The Sexual Division of Labor at the Sanders Site (41LR2), Lamar County, Texas

This study examined evidence of stress on human bone in order to reconstruct the sexual division of labor for the prehistoric Caddoan culture represented at the Sanders site (41LR2). Specifically, the repeated action stress seen in degenerative joint disease was quantified in order to infer culturally prescribed and habitual actions. Overall, the division of labor at the Sanders site was somewhat unusual among agricultural societies. Males from the Sanders site had evidence of greater stress loads. In particular, agricultural activities appear to have regardless of sex. Although the Sanders site burial population was believed to be of high social status based on burial location within the mound and grave goods, degenerative joint disease was high and activities evident from patterns of stress could not be associated with a life of leisure.


INTRODUCTION
This study examined evidence of stress on human bone in order to reconstruct the sexual division of labor for the prehistoric Caddoan culture represented at the Sanders site (41LR2).
Specifically, the repeated action stress seen in degenerative joint disease wu quantified in order to infer culturally prescribed and habitual actions.
Overall, the division of labor at the Sanders site was somewhat unusual among agricultural societies.Males from the Sanders site bad evidence of greater stress loads.In particular, agricultural activities apppear to have regardless of sex.Although suggest nearly continuous use of the site up to historic times.
Excavation of the site took place in the summer of 1931 under the direction of A.T. Jackson and B.B. Gardener from the University of Texu.Jackson fully excavated Mound 1 and trenched Mound 2 and the midden area (Figure 1).
Jackson recovered some 60 individuals from two locations in Mound 1.One group wu scattered over the top of the mound, whereas the other group wu located on the eastern slope of the been participated in by all, t" -• the Sanders site burial population wu believed to be of high social status hued on burial ..... """.-:;-:--::::-====-location within the mound and grave goods, degenerative joint diseMe wu high and activities evident from patterns of stress could not be Msoeiated with a life of leisure. ,

THE SANDERS SITE
The Sanders site is located in the northwestern corner of Lamar County, Texu, on Bois D'Arc Creek, a tributary of the Red River.The known site consists of two mounds and a low ridge of midden located on the eastern bank of the creek.Absolute dating within the Caddoao region bu been limited and the Sanders site is no exception.The Sanders site bu not been radiocarbon dated, although in general it is believed to be representative of the earlier manifestations of the Caddoan culture, and wu initially placed in the Gibson upect.Artifacts recovered from the site mound (Figure 2).There wu no difference between the two groups in burial form or offerings.
Both groups contained single or multiple burials, and all individuals were extended supine or on their side.Additional burials were recovered from later excavations and potbuoter backdirt.In total, 29 adult females, 25 adult males, 18 adults of indeterminate sex, and 20 children were disinterred.

DEGENERATIVE JOINT DISEASE
Degenerative joint disease (DJD) is often termed osteoarthritis, although the latter represents a more generalized condition.DJD is caused by the wear and tear of repeated action stress on joints and by the aging process.In the joints, the first response to stress is cartilage deterioration, followed by a bone reaction.The reaction of bone is the only evidence of DJD found in the archaeological record.Bony outgrowths (termed osteophytosis and lipping) are the first and most common response to cartilage destruction (Miller 1985:392).Bone involvement, therefore, only occurs after stress in the afflicted joint when it has had some time to transpire, unless it is the result of a stress related, single event injury.Thus, reconstruction of past activities based on DJD patterns can only give an indication of the activities participated in, not the duration of time spent engaged in the activity.
Since DJD is linked to repeated actions, it has been used to reconstruct activity patterns for skeletal populations (Angel 1966;Merbs 1983;Pickering 1984;Miller 1985;Dutour 1986;Striland 1987;Bridges 1990).The term activity patterns has been taken from Merbs' (1983) study to mean habitual or culturally prescribed and sanctioned postures and practices.In this study, osteophytosis has been used to reconstruct some of the activity patterns engaged in by the Sanders burial population.

METHODS
All 55 adults curated at the Texas Archeological Research Laboratory were examined for presence of joint surfaces.Juveniles were not examined because DJD is primarily associated with repeated actions occurring over an individual's lifetime.Of these 55 adults, 43 were selected for use.Of these individuals, 14 were female (two over 50 years of age), 19 male (three over 50 years of age), and 10 were of indeterminate sex.
Bone lipping was scored from O to 2, in increments of 0.25, with 0 representing condition absent, 1 representing moderate expression, and 2 representing severe expression of the condition.
Scores were recorded on 113 distinct joint surfaces.Separate articular surfaces were examined rather than regions.For example, the ankle was not scored as one region, but the tibia -8-articular surface for the talus, fibula articular surface for the talus, superior articular surface of the talus, and the calcaneal tuberosity were individually scored.By doing this, a more accurate account of specific activities could be gained.
The joints highly effected by preservational and recovery bias were omitted.These included carpals, tarsals, and distal phalanges.Other joint surfaces were lumped into group categories.These included ribs, and proximal phalanges of the hands.
Although the Sanders site was initially chosen for its good state of preservation, only 23% of the possible (if all individuals were complete) joint surfaces were present.Very few individuals were complete.This meant that examination of specialists within the population was impossible.Examination of individuals by sex, however, was feasible because individuals represented by specific joint surfaces could be compared with other individuals of the same sex.From this, behavioral differences between the sexes could be inferred.

RESULTS
The incidence of DJD was high among the Sanders population, with 34 (78 % ) of the individuals showing evidence of affliction.In general, males were more afflicted than females, exhibiting both higher relative frequencies and higher average scores.
Activities with mean scores of 1 or more.This category contained all activity patterns with average scores for all involved joint surfaces of 1 or greater.In this way, activities that were performed most often and/or were most stressful could be inferred.
Ground clearing, planting, and soil tilling.DJD of the back was very common in the Sanders population among both sexes.This is not uncommon for agricultural populations and the patterns exh ibited by both females and mal es were equivalent to other Mississippian related agricultural populations (Pickering 1984).Low to the ground activities such as soil preparation and ground clearing were indicated in the high levels of lower back stress seen in the lumbar and lower thoracic vertebrae.Such activities would require much bending and a stooped posture that tends to strain the lower back.
-) ) I J Because the Caddo were known maize agriculturalists, lower back stress was expected.There were many artifacts in the Sanders site assemblage that indicated that ground clearing and soil tilling were taking place, such as bison scapula hoes and stone hoes.
Both lumbar and lower thoracic body scores were examined in the determination of these activities and both were well represented.Differences between the sexes were slight, indicating that both sexes participated in these activities.Both had higher scores in the lumbar region (females averaged 1.12 and males averaged 1.33; Figure 3), than in the lower thoracic (females averaged 1 and males averaged 0.94).
Differences between the sexes were slight in this region and may imply that females and males participated in some different activites that resulted in lower back stress.Carrying loads on the back may have accounted for these differences.
Carrying loads on the back.Osteophyte development on the sacroiliac facet has been shown to result from carrying heavy loads on the back (Trotter 1937).Males from the Sanders site had bony growths at this location, as well as high scores in their lumbar vertebrae.High scores in the sacroiliac joint suggest that weight from the load carried was directed downward through the back.
There was also evidence that females carried loads on their backs, although in a different manner than males did.Carrying loads on the back can also result in stress in the lower thoracic region of the back (Merbs 1983).In this case weight from the objects transported would be dissipated across the lower back, rather than being directed downward.
The different patterns suggest that females and males used different mechanisms to support items on their backs.This may have been a result of dissimilar objects that were normally transported.
Supporting loads •on the head.Use of the head fo r supporting objects has resulted in cervical vert~brae spinous process bone lipping (Pickering 1984:78,97), anterior subluxation of the body (Scher 1978:97), body compression (Pickering 1984: 164), locked facets (Scher 1978:97), and dorsal body compression (Merbs and Eul. er 1985:389).The upper cervical vertebrae were The DJD evidence is consistent with balancing loads on the top of the head in an unaided manner.Balancing a load in this way pushes the head forward and into the chest, creating pressure and consequent osteophyte development along the anterior margins of the cervical vertebrae, as well as anterior subluxation.Osteophyte development within the Sanders female population followed this pattern.None of the individuals examined exhibited osteophyte development along the dorsal margin, as would be expected with the use of a tumpline (Merbs and Euler 1985:389).It is interesting that supporting loads on the head was not documented in either ethnohistoric or traditional accounts.
Long distance walking, running, or dancing.
The posterior calcaneal tuberosity often produces a large bony outgrowth in response to long distance travel by foot (Dutour 1986:222; Figure 4).Both females and males had high scores at this locality, and both with a considerable degree of unilateral expression.Females had a more bilaterial distribution with mean scores of 2 on the right and 1.33 on the left, while males averaged 1.5 on the right and O on the left.The unilateral distribution seen in males may simply reflect the low number of observations on the left side (four among the males, JO among the females) .
The unilateral distribution was not found as strongly among males in other joint surfaces associated with these activities.The femur patella articular surface, femur tibia articular surface, and thel)atella lateral and medial articular facets all exhibited mean scores above one for both sexes (Figure 5).Results from these surfaces are all indicative of much movement at the knee, which is also associated with walking, running, and/or dancing.
The material culture from the Sanders site suggests that the members of this population may have traveled great distances to participate in exchange.The Caddo from the Sanders site may also have traveled great distances on foot to gather and/or hunt.Both females and males may have participated in these activities.Ethnohistoric accounts described women accompanying men on hunting expeditions, although only females were said to gather (Swanton 1942).Unilateral distribution irt the calcaneal tuberosity, particularly among the males, may have resulted from frequent or stressful par- The rugged looking marginal bone growth is osteophytosis on the patella anicular -10- ticipation in dances that required specific foot movements balancing more weight on the right heel than on the left.
Paddling.Evidence of paddling is found in the elbow region where lipping scores are greater on the trochJea than on the capitulum of the same hand.This type of degeneration is found on both sides of the body when individuals are responsible for steering in both directions, as in paddling a canoe or rowboat.The movement is pr.imarily flexion and extension, with rotation occurring when only one paddle is employed.This was found at the Sanders site among males with mean capitulum scores of 1.05 and mean trochlea scores of 1.19.' Oddly, scores were slightly higher on the left side.
Since the Sanders site is located on Bois o• Arc Creek, approximately three miles south of the Red River, it is likely that the river was used as an important transportation route.
Activities with mean scores of 0. 75 or greater.Lowering the level of stres severity allowed the analysis of activities that were performed less often and/or were less stressful.
Sewing, basket weaving, ceramic production, and shell working.The high level of stress seen in the palmar side of the proximal phalanges of the fingers among females (0.94 average) suggests that females were involved in activities associated with fine finger manipulation of objects.It has been directly associated with sewing (Merbs 1983: 155).Bone awls were found in burials of both sexes, and pottery disks that may have functioned as spindle whorls were recovered from the Sanders site midden.The level of stress in males was much less (0.25 average).
Ceramic manufacture and basket weaving both require extensive manipulation of small objects and the use of the fingers as tools.Both of these activities were attributed to females ethnohistorically (Swanton 1942).
Other possible activities include flint knapping and shell work, both suggested from the artifact assemblage.Further activities that would require extensive finger manipulation include husking maize, although this activity was documented traditionally as performed by both females and males (Newkumet and Meredith 1988).
Wood cutting.Osteophyte development seen in the olecranon process of the right ulna (0.83 mean) indicated that males may have performed wood cutting.The spicule-like growth observed on the olecranon process results from wood cutting and other activities where the arm is horizontal and the elbow flexed (Dutour 1986:222).Many stone axes were recovered from the Sanders site in the midden.
Hide scraping.DJD data indicate •that scrapers were held palm-down and used in a back and forth, trough-like motion among males.This type of movement affects the distal humerus, proximal ulna, and marginal areas of the radial head (Merbs 1983:154,156;Miller 1985).The sternal and acromial facets were involved as well.Artifactual data indicate that hide scraping took place.
Carrying loads with the arms bent.Males show evidence of carrying heavy loads with their arms bent.This activity is suggested by radial tuberosity degenerl;ltion (Dutour 1986:222).However, degeneration of the radial tuberosity is the result of tlexion of the elbow, hence it could also have been the result of paddling behavior.

DISCUSSION AND CONCLUSION
The activities described above are those recognized as most stressful to the Sanders burial population.The amount and frequency of DJD Volume IV Number 3 does not necessarily correspond to the amount of time engaged in a particular activity, although time is often a primary factor in stress induction.Many activities were not discerned here that were expected (bow and arrow use, for example).The activities not exhibited by stress markers on the bone (DJD) were therefore either not emphasized culturally or were not stressful to individuals who may have performed such activities, either rarely or often.
The activities emphasized by the Sanders population supported an agricultural subsistence economy with a great deal of travel, most likely due to participation in an extensive trade network.
Sex and status differences in food production and related activities were not as expected.Participation in these activities may seem surprising since the Sanders population has been considered an elite group.Participation in agricultural activities was found in both females and males, including planting, cultivating, ground clearing, and hoeing.However, participation in food preparation and hunting techniques was not clarified.
The patterns of DJD also may suggest a limited degree of differentiation between elite and nonelite among the Caddo.Unfortunately, a village population has not yet been analyzed for comparative purposes.Since so many individuals were afflicted, either a considerable amount of time was spent in the activities outlined, or high stress loads were endured by the population.
Stress loads were examined for females and males.It was expected that females would exhibit greater stress loads than males as has been found in Mississippian popu lations (Bridges 1987;Pickering 1984).The opposite was true for the Sanders population.Males exhibited higher stress loads in terms of greater mean scores and greater relative number of articular surfaces afflicted.
The DJD evidence may suggest a ritual importance of maize agriculture.This would explain participation in activities associated with agriculture, but limited food preparation and hunting activities.Ritual importance of agriculture is also suggested by Newkumet and Meredith's (1988) depiction of men beginning each day by walking the fields of maize.
This study was unable to determine precisely which members of the population endured greater workloads than others because the representation of articular surfaces was low.Instead , individuals were grouped by sex in order to determine a sexual division of labor.The sexual division of labor does not appear to have been strict because individuals of both sexes practiced many of the activities associated with agriculture.However, some activities not directly associated with agriculture appear to have been sex specific.
Volume IV.Number 3 highly affected by DJD among the Sanders population.

Figure
Figure 3. lumbar venebra from an adult male (397) with osteophyte on the superior margin of the body (drawing by Kaylee Stallings).

Figure 5 .
Figure 5. left distal femur of an adult male (787).The rugged looking marginal bone growth is osteophytosis on the patella anicular

Figure 4 .
Figure 4. Adult male's (787).right calcaneus with large hook-shaped bony lip on calcaneal tuberosity.The bone in this region is normally smooth (drawing by Kaylee Stallings).