Summary counts of the artifacts recovered from the Bundy-Voyles site are presented in Appendix 1. A total of 27,209 pieces of material, excluding faunal and floral remains which were recorded by weight, were cataloged by the Indiana University field school. The artifact assemblage was dominated by Late Prehistoric materials: All of the ceramics recovered consisted of Oliver Phase, or Heaton Phase (Tomak 1983), vessels; and all of the projectile points were triangular forms, except for one blade section. The following subsections summarize and provide extended descriptions for three major categories of cultural materials: sherds; lithics; and bone and antler.

GO TO:[Sherds][Stone Tools][Bone Tools}

Sherds

The pottery assemblage was divided into four categories: rim sherds, neck sherds, body sherds, and pipe fragments. (See Appendix 1.) The rim sherds category consists of pottery fragments from the top of the vessel that retain enough surface area to distinguish the lip portion. Larger sherds that retain the rim and either the neck or body portion are also classified as rim sherds. A neck sherd is a pottery fragment that is missing the rim portion but includes enough curvature to identify it as a Late Prehistoric constricted orifice jar. Neck sherds were separated because on Fort Ancient-like vessels, the neck is the preferred field for decoration. Body sherds were any fragments without a rim or neck portion. Pipe sherds included any fragments of smoking pipes. A fifth category, indeterminant, contains items too small and fragmentary to be accurately distinguished and is not included in the percentage totals or the following discussion. (For examples of rim and neck sherds, see plates 3.2b, 3.3a, and 3.3b.)

Appendix 1 summarizes the frequencies of rim sherds, neck sherds, body sherds, and pipe fragments both by unit and by feature proveniences. Table 3.2 presents the frequency of rim, neck, and body sherds for temper and type of surface treatment and/or decoration. A distinction needs to be made between surface treatment and decoration. Surface treatment is the result of finishing the surface of the vessel through smoothing and texturing and can include brushing, burnishing, cordmarking, fabric roughening, or smoothing (Rice 1987: 136-37). Decoration refers to the "embellishment of a vessel beyond the procedures used in forming the clay mass into the final vessel shape and finishing its overall surface" (Rice 1987:144). In the initial cataloging system, surface treatment and decoration categories were a combined classification, since each sherd has a surface treatment and may have additional decoration. In cases where the sherd exhibited decoration, this was noted instead of the surface treatment. Individual attributes have been recorded for each sherd larger than four square cm, making separate observations possible for decoration and surface treatment for lip, rim, neck, and body sherds. An analysis of this data is beyond the scope of this report, but the observations are part of ongoing research by the senior author (McCullough n.d.b).The temper of the great majority of vessels found at the Bundy-Voyles site was crushed stone- or grit-tempered. Approximately 96.9 percent of the sherds recovered were grit-tempered; only 2.3 percent were shell-tempered. The remaining sherds contained minor amounts of other types of temper, or the temper type could not be ascertained (indeterminant). This high percentage of grit temper along with minor amounts of shell temper is typical of other Oliver Phase sites that have been investigated (Dorwin 1971; Redmond 1994b; Redmond and McCullough 1996).

Decorated ceramics at the Bundy-Voyles site consisted of designs executed by incising or by cord impression, including cord-wrapped dowel impressions, which are typically associated with Oliver Phase sites from the East Fork of the White River (Redmond 1994b; Redmond and McCullough 1996) and the upper West Fork of the White River (Dorwin 1971; McCullough 1991, 1992). The pottery with the incised decoration resembled the vessels commonly associated with the Anderson Phase (Griffin 1943; Essenpreis 1982) in southern Ohio and southeastern Indiana. The vessels with cord-impressed designs exhibited a wider geographical distribution but were often found on vessels with collars and were similar to types from the southern Great Lakes region. Decoration from the Bundy-Voyles site conformed to what would be expected from an Oliver Phase site, but the frequencies of occurrence demonstrated some interesting changes.

The surface treatment/decoration table (Table 3.2) indicated that well over 80 percent (81.8 percent rims, 82.3 percent necks) of the rim and neck sherds recovered from the Bundy-Voyles site were undecorated (categorized as plain or cordmarked). This relative lack of decoration was in contrast to vessels recovered from the Cox's Woods site, where about 80 to 85 percent of the rims carried some type of decoration. At the Sugar Creek site only about 26 percent of the rims carried intentional decoration (Table 4.4). While only 7.5 percent of the necks were decorated, the frequency of rims with cordmarking from the body onto the lip was much higher at the Sugar Creek site than at the Bundy-Voyles site. At Bundy-Voyles, the frequency of plain or smoothed rims (smoothed-over-cordmarked is underrepresented in this analysis) was about 38 percent, compared to 12 percent at the Sugar Creek site.

Cord impression (including cord-wrapped dowel impressions) appeared to dominate the decorated rim and neck sherds at the Bundy-Voyles site. Middle Fort Ancient-type designs, such as guilloche (or sherds more frequently classified as thin- or broad-line-incised because this motif is difficult to determine on fragmentary pieces), were a relatively minor decorative element. They constitute only 20 percent of the decorated rim and neck sherds (compared to 33 percent at the Sugar Creek site). However, it is important to note that neck portions carrying designs that extended onto the shoulder, such as in the Fort Ancient Tradition, were classified as a body sherd, which may have underrepresented the importance of these designs.

Some sherds found in the Oliver Phase ceramic assemblage are classified as "fabric roughed," exhibiting a roughened surface that appears to have been created with a wad of cloth (Dorwin 1971). Sherds with this type of surface treatment often can be overlooked during the cataloging process. As part of the individual attribute analysis (McCullough n.d.b) and to ensure consistent surface treatment observations, body sherds larger than four square cm were observed for temper type and surface treatment. Body sherds were chosen because these are usually not decorated, making only surface treatment (and temper) observations possible. Table 3.3 presents the results of this analysis and demonstrates that the fabric-roughed surface treatment was only slightly underrepresented in the cataloging process. However, with a frequency of 4 percent, this type of treatment was not a significant part of the assemblage. Cordmarked, grit-tempered ware was by far the most frequently encountered. Interestingly, fabric roughening was more popular at the Sugar Creek site (see Table 4.5), with a frequency of over 10 percent. Small quantities both of cord-marked and plain shell-tempered ware were still represented in the size-sorted sample. In terms of morphology, Figure 3.23 demonstrates a few examples of the rim profiles from vessels recovered from the Bundy-Voyles site. An initial examination of profiles suggests that the vessel morphology at the site includes a higher frequency of forms with a rim-neck juncture that folds outward sharply, or is strongly everted. (For examples, see all but the center specimen in Figure 3.23.) Also, on many vessels, a thick, flattened shoulder is present; at least one vessel exhibits decoration on the shoulder. The shouldered rims on these specimens are thick enough to suggest that the vessel was manufactured in an inverted position.

While refitting of pottery sherds was not a formal aspect of this investigation, sherds that join together but are from separate proveniences suggest contemporaneity of deposition and may reflect degrees of social interaction such as has been attempted at the Incinerator Site (Heilman and Hoefer 1981; Nass 1988). Even though only a small amount of time was spent looking for ceramic refits, the analysis substantiated that the features on the site coexisted with the overbank midden, that segments of the overbank midden were deposited at roughly the same time, and that the burial on the site was interred at or after the time the features were in use (Table 3.4).

From the above descriptive analysis, it is clear that the Bundy-Voyles site ceramic assemblage differs from other Oliver Phase sites, even though the basic styles are the same. These differences deserve additional investigation and may be useful in developing a temporal sequence for Oliver Phase pottery. Specific kinds of decoration aside, the most noticeable difference is the number of vessels exhibiting decoration. Changes in the frequency of decoration may indicate diachronic changes in the ceramic assemblage. The data suggest that additional temporal indicators could be the frequency of cord-marked or plain rims, the frequency of a fabric-roughed surface treatment, and the morphology of the rim and shoulder. An individual attribute analysis of selected ceramic assemblages from south central Indiana is currently being conducted by the senior author (McCullough n.d.b), and these observations will be evaluated.