A high priority of the current research was to determine whether a stockade wall was present on the Sugar Creek site. Trenches were placed specifically for the purpose of identifying such a structure. Whenever soil anomalies extended all the way across the base of the trenches, blocks were expanded to determine whether the anomalies ended fairly quickly or extended in long linear lines indicative of a stockade wall. No evidence of a stockade-type structure was identified on the Sugar Creek site.
However, in the course of expanding the trenches to identify stockade walls, several features suggestive of domestic structures were identified. The strongest evidence for a domestic structure was recovered from Feature 5, where there was what appeared to be the corner of two wall trenches that are similar to a structure identified at the Cox's Woods site (Redmond and McCullough 1996: Fig. 21). The structure at Cox's Woods extended approximately 6.0 m by 3.0 to 4.0 m wide (outside dimensions) and was subrectangular in shape. The edges of the wall trenches were irregular in both plan and profile, with the straight segments of the wall trench varying from 50.0 to 75.0 cm wide. The depth of these segments was about 25.0 to 35.0 cm below plowzone. Sporadic posts were identified along these trenches with large deep-set posts found on the "front" corners. The trench segments have a similar morphology to the wall trenches identified in Feature 5, which averaged about 52.0 cm across and 38.0 cm deep. A possible interpretation for the construction of such a structure is that logs were laid in the wall trench as sill plates. Next, saplings could have been attached to the insides of the log and bent over, pushing against the inside of the sill. The structures could then be covered with skins or bark and soil banked against the bottom.
Caution needs to be exercised in accepting the structure identified at Cox's Woods until other Oliver Phase domestic structures can be exposed fully in plan view, since it was so different from others structures identified in the lower or middle Ohio Valley. Typical Fort Ancient structures are made of a single line of wall posts spaced about 20.0 cm to 60.0 cm apart (Cowan 1987; Heilman and Hoefer 1981; Turnbow and Sharp 1988). On the other hand, Mississippian type domestic structures are covered with wattle-and-daub, with walls set in narrow trenches and narrow sapling poles set through the base of the trench (Black 1967; Muller 1986). The structure at the Cox's Woods site and features 2 and 5 lacked any evidence of wattle-and-daub construction.
The limited excavations conducted at the Sugar Creek site were inadequate to confirm that the structure type identified at the Cox's Woods site was also present on this site. Wide-area excavations around the domestic structures at the Sugar Creek site are needed to confirm this. If the middens with high densities of domestic refuse (for example, pottery and bone) are associated with domestic structures (as we believe to be the case with the Unit K midden and Feature 5), then it is likely that the midden scatters identified on the surface represent other areas of domestic structures (or possibly pit house structures).
By examining the general distribution of the surface midden scatters (Figure 4.10) and noting the locations of the middens around units K, E, and H, one can get an idea of the site structure (that is, the intrasite patterning) at the Sugar Creek site. In general, the distribution suggests a loose clustering of houses around a relatively open plaza. The shape appears to be somewhat oblong, with a possible opening toward the east. This pattern suggests a settlement that is not tightly nucleated as are the Cox's Woods site (Redmond and McCullough 1996) and the Clampitt site (Redmond 1994b). The distribution of the midden scatters and the excavation results indicate that no stockade wall was constructed at 12 Jo 289.
As with the Bundy-Voyles site, this site is located to provide easy access to a wide range of resources useful to the Oliver Phase peoples. An upland physiographic area is west of the site, easily accessible across the rock-bottomed shallows in Sugar Creek. The site itself is on flat, highly productive farmland between Sugar Creek and the Blue River. These ecological borderlands were excellent locations for prehistoric sedentary settlements, because the variety of resources allows for easy logistical mobility as opposed to residential mobility. Species recovered during the excavations, such as elk, deer, black bear, beaver, raccoon, fox squirrel, coyote, turkey, and fish, attest to the utilization of a variety of natural environments, from open clearings to closed-canopy forests and riverine environments (see Appendix 3). Like other Oliver Phase peoples, the inhabitants of the Sugar Creek site were farmers. The botanical assemblage at 12 Jo 289 exhibited poor preservation and a low density of floral remains. However, maize was identified in all of the samples examined, as well as one fragment of a common bean and sumac, bramble, and hickory nut. For a complete list of the taxa identified, see Appendix 2.
Along with subsistence data, the current research also attempted to assess seasonality of Oliver Phase village sites. Both the faunal and floral analyses were unable to determine this with certainty. However, the high density of cultural deposits, the presence of domestic structures, and the evidence of agricultural activities strongly suggest a settlement that was occupied for a large portion of the year, as at the Bundy-Voyles site.
Three carbon samples were submitted for radiocarbon assay (see Appendix 4). The sample from Feature 2 yielded an intercept date of AD 1470, which is probably an anomalous result, most likely due to the sample's small size and the feature's location beneath a historic campground. The other two dates overlap with those from both the Bundy-Voyles site and the Crouch site but appear to be slightly earlier, tentatively suggesting an early fourteenth-century date for the occupation. Again, information from one site is not adequate to demonstrate diachronic change for an entire cultural manifestation, especially since the tentative earlier positioning of the Sugar Creek site is associated with both similarities to and differences from other Oliver Phase sites.
These differences and similarities can be tested for temporal significance in later investigations. First, this site exhibited a roughly circular structure that has been identified at some Oliver Phase sites (Dorwin 1971; Redmond 1994b; Redmond and McCullough 1996) but not at others (Redmond and McCullough 1993; this report, Section 3.) Second, this site does not appear to be as tightly nucleated as the Clampitt site or the Cox's Woods site and apparently lacks a stockade wall. A similar loose type of village structure, also probably lacking a stockade wall, was identified at the Bowen site (Dorwin 1971). While the presence or absence of a stockade wall at the Bowen site can never be known (units were not placed specifically to identify such a structure), the disposal patterns on fortified Oliver Phase villages have shown that trash accumulated along the inside edge of the stockade wall (Redmond 1994b; Redmond and McCullough 1993). After five seasons of excavations that were placed over areas of high artifact density, it seems likely that if such a structure were present at the Bowen site, it would have been identified. Obviously, the distribution and temporal significance of stockaded Oliver Phase villages is still not well understood: Is the presence of stockade walls related to spatial proximity to valuable resources or did their presence become a necessary feature only during the mid-fourteenth century?
Finally, as discussed earlier, the ceramic assemblage at 12 Jo 289 reflects both continuity and variation, allowing the use of seriational techniques to order this site in a relative chronology. The vessels appear to have a lower frequency of decoration than those from the Cox's Woods site and a higher frequency than those from Bundy-Voyles. Also, a higher frequency of cordmarked rims and Fort Ancient-influenced vessels were recovered from the Sugar Creek site than from Bundy-Voyles.
While this site appears to predate the possible Vincennes Phase-like site at 12 Jo 5 (this report, Section 5), any conclusions are necessarily tentative. Only a small amount of the surface area was excavated at the Sugar Creek site, nor is it known how representative Sugar Creek is of sites located along the secondary drainages in the till plain region.