METHODS

Demographic Criteria Assessment

The analysis utilizes traditional approaches for the assessment of the general physical characteristics of the age at death, sex, and stature of the individuals in the sample (e.g. Stewart 1979; Brothwell 1981; Bass 1987; Isçan and Kennedy 1989). It consists of a five phase procedure, as follows: 1) a general description and confirmation of the human origins of the material; 2) the estimation and segregation of the minimum number of individuals; 3) the determination of sex; 4) the estimation of age, and; 5) the estimation of stature.

General Description of the Skeletal Material

The human remains were separated from non-human remains, and the individual bones and bone fragments were identified and classified. Observations were made on their condition and preservation, and any evidence of post mortem damage to the material was noted.

Minimum Number of Individuals

During this phase, the minimum number of individuals (MNI) was estimated, and the skeletal material was separated into discrete individuals. An a priori assumption for the determination of the MNI is that an individual's remains will not be spread out over more than one room. This assumption was fundamental in determining and maintaining chronological control during the initial excavations, and is followed in this research. Determination of discrete individuals was based on diagnostic bones indicative of at least one individual. These include, for example, the sternum, sided limb bones, and crania.

Determination of Sex

Observations regarding the sex distribution of the sample were compared with metrical and morphological standards developed to delineate the sex of an individual within a broader dimorphic distribution. Determination of sex is based on direct observation of sexually dimorphic criteria. Adult males and females differ in both general size and shape, and this variation is reflected in the skeletal anatomy. The determination of the sex of subadult skeletons is more problematic than those of adults. However, the use of an inclusive approach combining morphological and metrical characters has been tested with known (sexed) series of modern skeletal material, and Jackes (1992:195) has reported an accuracy rate of 85-95 percent.

Morphologically dimorphic features include such characteristics as the shape of the pelvic girdle and the crania (Buikstra and Ubelaker 1994:16). The pelvic girdle is the most sexually dimorphic region of the skeleton, and it can be used to determine sex with a high degree of accuracy (Bass 1987). The sexual dimorphism of the pelvis is primarily the result of reproductive mechanics, and is not readily apparent until adolescence. Beginning in adolescence, the female pelvis expands relative to its height, while the male pelvis continues along trajectories established at birth (Buikstra and Ubelaker 1994:16). Holcomb and Konigsberg (1995:113) have reported a greater than 66 percent accuracy utilizing the greater sciatic notch of the ilium to determine the sex of known infant skeletons, and these authors have suggested that genetic morphological traits such as these may offer further avenues of research for the sexing of immature skeletons.

Adolescent characters can begin to develop as early as nine years of age. Detection of female aspects in a very young os coxae would indicate a high probability of accuracy (Coleman 1969). Conversely, male patterns observed in an adolescent os coxae are to be considered inconclusive, as the remains may represent a female as yet undeveloped (Buikstra and Ubelaker 1994:16). In such cases, additional corroborative evidence must be sought to make a firm determination of sex. In addition, male crania may retain a gracile, female form during early adolescence. Hence, detection of male characters in adolescent skeletal material is suggestive of its masculinity (Buikstra and Ubelaker 1994:16).

Other metrically dimorphic attributes include the maximum diameter of the femur head and the maximum facial breadth (Buikstra and Ubelaker 1994:16). This form of variation is often continuous in nature, with males being longer or larger than females. Discriminant function formulae are used to segregate individuals based upon patterns of sexually dimorphic growth and musculature (e.g., Snow et al. 1978; Stewart 1979; Richman et al. 1979; Kelley 1979; Steele 1980). The formulae are typically applied with the most success to the long bones. Muscular development is also sometimes evaluated through these formulae, but accurate measurement of such development is problematic (Workshop of European Anthropologists 1980; Brothwell 1981; Isçan and Miller-Shaivitz 1984; St.Hoyme and Isçan 1989).

An abbreviated listing of the metrical and morphological characters utilized in sexing the skeletal material is presented in Table 2. A complete list of all characters recorded and the organization of analytic methods is presented in Appendix A (Stewart 1979; Brothwell 1981; Isçan and Loth 1986; Isçan and Miller-Shaivitz 1986; Bass 1987; Bennett 1988; Ubelaker 1989).

Estimation of Age at Death

The determination of age relies on the assessment of the physiological age of the skeleton, as opposed to the chronological age of the individual. The physiological age is based upon relative growth patterns, and is hoped to give an accurate estimate of chronological age, but environmental, nutritional, and disease stresses often cause changes in the skeleton which will mask the true age of the individual. In addition, the accuracy with which age can be estimated varies inversely with the age of the individual at death. In younger years, with age being estimated primarily upon observed developmental changes, more precise estimates are possible, whereas in older individuals, age estimates are more often accomplished via the observation of degenerative changes, which offer less accuracy.

Age determination can be accomplished through many means, and a holistic analysis of all possible age-related attributes is best for an overall estimate. Some of the more typically utilized attributes include:

1. Dental Eruption and Occlusion 2. Cortical Bone Histology 3. Cranial Suture Closures 4. Postcranial Epiphyseal Unions 5. Pubic Symphyseal Face Morphology 6. Age-Related Degenerative Conditions 7. Phase Changes in the Sternal Rib 8. Potpourri

1. Dental Eruption and Occlusion Age estimates are based on the age of eruption of the deciduous and permanent dentition. This method is useful in age estimates of up to about 15 years. The third molar (wisdom tooth) erupts after this time, but is so variable in age of eruption, if it erupts at all, that it is not a very reliable age indicator. See Bass pp. 289-290 for an illustration of Ubelaker's eruptive phases, noting the standard deviations. Occlusal wear has also been offered as an indicator of age, but this has been shown to be highly inaccurate, especially in archaeological context, where high-grit content diets (such as from the use of natural stone mano and metate) can wear down the occlusal surface of the tooth by the end of puberty - see Bass pp. 286-87, after Brothwell (1965).

2. Cortical Bone Histology Kerley (1984) developed a system of aging based on osteon counts taken from midshaft long bone sections. This process involves counting the number of whole osteons and osteon fragments (which increase in number with age), and nonhaversion canals and the percentage of circumferential lamellar bone in the cortex (which decreases with age, completely disappearing around age fifty). These estimates are taken from the outer one third on the cortex, with a normal light microscope in four fields at 100X. A percentage estimate is calculated, and what is sought after is the rate of osteon turnover or replacement. These percentages are plugged into either a regression formula or a pre-calculated age\profile chart. Kerley has obtained a reliability of almost 90% with a standard deviation of +/- 5 years, with the best correlation coming from the fibula, then the femur and tibia. 3. Cranial Suture Closures This method bases age upon the degree of closure, union or ossification of the cranial sutures. These methods have until recently been considered inaccurate, but Meindel and Lovejoy (1985) have introduced new evidence to indicate parietal ectocranial sutures are reliable indicators of age over 40 years. In addition, Mann et al. (1987) have offered the four maxillary sutures and their rates of closure as reliable age estimators - see Bass pp.47-48.

4. Postcranial Epiphysial Unions (see handout #1) Endochondral bones of the postcranium form via the union and ossification of cartilaginous bridges between growing bones. This process can be seen to occur along a growth algorithm, and can be used to estimate age at death. Handout #1, as well as Bass (1987), lists some of these locations of epiphyseal union, as well as the approximate age ranges for which these unions occur. This data can be used on a union/non-union basis, and McKern and Stewart have define five grades of epiphyseal union: unobservable (0), beginning (1), active (2), recent (3), and complete (4), and these offer a possibly more accurate estimate of age.

5. Pubic Symphyseal Face Morphology (see handout #2) The pubic symphyseal face in the young is characterized by an undulating surface, such as the crennulated surface of a typical non-fused epiphyseal plate. This surface undergoes a regular progressive metamorphosis from age 18 onwards. The phase system diagrammed in the handout, was developed by Suchey and Brooks for the male pubic symphysis.

6. Age-Related Degenerative Changes in Skeletal Features Many non-pathogenic conditions such as certain expressions of arthritis and osteoporosis become more prevalent and pronounced in old age, and can be used to give corroborative evidence in the determination of age. These occurrences are not entirely reliable in themselves, however, as injury and pathological expressions of these conditions can mimic the degenerative condition. An illustrative case can be seen in the osteophytic growths of the vertebral body (via osteoarthritis). These growths form on the outer margins of the centra, and Steward (1958) has computed an age progression histogram for humans over 21 years based on the percentage of extra-central lipping as a function of age for the lumbar and thoracic vertebra - see Bass pp. 20-21.

7. Phase Changes in the Sternal Ribs Iscan and Loth have developed a system of age estimation based on sequential changes at the sternal end of the fourth rib. These changes are similar to those that occur on the pubic symphyseal face. They are of a specific morphological nature and occur on the costochondral joint between the rib and sternum. They consider that these phases are not as subject to variation due to sex, pregnancy and activity patterns as is the pubic symphyseal face. See Bass (pp. 135-142) for photos of Iscan and Loth's phases, with the general progression illustrated as an increase in the depth of the articular depression and the degenerative fragmentation, thinning and increased porosity at the edges of the articular surface over time.

8. Potpourri: a. Note that generally females are more advanced than males with regard to physiological age, being about two years advanced at puberty, five years at maturity, and seven to ten years in old age. b. The sacroiliac joint undergoes changes in morphology similar to those at the pubic symphysis, Lovejoy et al. (1985) offers a phase system based on these morphological changes. c. Krogman (1949) offers a system of aging based on transillumination through the scapular body to chart the occurrence and amount of atrophic (thinning) centers, basically, the more that are present, the older the individual. d. Various radiographic analysis techniques focus on age related changes to interior bone structures, such as at the costo-chondral juncture, the metaphyseal plates of the long bones, and Walker & Lovejoy's (1985) radiographic analysis of trabecular bone involution in the clavicle. e. Bass (1987) and Ubelaker (1989) offer age estimates based on long bone lengths, but these have a wide range of variation even within a single relatively homogenous population.

Estimation of Stature

Estimation of stature is based on extrapolation formulae derived from populational averages of long bone lengths. These averages are computed from known populations, resulting in some degree of error when applied to a sample from a prehistoric population. The formulae utilized are based on long bone lengths and corresponding statures derived from data from historic Native American populations (Steele and McKern n.d.; Steele 1980). Bones utilized in this phase of analysis include the humerus, radius, ulna, femur, tibia, and fibula.