could explain the youthful age-structure of the people in the kiva. Since turkey bones have been found at Woods Canyon Pueblo, the disease could have been salmonellosis, a highly contagious bacterial infection. However, the prevalence of older children through young adults in Structure 5-S argues against this as the cause of death, because individuals of this age usually do not succumb to salmonellosis. Another possibility is viral infection. So far, medical researchers have traced the hantavirus in the American Southwest only as far back as 1959, but the Centers for Disease Control (2000*1) reports that the Navajo "recognize a similar disease in their medical traditions, and actually associate its occurrence with mice." Even if hantavirus was present in ancient times, however, it is unclear whether it would have produced the demographic pattern seen at Woods Canyon Pueblo; at least in the modern era, this quickly lethal virus usually does not affect children.

Poisoning—either accidental or intentional—is another possibility, because every individual found in Structure 5-S was old enough to have consumed prepared food. According to Micromedex (1999*1), there are not many dangerous plants in the northern San Juan region. Water hemlock (Circuta maculata) grows in swampy areas, as currently exist near Woods Canyon Pueblo, and along streams. This poison produces a severe gastrointestinal reaction and respiratory arrest. A mouthful of this plant, which looks like wild celery, could be fatal. A cow can be killed by a walnut-sized portion, so even a handful of water hemlock in a cooking pot has the potential to cause the deaths of many people. The ingestion of death camus (Zigadenus sp.), which looks like a wild onion, can result in blocked electrical impulses to the heart, leading to a greatly decreased heart rate and, in some cases, death. However, it would take more than 200 bulbs to kill the number of people found in Structure 5-S (Rocky Mountain Poison Control Center, personal communication 1999), and it seems improbable that an error of such magnitude would have been made by experienced foragers. Of these two plants, water hemlock is the more likely potential poison, both in terms of toxicity and the small amount needed to affect so many people. Ingestion of water hemlock could explain the presence of so many youthful, generally hardy individuals who normally would survive an infection. Death from poisoning is usually accidental, but intentional poisoning is another possibility.

I consider it implausible that these individuals were attacked or executed, even though it is possible to kill someone without harming the skeleton. Forensic data show that lethal knife wounds often do not harm bones, and infants and young children are especially susceptible to lethal kicks or punches to the abdomen (Cohle et al. 1995*1; Fossum and Descheneaux 1991*1). Nonetheless, with so many bones present, it is highly unlikely that there would be no evidence of skeletal damage if human-initiated violence had played a role in the deaths of the individuals in the kiva.

In my opinion, these people probably died during a highly contagious epidemic. In nonantibiotic societies, a virulent infection is what usually kills people, and there is nothing in the Structure 5-S biological or mortuary data to suggest otherwise. Death from poisoning seems less likely but cannot be ruled out. Violent death is by far the least likely of the three possibilities. Regardless of the cause, the deaths of at least 10 individuals in a short period of time must have been a dramatic occurrence in the village.

The adult female (HRO 3) would have been old enough to have been the mother of one or more of the children aged 12 years and under. Given the latitude within each individual's age range, it is biologically feasible that she was the mother of anywhere from one to all of the children found in Structure 5-S. It is less likely, but still possible, that HRO 3 was the mother of HRO 10, a youth of unknown sex. The age of this adolescent was estimated on the basis of skeletal development to have been the late teens; if a male, he may even have been in his early twenties.

The presence of the familial-based fusion anomalies in HRO 3 and several of the children certainly allows for the likelihood that she was the mother, or close relative, of one or more of the affected children. Given that multiple pregnancies are likely to exacerbate maternal mineral deficiency, the occurrence of sacral anomalies in several of the children is somewhat suggestive that the adult female may have been their mother. Because it is possible that at least one other adult female (HRO 6) is present, it could be that all or some of the children were actually not the offspring of HRO 3. Unfortunately, no other adults could be assessed for congenital traits.

The relationship of the child buried in the talus slope (HRO 11) to the 10 people in the kiva is unknown; however, because individuals most likely were buried near their own households, this person's proximity to Structure 5-S may well represent some degree of relatedness. His or her prematurely fused capitulum could be another indication of kinship to the people in the kiva.

Illness usually affects people living close to one another—family members first, then other members of the community. This fact further supports the notion that at least several of the people in the kiva might have belonged to the same household. Although the composition of the typical household at Woods Canyon Pueblo is unknown, the archaeological assumption that kiva suites represent household architecture lends additional credence to this argument. All in all, the combination of the individuals' estimated ages at death and the frequent occurrence of familial traits allows for the possibility that many, if not most, of the children and the adult female (HRO 3) were closely related. The close—sometimes overlapping—placement of individuals in the excavated portion of the kiva and the discovery of a cranium just beyond the limits of the test trench indicate that there are more skeletal remains in the unexcavated portion of the kiva. If this is so, it is possible that multiple households were interred there, especially given that disease could easily spread through a compact and highly interactive community.

In conclusion, the number and mortuary treatment of the individuals found in Structure 5-S is a highly unusual circumstance in the prehistoric Pueblo archaeological record. In this case, it could well be that the story of these Woods Canyon Pueblo inhabitants involved epidemic illness or accidental death, affecting at least one family. The mortuary context suggests that their remains were treated with consideration and respect during what undoubtedly would have been a difficult and distressing event. It is also clear that other villagers continued to live at Woods Canyon Pueblo for some time after.

References cited | To borrow, cite, or request permission

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