Human Remains and Mortuary Analysis DEBORAH E. BLOM AND MATTHEW S. BANDY T HIS CHAPTER PRESENTS the bioarchaeological analyses of the human remains from both Formative and Tiwanaku period contexts from Chiripa. The research compares mortuary prac- tices, inherited skeletal traits, demography, mark- ers of activity patterns and diet, and non-specific indicators of health for these populations to other populations from the Katari Basin and the Tiwanaku Valley in the altiplano and the Moquegua Valley on the western slopes in South- ern Peru. We discuss variation and continuity in several biological and cultural attributes for the inhabitants of the Chiripa settlements. While the recovery of a large mortuary sample was not one of the primary goals of the 1992 and 1996 excavations at Chiripa, some human remains were encountered in the course of the work. Human skeletal material was found in a total of sixty-two archaeological contexts. These contexts include formal burials (n=25), as well as bone dispersed throughout other contexts (n=37). For purposes of this study, a location and arrange- ment of remains that appear to have been the direct and intended result of human action are defined as "burials." This designation was prima- rily based on the relative completeness of the remains and evidence of a specially prepared, depositional environment, such as an excavated pit or an associated artifact assemblage. The addi- tional human bone, which was not found in burials, was generally isolated teeth or bone fragments whose location and arrangement seem to be the consequences of actions directed to separate and unrelated ends, such as bone dis- persed during construction episodes. In these terms, more than half of the human remains encountered (37 loci) represented isolated finds (see appendix 5). These isolated finds were encountered in fill and midden, as well as on cultural surfaces. Such a high frequency of disarticulated and dispersed human bone is not that surprising in a site like Chiripa, with a long and intensive occupation, and a very complex deposition history. It is also very common in the site of Tiwanaku. While others have looked at the patterning in isolated remains to infer status of individuals, this can only be done through in-depth study of the overall context and the demographic 118 Early Settlement at Chiripa, Bolivia profile (e.g., Storey 1992). At present, we will confine this discussion to remains identified as "burials."' Twenty-five occurrences of human remains were identified as burials during excavation. Burials were identified from all Formative and Tiwanaku phases of the site's occupation. No graves were encountered pertaining to later periods. Of the twenty-five identified burials, four could not be assigned to a definite phase, five were from the Early Chiripa phase, one each was from the Middle and Late Chiripa phases, one was from the Tiwanaku III (Qeya) phase, and thirteen pertained to the Tiwanaku IV and V phases. Therefore, our sample consists mainly of individu- als from the Early Chiripa and the Tiwanaku phases. For purposes of analysis, then, we will speak simply of Chiripa/Formative vs. Tiwanaku period (Tiwanaku IV and V) burials. It should be borne in mind, however, that the Chiripa burials mainly pertain to the Early Chiripa phase. All of the burials encountered from Chiripa were located in the Santiago area. In the Tiwanaku period, this area almost certainly was used as a formal cemetery due to the high concen- tration of burials. The large number of tombs of the Early Chiripa phase suggests a possible association between the burial locations and the semi-subterranean court (ASD- 18) located in this area. In the Late Chiripa phase on Monticulo a clear association is observed between public constructions and high tomb concentration. Such relationships may have been the case earlier in the Formative occupation as well. CHIRIPA PHASES The Chiripa Formative mortuary sample consists of 7 burials containing 7 adults and 7 subadults. Of the 6 adults for whom sex could be determined, all but one are female. However, this 5:1 sex ratio on a sample of six does not different significantly from a 3:3 ratio (Pearson's x2, p=0.22). One of the interesting aspects of this sample is the rather large number of multiple burials. Three of the seven burials relating to this period include the remains of multiple individuals. One (Locus 843) contained the remains of two subadults, both between the ages of 2 and 5 years. Another (Locus 1236) contained the remains of four individuals, one female, another adult, a juvenile, and a fetus. In addition, the third (Locus 1404) contained a relatively complete adult female, the crania of two additional adults (likely females) and fragments of a single subadult. This observation accords well with evidence from prior excavations at the site. Both Bennett ( 1936) and Portugal Zamora (1940, also Portugal Ortiz 1992) note the common occurrence of multiple inter- ments in House and pre-House strata. These include the relatively well known 'tumba de los amantes,' excavated by Portugal Zamora near House C on the mound. Because of the arrange- ment and differential preservation of the individu- als in the tombs, we cannot rule out the possibility that the tombs were reused over a period of time. In terms of tomb construction, a third of the Chiripa phase sample are cobble-lined cysts, with two-thirds being plain, unlined pits, or simple fossae. In addition, half of the tombs included large fragments of heavily worn, flat grinding stones, presumably employed as capstones. It is worth noting that virtually all of the grinding stones recovered in the excavations are from tomb contexts, suggesting that use as capstones was the typical form of discard for this artifact class. One of the tombs also contained remnants of vegetal material around the skeleton, reminiscent of Bennett's (1936) 'straw-lined' graves. The vegetable material may certainly be from mats or textiles or cords that were wrapped around the corpse; most likely of lake reed. This is a common occurrence in Formative period burials elsewhere (e.g., Cardona 1997). Because the preservation is not ideal, the position of the body was not always discernible. However, a fair amount of mortuary variation is observed, with positions including facing south or north and flexed, either seated or lying on the left side. One individual was lying face down. No tombs of the Chiripa period contained ceramic offerings. Again, this is in line with the observations of the previous investigators. The most common offering type, included in half the graves, is beads, mostly of sodalite. However, beads of a turquoise-like stone are also present. Browman (1984:125) has attributed this sodalite to the Cerro Sapo source in Cochabamba. If this is the case, the frequent occurrence of the material in Human Remains 119 Early Chiripa graves suggests 1ong-distance exchange networks prior to 1000 B.C. Finally, two of the Chiripa phase tombs contained traces of red pigment, which is likely hematite. TIWANAKU PHASES The assemblage from the Tiwanaku IV and V phases consists of thirteen burials, containing the remains of nine adults and seven subadults (including skeletal remains that were not removed from Locus 816). Of the adults complete enough to determine sex, three are female, with only one male identified. In contrast to the Chiripa period tombs, multiple interments seem more rare in the Tiwanaku period, with only one clear example encountered (Locus 536), a tomb containing an adult male and an adult female. In tomb form and construction, the Tiwanaku period burials exhibit considerable continuity with the earlier Chiripa period. Five of the II pit burials were stone-lined cysts, the remainder being plain, unlined fossae. Body positions include flexed and placed on the left side or seated. Grinding stones served as capstones in two of the burials, and remains of vegetal material occurred in two others, again indicating continuity of earlier practices. The Tiwanaku period burials are distin- guished from the Chiripa period principally in terms of their associated artifact assemblage. Only one Tiwanaku period burial contained beads (8% of the sample), in contrast to 57% of the Chiripa period burials containing beads. The principal mortuary offering in the Tiwanaku phases was ceramic vessels, which were present in 6 graves (46% of the sample). The types of vessels present include plainware ollas, keros, tazones, and one broken annular-base incensario. A considerable degree of continuity in the mortuary practices at the site of Chiripa can be seen through the course of almost two thousand years of occupation. The principal changes in the Tiwanaku period seem to be the decrease in the number of multiple interments, and the substitu- tion of ceramics for stone beads as the most common mortuary artifact type. This continuity of tomb construction practices could potentially indicate the existence of a 'local tradition' of the sort identified at Lukurmata by Bermann (1994) or possibly a long-term altiplano burial tradition. HUMAN OSTEOLOGY Standard inventories and demographic variables were recorded on all skeletons (see appendix 5). Determination of sex relied on pelvic and cranial morphology, while age-at-death estimates were based on dental eruption, epiphy- seal and suture closure, and changes in the pubic symphyses and auricular surfaces (Brooks and Suchey 1990; Buikstra and Ubelaker 1994; Lovejoy et al. 1985; Meindl and Lovejoy 1985; Scott 1979; Smith 1984; Suchey and Katz 1986; Todd 1921 a, b). In addition to that detailed in the description of the mortuary practices, the demo- graphic profile of the sample can be broken down as follows. If we assume that the individuals do not crosscut loci, the 37 occasions of dispersed bone found in 'non-burial' contexts contain at least 22 adults and 14 subadults. At this point we do not have a count of total contexts, but it is clear that human bone appeared only rarely in non- burial contexts. Of the 25 contexts defined as burials, we have 20 adults (nine females, four males, and seven of undetermined sex) and 17 subadults. Therefore, the 1992 and 1996 excava- tions uncovered an estimate of 73 individuals represented by partial to complete remains. PALEOPATHOLOGY AND ACTIVITY PATTERNS In order to analyze the changing patterns of health over time, three non-specific markers of health and disease (linear enamel hypoplasias, porotic hyperostosis, and periostosis) were mea- sured. Comparison samples used include Tiwanaku IV/V contexts from the site of Tiwanaku, Late Formative and Tiwanaku IV/V contexts from sites in the Katari Basin such as Kirawi (CK65) and the Tiwanaku V phase site of Chen Chen in the Moquegua Valley of Southern Peru (figure 33). Several of the individuals from Chiripa displayed porotic hyperostosis. Porotic hyperosto- sis is evidence of the body's response to anemia during childhood (Stuart-Macadam 1985; Stuart- Macadam and Kent 1992). These porous lesions on the skull vault are the result of pressure from within on the outer table of the skull. The pressure is caused by marrow hyperplasia in the attempt to 120 Ear/l Settlement at Chiripa, Bolivia FIGURE 33 Map of the Southern Central Andes, and the locations from which the human data come. dispel anemia by producing additional red blood cells. Although porotic hyperostosis does not affect the heavier bones of adults as a rule, the healed childhood lesions are visible in adult remains. Because genetic anemia is very rare in New World populations, the anemia is very likely due to iron deficiency. While it was first assumed that the deficiency was the result of an iron poor diet, Stuart-Macadam and colleagues (ibid.) have argued that more convincing causes are chronic infection and parasites. Since these pathogens often cause bleeding and/or diarrhea, they impede iron absorption since the food moves through the digestive tract too rapidly. Porotic hyperostosis was present in 40-50% of all the Chiripa remains (Chiripa phases 3/7=42.9%; Tiwanaku phases 4/ 4=50%). There was no significant difference over time (Pearson's X2=0.032, df=1, p=0.858). This is similar to the Katari Basin Formative sites (3/ 7=42.9%) and lower than the other Tiwanaku period samples analyzed (Katari Basin 11/ 14=78.6%; Tiwanaku 24/39=78.6%; Moquegua 243/315=77.1 %). While many of the samples are small, and in-depth analyses considering demo- graphic variables as factors could not be per- formed, these data indicate a general trend of samples from later period contexts suffering from a higher prevalence of anemia. This trend is not apparent, however, in the Chiripa sample. Another indicator of general disease is the presence of linear enamel hypoplasias on the teeth (Goodman and Rose 1991; Rose et al. 1985). Humani Remains 121 These defects in the enamel are caused by inter- ruptions in enamel formation during tooth devel- opment. These lesions generally are secondary to episodes of acute childhood illness, such as those resulting in a high fever. Incidences of linear enamel hypoplasia were recorded on the upper central incisors and the lower canines. Of the individuals from burials with these teeth present, linear enamel hypoplasias were present approxi- mately 60% of the time in both the Tiwanaku (3/ 5=60%) and Chiripa (4/7=57.1 %) period samples from Chiripa. This is similar to the percentages in the Katari Basin sites (Formative phases 5/ 8=62.5 %: Tiwanaku phases 8/14=57.1 %), higher than the site of Tiwanaku (9/21=42.9%), and lower than in Moquegua (89/119=74.8%). Linear enamel hypoplasias, like anemia, indicate that at Chiripa there was a relatively average amount of illness for these time periods and stability over time. Likewise, periostosis, a bony growth often linked to generalized infection (Mensforth et al. 1978; Ortner and Putschar 1985), was equally common in both time periods at Chiripa, affecting 50% of the adults with observable lower limb bones (Chiripa phases = 1/2; Tiwanaku phases = 2/ 4). Healing or healed periostosis indicates reha- bilitation from the infection while active growth indicates ongoing infection or healing of a trauma. Once again, we are dealing with a small sample, and the difficulty in interpreting these results is magnified by the fact that one case is a localized patch, likely due to trauma; one is an active case; and two cases are slight and moderate, mixed (active and healing) reactions. Specific descrip- tions can be found in appendix 5. This differs from Dale Hutchinson's (1997) observations at other sites with late Chiripa and Tiwanaku compo- nents on the Copacabana Peninsula. He reports an increase of lesions over time. In fact, he sees an increase over time of most indicators of illness. The collections of human remains from the site of Chiripa indicate a pattern of stability in health over time, in spite of a changing social and physical environment including an intensification of agriculture and increased urbanization. The analysis of arthritis is also consistent with the trend of stability over time. Repetitive activities, such as carrying loads, scraping skins, using specific weapons, and planting and tending agricultural fields, often impact bone as osteoar- thritis of the joint surfaces (Kennedy 1989; Merbs 1983; Tainter 1980). In the Chiripa sample we see no clear pattern of arthritis which can be used to infer behavior. The sites of arthritis include all major joints, and all adult individuals with more than three joint surfaces present from Chiripa's Chiripa (n=4) and Tiwanaku (n=3) period contexts are affected. Individuals with three or fewer joint surfaces observable did not always exhibit arthri- tis, but whether it was present on the missing joints is not known. Clearly a larger sample is needed to learn more about activity patterns at Chiripa. In contrast to the paleopathological vari- ables, we do see a change in time in dental caries and attrition, suggesting that these two temporal groups at Chiripa were engaged in different activities and diets. Tooth wear and caries can be used as indirect means of observing dietary changes (Larsen et al. 1991; Powell 1985; Turner 1979). The collection of attrition data combine the techniques of Murphy (1959a, b), Scott (1979), and Smith (1984), as recommended by Buikstra and Ubelaker (1994). Recording of caries data follows a modified version of Moore and Corbett (1971). Dental caries are commonly associated with a high carbohydrate diet, while tooth wear generally indicates dietary items which are tough or fibrous, such as charqui (dried meat) and potato skins, or the introduction of abrasive materials in the food, such as grit and sand that may enter when grinding or perhaps when drying food items. The Chiripa people overall had a fairly severe amount of attrition (tooth wear).' The wear is greater than that for all other samples analyzed except the Katari Basin Formative contexts (Chiripa Formative x=8.5 (n=6); Chiripa Tiwanaku x=8.8 (n=4); Katari Basin Formative x=9.0 (n=3); Katari Basin Tiwanaku x=7.3 (n= 14); Tiwanaku x=6.0 (n=20); Moquegua x=5.1 (n= 113)). The representation of 5-year-interval age groups across these samles was not signifi- cantly different (Pearson's C =46.443, df=49, p=0.577). Within the Chiripa collection, the Tiwanaku period sample has slightly more wear than during the earlier Chiripa phases. The adults from Chiripa phases were also affected with caries in their molars or premolars 57.1 % of the time (4/7 individuals, 6 /54 teeth=l 1 %), while none of the 122 Early Settlement at Chiripa, Bolivia later Tiwanaku phase individuals were affected (0/ 4 individuals [0/34 teeth]). The caries data alone would normally indicate that the diet was com- posed of fewer carbohydrates over time. It should be noted, however, that we observed slightly more wear and fewer caries over time. Because the same types of food items that cause wear will reduce the caries rate because they will mechani- cally clean the teeth, this is consistent with either tougher or harder foods increasing in the diet or more abrasive inclusion entering the food, possi- bly though grinding or drying. Other data that were not presented here include measurements of crania, postcrania, and teeth, fractures and other trauma, cranial modifica- tion through the intentional use of pads or boards, and evidence of lip plug wear and the use of the teeth as tools, as well as post-mortem dismember- ment, defleshing and display. While no incidences of lip-plug use, teeth-as-tools, or post-mortem modifications of bone were seen in the Chiripa collection, the data are presented for each indi- vidual in appendix 5. BIOLOGICAL DISTANCE Finally, genetic relationships can provide more information about local continuity and change. Many features in bone can be used to measure genetic relatedness between groups of individuals. Because the preservation of the sample is generally poor and not ideally suited to metric studies, non-metric trait analysis of teeth and bone (Berry and Berry 1967; Corrucini 1972; Finnegan 1978; Turner et al. 1991) is one of the best ways to look at these relationships in the Chiripa remains. This analysis involves inherited, discrete variations in bone, for example, one vs. multiple foramina, bony bridging within a large foramen, or the presence of ossicles, which are small bones in the sutures of the skull. Techniques for the collection and analysis of the 'non-metric trait' data followed that described in Blom et al. (1998). When we compare the two temporal groups in the Chiripa sample, the dis- tance measurement is not significant. Therefore, there is a high probability that the distance mea- sure could be lower. Until we have a larger sample, we have no reason to expect that the two were not from a homogeneous population. In the Formative period comparison, Chiripa is quite distant from the other sites. A distance measure- ment of 0.331 is observed between Chiripa and the nearby Katari Basin sites of Lukurmata and Kirawi. The distance between the samples from the Chiripa phases and the Moquegua (Huaracane) Formative phase is also high. Overall, a pattern exists in the Formative of Chiripa as a separate group. In the Tiwanaku phases, the data indicate a pattern of gene flow between the Tiwanaku components of Moquegua, Tiwanaku, and the Katari Basin sites. Chiripa, on the other hand, does not appear to clearly follow this pattem. Again, likely because of the small sample size, the Chiripa measurements are not significant for those with the altiplano groups. The p-values are lower than those between the two temporal components of Chiripa are, but in order to explore this further, a larger sample is necessary. In summary, mortuary practices, paleopathological indicators, and genetic data suggest few differences between Chiripa and Tiwanaku phases at Chiripa. However, archaeo- logical and dietary data indicate change. The lack of change in health with an increase in urbanism is unexpected. These two factors are almost invari- ably associated with an increase on the amount of pathological lesions seen on bone. One possible explanation for this would be an 'osteological paradox' (Wood et al. 1992). It is possible the individuals were so compromised that they were dying as soon as they contracted a disease, so there was no time for the bones to be effected. How- ever, many of these individuals were dying in their 40s and 50s, essentially 'old age' for these popu- lations. A more likely hypothesis is that other social or environmental conditions were insulating the populations from the common problems of urbanization. The mortuary sample from Chiripa is small. Therefore, we cannot rule out the possi- bility that change did occur which could be detectable with a larger sample. These findings are suggestive, however, and should provide a foundation for continued study in this area.