by Sarah E. Price

Jump to: References

In addition to the traditional assignment of sherds to types, the Madison Park ceramic analysis concentrated on three questions. First, since Dead River and Hope Hull phase ceramics are not usually recovered from the same site, why are both present in the Madison Park assemblage? In essence, do these types represent distinct Late Woodland phases? Second, what functions/uses did the ceramics serve at Madison Park? Third, did vessel function change through time? The following chapter presents a discussion of the Madison Park ceramic assemblage in regard to these research questions.

Culture Historical Types

Dead River Phase
The early Late Woodland Dead River phase is known only from the area of Montgomery, Alabama, where the occasional recovery of Weeden Island ceramics, or locally produced Weeden Island-inspired ceramics, indicates contact (direct or indirect) with that coastal society to the south (Chase 1998a:69). A single radiocarbon date from Feature 9, Zone C of the Madison Park site coincides with the estimated date range of the Dead River phase (ca. AD 700-800; Chase 1998a:69). This date was obtained from carbonized wood and does not correspond with two other dates from the same feature. The current hypothesis to explain this date represents mixing in the feature fill and is not associated with the main occupation or actual filling of the pit.

Kilby Plain is one of two principal pottery types associated with the Dead River phase. This sand tempered ware occurs in a jar form, described by Chase as “an amphora-shaped vessel with a deeply indented rim which formed a collar” (Chase 1998c:68). These jars have a sharp inflection at the neck forming a flaring everted rim. There are often deep tool marks or finger-smoothed grooves in the inflection (Figures 5-1 and 5-2). Vessel walls are almost straight sided until meeting a rounded or conoidal base (Figure 5-3). The exterior surface is smooth, often with pebble smoothing marks (Chase 1998c:48). Kilby Plain vessels vary greatly in size.

Figure 5-1. Example of Kilby Plain rim (FS 583).

Figure 5-2. Example of a Kilby Plain jar rim with tooling (FS 598).

Figure 5-3. Reconstruction of a Kilby Plain jar.

 The other principal diagnostic pottery type associated with this phase is Dead River Red Filmed, which are deep hemispherical bowls with round bases and straight sides. The sand tempered ware is similar to that of Kilby Plain, but surfaces are smoother and often burnished. Interior and exterior surfaces are covered with reddish-orange film (Figures 5-4 and 5-5). This type subsumes an earlier defined type, Dead River Plain. Neither Kilby Plain nor Dead River Red Filmed vessels are decorated (Chase 1998c:48).

 

Figure 5-4. Example of Dead River Red Filmed rim (FS 382).

 
Figure 5-5. Reconstruction of a Dead River Red Filmed bowl.

Hope Hull Phase
The Hope Hull phase has been thought to have derived from the Dead River phase after AD 800 and to have continued until around AD 1100 (Chase 1998a:72, 1998b:26). Hope Hull phase sites are known only from Montgomery, Elmore, and Lowndes counties in central Alabama, and appear to be centered on the junction of the Coosa and Tallapoosa rivers, mainly based on evidence from site 1EE1 (Chase 1998b:12). Chase interpreted the Hope Hull phase as a direct successor of the similarly geographically confined Dead River phase, two phases of a single cultural tradition, based largely on the stratigraphic relationship of similar ceramics observed at site 1MT48 (Chase 1968b, 1998a:70), and the transitional nature of some early Hope Hull ceramics found at that site (Chase 1998b:20).

Chase defined two principal Hope Hull ceramic types, Adams Plain and Montgomery Red Filmed. In his last writings on Hope Hull, he further refined those ceramic types to define early and late varieties of each.

Adams Plain is a sand tempered ware on large, plain, conoidal jars (Figure 5-6). Chase defined Adams Plain, var. Kennedy, the early variety, as having a “somewhat grainy” surface texture, smooth inside and out but not burnished, with pebble “streaks” visible on the exterior (Figures 5-7 and 5-8). The rim flares and necks are tooled and slightly constricted (Chase 1998b:20, 1998:48). From his descriptions, it is difficult to visualize how Adams Plain, var. Kennedy, differs systematically from Kilby Plain. The later variety, Adams Plain, var. Adams, has a harder and more compact paste than var. Kennedy, and occurs on larger vessels (e.g., 60 cm tall and 30 cm diameter mouth). Vessel exteriors are pebble streaked and often burnished. The rim is everted, but the sharp neck inflection seen on Kilby Plain is replaced by a gradually curved neck. Lips are rounded, squared, and sometimes beveled (Chase 1998b:21, 1998c:48).

Figure 5-7. Example of Adams Plain jar rim (FS 822).

 
Figure 5-8. Example of Adams Plain rim (FS 617).

 Montgomery Red Filmed occurs on a sand tempered ware that has reddish-orange slip on interior and exterior smoothed, often burnished, surfaces. This type is distinguished from Dead River Red Filmed by vessel form, which is invariably an inverted rim bowl in Montgomery Red Filmed. The early variety, Montgomery Red Filmed, var. Montgomery, is undecorated while the later Montgomery Red Filmed, var. Froggy Bottom has fine lined (“scratchy”) rectilinear or curvilinear incising or engraving on the exterior surface (Figures 5-9 and 5-10). Punctations (often as short slashes) co-occur with this shallow incising, frequently in incised zones. Bowl shape is often elongated, not circular. A rare form of Montgomery Red Filmed, var. Froggy Bottom bowl, known colloquially as a “pillow pot,”  (Figure 5-11) has four appliqué knobs or projections at the vessel shoulder (Chase 1986a, 1998a, 1998b:21, 1998c:49).

Figure 5-9. Example of Montgomery Red Filmed, var. Froggy Bottom rim (FS 234).

Figure 5-11. Reconstructed “Pillow Pot” vessel.

Madison Park Ceramic Assemblage
The Madison Park ceramic assemblage consists of 279,807 pottery sherds: 258,576 recovered from units and 21,231 from features. Three subsets of ceramics were derived from the Madison Park assemblage based on: (1) size grading all sherds; (2) an attribute analysis of all sherds, with particular detail recorded for rims; and (3) a study of finishing procedure, wear, and wall thickness (referred to as the FWW analysis) of a small sample of sherds. The first step in this pottery study was determination of size grades for each provenience. Sherds (n=94,094)—defined as pieces of pottery larger than ½ inch—were separated from sherdlets (n=185,713) by use of a ½-inch screen. Sherdlets were then counted and set aside. Only sherds received further ceramic analysis. The detailed size grade data, which include sherdlets, were used to investigate site formation processes (see Chapter 3).

The following attributes were recorded for all sherds: surface treatment, presence or absence of surface film, decoration, and vessel part (i.e., body [n=89,641], rim [n=4,012], or base [n=441]). For rim sherds (n=4,012), additional attributes were recorded when possible: temper inclusions, vessel form, vessel diameter, rim thickness, rim profile, rim angle, tool impressions, and lip type. For the FWW analysis, a small sample (n=2,582, including rims [n=918], body sherds [n=1,613], and bases [n=51]) was measured for vessel wall thickness and observation of wear (fire clouded, erosion, and carbonization).

 Results
The following presents results of the various analyses conducted on the Madison Park ceramics, including vessel form, function, decoration, use and wear. Only sherds, not sherdlets, were used for these analyses, with the exception that only the rim (Appendix F) and use-wear (Appendix G) subsets consider decoration and wear analyses.

 Vessel Form
A key goal of the rim analysis was determination of vessel form. Three broad categories of vessels were recovered at Madison Park: bowls, jars, and plates/pans. A bowl is traditionally defined as a vessel with unrestricted or restricted orifice and a height varying from one-third the maximum diameter of the vessel to equal the diameter of the vessel. A jar is taller than its maximum diameter and has a restricted orifice, usually with an everted rim or collar. Plates or pans are very shallow and always have an unrestricted orifice (Rice 1987:216).

Vessel form is best determined by rim sherds; therefore the rim analysis subset was used for the following discussion. Vessel form could not be determined for 108 of the sherds in the rim analysis subset. Bowl rims (n=2,735) are the most frequent vessel type found in the assemblage at just over 70 percent of the assemblage. Jars account for under 30 percent (n=1,141), and plates are the rarest of the three forms (n=28). The counts and percentages of the three broad vessel types appear in Table 5-1.

Table 5-1. Vessel forms (based on rim sherds only).

Vessel Size
Orifice diameter and rim thickness were recorded when possible for all rim sherds in order to estimate vessel size (n=4,012). The width of the flare on Kilby Plain jars was also measured to determine if flare width was reflective of vessel size. Orifice diameters were then plotted for each vessel type to determine if there were bimodal distributions, indicating small and large vessels within each form (Figure 5-12). Kilby Plain flared jars cluster unimodally between 22 and 46 cm. Dead River Red Filmed bowls also cluster bimodally between 16 and 24 cm and 27 and 33 cm, with another possible cluster between 39 and 48 cm, which might be indicative of smaller and larger bowls within the assemblage. Adams Plain curved jars exhibit a unimodal plot, clustering between 26 and 38 cm. Montgomery Red Filmed inverted bowls have a unimodal distribution that clusters between 12 and 47 cm. The distribution of the orifice diameters indicates that there are some differences in vessel size between types, but also between phases. Jars overall are larger, presumably a reflection of function.

Vessel thickness is directly related to overall vessel size. Wall thickness also varies depending on the intended use of the vessel (Rice 1987:227). Taller vessels need thicker walls to support the weight of wet clay while drying; if the walls are too thin, then the vessel will collapse. On the opposite end, thinner walls disperse heat better, cook food faster, and reduce the amount of time and fuel needed to process food. When vessel rim thickness was examined for each of the types, there were no statistically significant differences between any of the vessel forms. In sum, all jars had slightly thicker walls than bowls, which was expected. Vessel rim thickness is positively correlated with orifice diameter; in other words, as diameter increases, so does thickness of vessel.

 Vessel Function

Based on vessel form, orifice diameter, and wall thickness, we can make several inferences based on the Madison Park ceramic assemblage in relation to site activities. Vessels with unrestricted orifices, such as the Dead River Red Filmed Bowls, were probably used for processing or serving. Open orifices make getting foodstuffs in and out of a vessel easy, and present the contents visually, as in the case of serving. The presence of red filming on almost all of the Dead River phase bowls indicates that restricting permeability was important to the function of these vessels. Slipping on either the interior or exterior surfaces can prevent moisture from getting in or liquid contents from seeping out. Slipping of the Dead River phase bowls indicates that for whatever foods they contained, permeability was an issue (Figure 5-13). Also, this vessel type may have been used for long term storage. Dead River phase bowls all have a rim thickness greater than 5 mm, and an average diameter of 28.23 cm.

The Hope Hull phase Montgomery Red Filmed bowls have a restricted orifice, or inverted form. Restricted forms are typically used for keeping things inside a vessel for an extended period of time or for containing liquids. Like the Dead River phase bowls, the majority of the Hope Hull bowls have both an interior and exterior slip, indicating that permeability was a consideration in the function of the vessels. Slips are more often functional than decorative, and greatly reduce evaporation of contents, or prevent moisture from getting into the contents. Montgomery Red Filmed bowls all have a wall thickness greater than 5 mm and an average orifice diameter of 24.97 cm.

Dead River jars have a flared rim, and based on vessel thickness and orifice diameter are relatively large. Vessel diameters fall between 22 and 46 cm and all of the sherds measured for rim thickness yielded values greater than 6 mm. Vessels with curved necks are adapted for storing and transport of liquids; the restriction of the vessel prevents spilling. On the other hand, flares are interpreted as spouts or funnels to make pouring or filling easier. The Hope Hull phase jars from Madison Park are curved without a flare. The orifice diameters of this set of jars ranges from 26 to 38 cm, and the rim thickness for all vessels is greater than 6 mm. None of the Hope Hull jars are slipped on the exterior or interior.

 Vessel Decoration
A small percentage of sherds in the rim subset assemblage exhibit decoration, incising, stab-and-drag, and punctation are the most common decorative modes. Incising is produced by dragging a pointed implement across the surface of a partially dried vessel before firing, and punctations are made by pushing an object into the surface. Stab-and-drag designs are produced by dragging a pointed object across the surface of an unfired pot and at somewhat regular intervals punching the object into the surface, creating punctations within the incision.

A majority (95.0%, n=3,812) of the rim sherds from Madison Park do not exhibit any decoration (Table 5-2). Of the decorated rims (n = 200), most are incised (n=104). Thirty-two rims are punctated and incised, and 18 are zone punctated. Twelve rims have stab-and-drag decorations, and three have a combination of punctations, stab-and-drag, and incising. One rim has incised lines and an appliqué node, and another rim is incised, scraped, and brushed.

 Table 5-2. Counts and percents of decorated sherds (from rim subset).

Of all the decorated rims, bowl rims (n=184) make up most of the sample. Only 4 percent of the decorated rims are jar forms. These data correspond with our current understanding of Dead River/Hope Hull ceramics, with bowls generally exhibiting a higher frequency of decorations and slip than do jars (Chase 1998a). When bowls are separated by form, the results are similar to the general bowl pattern, except for a spike in the incised incurved bowl category. Chi-square indicates that these distributions are significant. We expect to see a higher frequency of decoration on bowls than jars, but bowl rim shape (straight, incurved, or excurved) does not determine the type of decoration exhibited. David Chase’s observation—that decoration occurs only on incurved late Hope Hull (Montgomery Red Filmed, var. Froggy Bottom) bowls and not at all on earlier Dead River and Hope Hull bowls—evidently needs correction. The presence of decoration on straight-sided bowls indicates either some Dead River Red Filmed bowls are in fact decorated, or some straight-sided bowls date to the Hope Hull phase.

Decoration on body and base sherds was recorded for the sherd subset analysis (Table 5-3). As expected, none of the bases (n=51) exhibit any decoration, and a majority of the body sherds (88.9%, n=1,434) do not have decoration either. The most frequent type of decoration on body sherds is incised lines (n=54) or combinations of incised lines with punctation (n=28) and stab-and-drag (n=1), all or nearly all from bowls. Two body sherds are decorated with stab-and-drag only, 28 are punctated, and one red filmed interior-exterior sherd has a drilled hole.

 Table 5-3. Counts and Percents of Decorated Sherds (from body subset).

Many of the decorated body sherds classified as Montgomery Red Filmed have remnants of red slip on at least one surface. Of the 356 body sherds with red slip on the interior only, two are incised. Of the 94 body sherds with red slip on the exterior only, one is incised and another is punctated. In contrast, body sherds with red slip on interior and exterior exhibit more surface decorations. Twenty-seven red slipped body sherds are incised, seven are punctated, and one is double roulette punctated. There are combinations of these, such as punctated and incised (n=5), and stab-and-drag and punctated (n=1). The distribution of slip and decoration from the rim analysis is similar to their distribution on bodies and bases. Incised lines appear more frequently across all slip application categories. However, most of the decorated rims (n=263) do not exhibit slip. Decoration on the ceramics at Madison Park occurs on all types of sherds regardless of the application of slip.

Erosion on the exterior of a vessel suggests movement of a vessel while sitting upright, possibly during mixing of contents (Rice 1987:234). Two sherds (one body and one base) exhibit erosion on the exterior surface. Both were found in the same unit and level. Neither have an apparent slip on either surface, but the body sherd has stab-and-drag decoration on the exterior. The plain base with fire clouding was found in the same unit and level as the eroded sherds.

Finally, one incurved bowl rim was recovered with encrusted charcoal on the exterior. This vessel does not exhibit decoration or slip. The lip is square with overhanging paste on the exterior and the vessel has an orifice diameter of approximately 30 cm.

Discussion
Based on Chase’s original definition of Dead River phase ceramics types, vessel forms should include straight sided bowls with red filming, and (sometimes tooled) flared jars. Hope Hull Phase vessel forms should be restricted to inverted, red-filmed bowls and plain curved jars. We were interested in determining if the ceramic types originally identified for the two phases were distributed stratigraphically in the proposed chronological sequence at Madison Park, since it is rare to recover both from one site. Since the phases are essentially defined by vessel forms (flared jars and straight bowls are Dead River and curved jars and inverted bowls are Hope Hull), a vertical and horizontal examination of the ceramic distributions should address this question.

 The vertical distribution of types was examined across the whole site. Dead River Phase ceramics dominate every level (Figure 5-14); particularly Kilby Plain jars (Figure 5-15).

Figure 5-14. Distribution of ceramic types recovered from units by cultural phases and levels.

The vertical distribution of ceramics in the rim subset at Madison Park indicates that the largest proportion (n=1,046, 30.8 percent) of the sherd assemblage came from Level 2. Eighty-five percent of the sherds (n=2891) were recovered from the top three levels (the upper 15 cm of the site beneath the plowzone). All variables were analyzed per level, but distributions do not differ significantly from the overall assemblage. For example, the frequency of bowl forms does not fluctuate greatly between levels, suggesting little change through time; therefore the assemblage can be treated as representing a single phase (Figure 5-15). This pattern appeared again and again for every variable (overhanging paste, decoration, slip, rim shape, tooling etc) recorded in this study.

Figure 5-15. Distribution of ceramic types from units by level and types.

Dead River phase ceramics (Kilby Plain and Dead River Red Filmed) appear throughout the levels, as well as ceramics traditionally interpreted as Hope Hull (Montgomery Red Filmed and Adams Plain). Theoretically, Dead River phase should predominate in the lower levels. But Kilby Plain appears throughout the levels in a similar distribution as all the other ceramic types; Levels 1-3 contained the majority of the sherds and Levels 4-6 yielded smaller numbers.

Since Hope Hull is considered a later manifestation of Dead River, these similarities are surprising. Based on a traditional interpretation of the ceramic types present, occupation at Madison Park started during the Dead River phase and continued into the Hope Hull phase. However, the stratigraphic distribution of ceramics at Madison Park suggests continuity of all vessel types throughout the occupation. Further research on refining or combining these phases is warranted given the outcome of this study.

The Hope Hull occupation at Madison Park temporally overlaps with the Autauga occupation of central Alabama. The stab-and-drag decoration seen on a tiny number of Montgomery Red Filmed sherds at Madison Park share that decorative mode with Autauga phase Bear Creek Incised vessels. This is the first documentation of stab-and-drag decoration in a Hope Hull or Dead River ceramic assemblage. We propose that these stab-and-drag decorated vessels and similarly ornamented vessels found at Autauga sites have a common origin in contact with Weeden Island populations near the Gulf coast. The stratigraphic distribution of stab-and-drag sherds at Madison Park, albeit a small sample, indicates this mode of decoration existed throughout the Madison Park occupation. This suggests that stab-and-drag should be added to the type descriptions of Dead River and Hope Hull ceramic assemblages.

In conclusion, it appears that the original phase distinction between Dead River and Hope Hull based on the types presented by Chase needs revisions. Dead River phase ceramics dominate the entire ceramic assemblage. In conjunction with the radiocarbon dates obtained from this excavation, the site appears to have been occupied around AD 975, and again around AD 1009. Based on the previous date ranges proposed by Chase, this would place the site within the Hope Hull phase, not Dead River which dominates the ceramic assemblage. Unfortunately, none of the attributes examined here appear to have value in delineating phases of the Late Woodland occupation of 1MT318.