Social Structure

Bison hunting is an activity that can be accomplished by single hunters, small groups, or larger cooperative aggregates. For many time periods, there is a positive correlation between the number of animals killed and the number of hunters participating in the kill. Increases in time and energy required for preparation of a hunt are often met by increasing the number of participants. For example, the construction of a corral capable of containing and holding 30 bison requires the cooperative efforts of a large number of people. Additional coordination is necessary to accumulate and induce the necessary number of animals to follow the constructed drive lanes into the corral, dead end arroyo, or over a cliff. To meet the requirements of labor, Historic Plains groups timed the large scale kills to enable aggregation of two or more bands of hunters. Timing the events made allowances for harvest schedules, bison aggregation cycles, and the seasonal needs of the groups. Planting, tending, and harvesting crops took precedent over bison hunting. Planning a large scale bison hunt during the time when bison herds were dispersed or absent from a region doomed the endeavor before it even began. Groups that relied on large amounts of stored foods for winter survival timed hunts with the onslaught of cold weather that would preserve the meat and took advantage of berry collection schedules so that pemmican could be made and stored for the winter. Any or all of these considerations affected the timing of large scale bison hunts on the American plains. This is particularly noticeable on the northern Plains where the majority of hunts occurred during the fall and winter seasons.

General Characteristics of Communal Hunts

Are all large scale bison kills the result of communal hunts? This is a difficult question to answer. As we track hunting techniques back in time, it becomes even more difficult to address this question. Part of the problem rests in how we define "communal hunting". Driver (1990:12) provides three traits of communal hunting:

(a) Participation by more than two hunters (usually many more than this).

(b) Active cooperation between hunters such that they work together, as opposed to passive cooperation in which hunters agree not to interfere with each other’s activities.

(c) A system of hunting that requires all hunters to participate in a previously conceived plan.

This is a very broad definition of communal hunting that could conceivably make the cooperative venture of killing a single bison a communal hunt. Missing from this definition is the social aspect imposed by assembling hunters that normally do not hunt together. Many historic communal hunts involve participants from different bands. Such aggregations allow ritual feasting and dancing, betrothals, trading, and the renewal of old acquaintances and obligations.

Communal Hunting at Cooper

Is there evidence that the kills at the Cooper site were communal? Following the three criteria of Driver (1990) it is easy to argue that each of the Cooper kills required a) many more than two hunters b) working in cooperation, and c) following a preconceived plan of action. George Frison (1978) has demonstrated that an arroyo bison kill minimally requires one set of hunters to round up and drive the animals and another set to be in position to kill them. Evidence that the kills followed a set plan is rather obvious considering the coordination that must have existed between drivers and killers. But was the arroyo in which the kills occurred the only one manned by hunters or were there hunters also positioned at the head of adjacent gullies? The most convincing evidence that this gully was the premeditated terminus of the drive is seen in the painted skull. If our reconstructed sequence leading to the painting of the skull and its subsequent trampling by animals of the Middle Kill is correct, then the ritual was performed in the predetermined locus of the kill. The possibility that hunters were positioned along adjacent gullies is not precluded by this situation, but that would entail additional people.

Communal kills could be performed by members of the same group. And, while this fits the criteria, many kills documented in the ethnographic and early historic sources coincide with the aggregation of more than one group or band. In these instances, the large scale kill provided a means to feed a large number of people drawn together for social as well as economic reasons. This inter-group or social aspect of "communal" kills adds another dimension not covered by Driver (1990). This fourth dimension is very difficult to detect in the archeological record and is assumed through analogy. One aspect of the social dimension is the heightened ritual activity often including feasting, dances, curing ceremonies, and ceremonies honoring the dead. The painted skull at Cooper could fuel an argument of heightened ritual activity. But did similar rituals precede all hunts? If so, then this alone does not indicate heightened ritual activity.

Other lines of evidence for the amalgamation of more than one group might be drawn from the lithic raw material of recovered projectile points. Southern Plains Folsom assemblages are usually dominated by implements made from Edwards Plateau chert from central Texas. Points made from this material are often found over 400 km from the source. North of the central Texas area, additional lithic sources include Tecovas jasper and Alibates agatized dolomite. Tecovas Folsom points are seldom found, but Tecovas flake tools have been recovered in eastern New Mexico and the Texas panhandle in areas near the source. Alibates is a more common component of some Folsom assemblages. It is used in the manufacture of projectile points and flake tools. Like Tecovas, the Alibates quarries are in the Texas panhandle. Folsom assemblages in the Alibates area often include Edwards Plateau chert points and tools, yet sites nearer the Edwards Plateau do not include Alibates points or tools. In fact, only one Alibates Folsom point is known from a site near the Edwards Plateau. North of the Alibates quarry area, the number of Alibates Folsom points increases; yet Edwards points are still common. The mechanism that allows the northern distribution of Edwards Plateau chert but inhibits the southern use of Alibates is not currently understood. It has been proposed that Folsom groups leaving the Edwards Plateau area headed west and north in an annual round. As discussed previously, if the supply of Edwards Plateau chert is exhausted before the group returns to the Plateau, then alternative sources of stone, such as Alibates, are sought. One problem of this model is the paucity of discarded Alibates points and tools in the Plateau area. Are these materials jettisoned as the Edwards Plateau chert sources are approached?

An alternative explanation is that somewhere between the Edwards Plateau and the Alibates quarry is a boundary between two Folsom territories. Group composition and territorial boundaries were undoubtedly fluid and it is probable that territories overlapped. If this were the case, then it is possible that sites containing both Edwards and Alibates assemblages mark localities where members from both bands came together, perhaps for communal hunting. The distribution of sites containing both Edwards and Alibates assemblages suggest the location of this overlap or aggregation zone.

The social or fourth dimension of communal kills must incorporate a temporal factor since the participating groups must have an agreed upon time to meet. If bison hunting was a planned event at aggregations, then the scheduling of the aggregation should coincide with known bison movements/activities to ensure the presence of sufficient numbers of animals for successful hunts. The ability to determine the season of kill from bison dentition provides the means to determine the timing of aggregations.

Due to increased numbers of people to feed at aggregation sites, kills would probably contain larger numbers of animals than each group would normally procure on its own. Hence, the kills should contain large numbers of animals and should indicate a specific season of the year.

The distribution of Folsom bison kills of known seasonality indicate a cluster of large late summer/early fall kills in NW Oklahoma and adjacent portions of Texas. The lithic assemblages of these sites contain points and tools made from both Edwards and Alibates raw materials. The Middle Kill at Cooper is the only aberrant assemblage in this cluster. Here, all seven recovered Folsom points are made from Edwards Plateau chert. Sampling error related to the small size of the lithic sample may be affecting the results.

In addition to a temporal component, aggregation also relied upon selecting a meeting place that could be found by all groups concerned. Such a meeting place would allow access to sufficient amounts of food (i.e. bison), water, and wood. The meeting place may not have been a specific site, but may have been a broader area such as a stretch of river. The groups would find each other by moving along a river course or divide. It follows that areas that provided the requisite resources would be used repeatedly, although not necessarily for every aggregation or in successive years.

Recurrent use of an area, however, would allow repeated use of landforms where successful kills were made. The Cooper arroyo, with its three kill events, would be a predicted outcome of such a situation.

The Cooper kills indicate the events occurred every 3 to 5 years. If aggregations were annual, then intervening years were spent at other locations or at least alternate arroyos were used. The Lipscomb site, 65 km from Cooper, could be a kill or kills associated with another aggregation locus.

Clues from Butchering and Transport Decisions

The affect of multi-group communal hunting on butchering and transport decisions is difficult to assess. The butchering patterns from small Folsom kills such as those seen at Lubbock Lake entail carcass dismemberment and meat stripping from the entire animal skeleton. The larger kills including Cooper and Lipscomb seem to rely upon a butchering process that selected high quality meat packages with little skeletal dismemberment. On the one hand, the switch from intense butchering to gourmet butchering--or from heavy butchering to light butchering (Reeves 1990)--could be the result of simply the number of animals at a kill. A Folsom group would require X amount of meat and bison product for subsistence. If the number of members in the group is held constant, and thus the meat/products requirement held constant, then the differences in butchering techniques could simply relate to the number of animals killed: The more animals killed, the less extensive the butchering necessary to fulfill the meat requirement (X) of the group. In this scenario, the amount of bison product obtained from the heavy butchery of 5 animals would equate to the amount of bison product obtained from the light butchering of a larger number of animals. The question is, how many gourmet-butchered animals equates to 5 extensively butchered animals? Are the kills at Cooper in excess of this number to suggest killing by more than one Folsom group? The meat yield from a bison is roughly 45% of its body weight. Hence, roughly 50% of a bison carcass is usable foodstuff. If gourmet butchering removed half this amount from a carcass, then it would require twice as many animals to meet the minimum requirement of bison products. The meat from 10 gourmet butchered animals would equal the meat from 5 extensively butchered animals.

The location of cutmarks on the bones from the three kills at Cooper may provide a means to quantify the amount of meat obtained from these kills. Cut marks have similar distributions in all three kills. Such marks cluster on the ribs, humerus, femur, tibia, and thoracic vertebrae. Marks are occasionally found on the radius/ulna, lumbar vertebrae and pelvis. The single carcass quantity of meat and fat associated with the bones from Cooper displaying cut marks (ribs, humerus, femur, tibia, and thoracic vertebrae) totals approximately 58.5% of that from a complete bison carcass. The breakdown of meat and fat from Emerson’s (1993:142) food utility index model was used to figure the percentage of edible product from each skeletal element.

Using the 58.5% figure, the gourmet butchering of thirty-plus animals from a kill at Cooper or Lipscomb yields totals 3.5 times greater than the amount of meat and fat from the extensive butchering of 5 animals from a Lubbock Lake kill. Hence, even the selective butchering of animals at Cooper produces a surplus capable of feeding three Lubbock Lake size Folsom groups. This exceeds the expected situation of a Lubbock Lake size group employing a selective butchering technique because of an overkill.

Although the switch in butchering technique from extensive to gourmet may be related to the number of animals killed, the large kills at Cooper and Lipscomb could provide for the aggregate needs of either three Lubbock Lake groups through gourmet butchering or six Lubbock Lake groups through extensive butchering. Although not conclusive, this discussion anticipates the presence of more than one group at the Cooper site on the basis of butchering technology.

The shifts in butchering techniques could also relate to transport decisions. When the residential group establishes a camp adjacent to a kill--such as Lubbock Lake and Waugh--transporting meat on bone is not a problem. A multi-group aggregation site that is established prior to a kill would require that meat/products be transported to the camp. A meat stripping technique that leaves the bones at the kill is an efficient technique in this situation. When used in conjunction with a gourmet technique, butchering time is reduced to the extent that a processing camp is not needed prior to transporting the meat masses to the aggregation site.

No evidence of a processing camp has been found in the vicinity of the Cooper kills. Of course, the extensive modification of the landscape by the Beaver river over the past 10,000 years may have removed any evidence. However, it should be noted that no debitage or tools suggestive of cooking or processing beyond the use and on-the-spot sharpening of flake butchering knives was recovered.

The gourmet butchering and lack of a nearby camp suggest that a habitation site was some distance from the kills. Locating such a site may be impossible. A large habitation site placed along the river for ease of access to water and timber would not have been preserved. The flushing of flood plain deposits by the downcutting of the river valley would have destroyed any site in this setting.

In summary, the three kills at Cooper meet the criteria set forth by Driver (1990) for communal bison kills. However, if the criteria are extended to include the social aspect of aggregation of more than one group, then it must be demonstrated that the Cooper kills were the result of cooperation between hunters from two or more groups. Evidence that more than one group is present may be drawn from the presence of two spatially exclusive lithic types in the tool assemblages, seasonal redundancy in the timing of the kills, utilization of a gourmet butchering technique, repeated use of the same gully, and the lack of a processing area adjacent to the kills, suggesting that the filleted meat masses were transported to a camp at some distance from the kills. The painted skull may manifest heightened ritual activity associated with aggregated groups. Without confirmation of the existence of a multi-group aggregation site in Folsom times, however, this must remain but one possible explanation of the available data.


skullbut.jpg (5711 bytes)