Variation in Roasting Pit Design
Roasting pits comprise one of the most abundant and characteristic elements in the archaeological record of eastern San Diego County (e.g., Castetter et al. 1938; Cook and Fulmer 1980; May 1987; Wallace and Taylor 1958; Williams 2014). Some variability is evident in the manner in which the pits were constructed, and it may be hypothesized that some of this variability is attributable to differences among their makers in culturally defined mental templates for pit construction. If this is the case, archaeologically observable differences in pit design may prove useful in distinguishing boundaries between major ethnic groups, boundaries between smaller social units within the major ethnic groups, or cultural changes through time.
M. Steven Shackley (1984) reviewed the ethnographic evidence concerning agave roasting practices and discerned eight structural patterns among different ethnic groups in the Greater Southwest, southern California, and northern Baja California. He adduced archaeological evidence for two additional roasting pit designs within the Kumeyaay area, in the Mountain Springs area of western Imperial County and in Canebrake Canyon in eastern San Diego County. Shackley used data from the excavation of several pits in the Mountain Springs area to evaluate competing hypotheses that proposed that pit design was essentially idiosyncratic or that pit design was traditional within a group. His conclusion was that the “extreme architectural similarity” of the excavated pits supported the conclusion that group tradition predominated. Shackley also suggested that “perhaps the pits were traditionally structured across larger group sizes (e.g., Cahuilla versus Kumeyaay)” (Shackley 1984:149-150).
The ethnographic extrapolations and archaeological reconstructions of pit structures that Shackley presented suggest a number of empirical predictions about roasting pits in the San Diego area:
- Cahuilla roasting pits, according to ethnographic data from Bean and Saubel (1972), might be characterized by (a) a large rock in the center of the pit, probably not disturbed either by the opening of the pit to extract the roasted agave or by pit reuse; (b) a basal layer of rocks, underlying the coals, which would presumably not have been disturbed either by pit opening to remove the agave or by reuse, and which therefore would be found archaeologically in their original positions, with fire alteration on their upper sides only; (c) a second layer of rocks, overlying the coals but underlying the agave, which presumably would also not have been disturbed during agave removal but which would have been removed during any pit reuse and therefore would be shown, in the case of single-use pits, by rocks in place above the charcoal layer with only their lower surfaces fire-affected, and, in the case of reused pits, by rocks in place with multiple fire-affected surfaces and perhaps by rocks scattered on the ground surface near the pit; and (d) no rocks lining the vertical sides of the pit or forming a surface ring around it.
- Some Kumeyaay roasting pits, according to ethnographic data from Chase (1919), might be characterized by (a) rocks lining the base and sides of the pit, and presumably not disturbed by pit opening or reuse and therefore with only their upper or inward-facing surfaces fire-affected; (b) a layer of rocks, which overlay both coals and agave and which presumably would not be fire-affected and would be found scattered on the surface following pit opening; and (c) no large central rock, no surface ring of rocks, and no rocks in the pit with multiple fire-affected surfaces.
- The Mountain Springs archaeological model (Kumeyaay) suggests that pits might be characterized by (a) a basal layer of rocks underlying the coals, which presumably would not have been disturbed by pit opening or reuse; (b) a surface ring of larger rocks; and (c) no large central rock, no rocks lining the pit sides, no rocks overlying the coals, no surface scatter of rocks (except from subsequently disrupted surface ring rocks), and no rocks with several fire-affected surfaces.
- The Canebrake Canyon archaeological model (Kumeyaay) suggests that pits might be characterized by (a) a basal layer of rocks underlying the coals, which presumably would not have been disturbed by pit opening or reuse; (b) a large central rock, reaching to the surface, perhaps not disturbed by pit opening or reuse; (c) large rocks lining the pit walls and reaching from the base of the pit to the ground surface, also presumably not disturbed by pit opening or reuse; and (d) no rocks overlying the coals, no surface scatter of rocks, and no rocks with multiple fire-affected surfaces.
Dayle Cheever and Dennis R. Gallegos (1988) excavated a disturbed roasting pit feature in the Table Mountain area. The pit contained a single layer of rocks with evidence of burning on their upper surfaces; fewer than ten thermally-affected rocks were observed in the vicinity. This suggests a simpler pit design than any of the models discussed above.
Brian Williams (2014) reported on thermal features at site SDI-7074 in Jacumba Valley. One hundred and thirty features were classified as circular earth ovens (n = 53), oblong earth ovens (n = 13), ringed earth ovens (n = 2), amorphous earth ovens (n = 3), a circular earth oven and hearth (n = 1), hearths (n = 7), an historic-period hearth (n = 1), threshing floors (n = 3), and undetermined features (n = 47).
PROSPECTS
Future archaeological investigations may be able to document the range and relative frequency of different roasting pit designs and to use their geographical distributions to shed light on ethnic boundaries and settlement systems. Replicative experiments may shed light on the relative efficiency of alternative pit designs.