lithic analysis

by Tara L. Potts and Philip J. Carr

Jump to: Lithic Analysis and Organization of Technology - Culture History
Lithics and Reconstructing Cultural Lifeways - Flakes, Tool Use, and Other Lithics - Conclusions
- References

Culture History

Our understanding of the culture history of central Alabama is hampered by a number of factors, especially with regard to diagnostic stone tools. First, no well-stratified, multicomponent sites have been excavated in the region with a suite of absolute dates to provide a firm basis for forming culture-historical types. Second, radiocarbon dates in association with hafted bifaces are rare from any site in the region. For these reasons the current culture history is based on surface finds and limited excavations or adopted from other regions, which leads to a third problem: we do not have sound understandings of the range of variability within a given type or the relationships of different types.

Obviously the excavation of a stratified site that produced diagnostic stone tools in context would be significant for this region. For this study we have classified diagnostic bifaces to culture-historical types using standard references for the Southeast (e.g., Cambron and Hulse 1964; Justice 1987; McGahey 2000) and others with specific discussions of central Alabama prehistory (Chase 1998; Walthall 1980).

Radiocarbon dates and ceramic analysis (see Chapters 4 and 5) indicate that Madison Park is a Late Woodland site with Dead River and Hope Hull phases present. Discussions of artifact types representative of these phases tend to focus on ceramics; little effort has been devoted to defining temporally diagnostic lithic tools. The Madison Park site assemblage offers an opportunity to better define stone tool types with culture-historical significance.

Dead River Phase
Walthall (1980:176) considered the Dead River phase to date from AD 500 to 700, and that sites belonging to this phase have “small, Hamilton-like triangular arrowpoints.” In a more recent overview of the Dead River Phase, Chase (1998:68-70) revised his date range for the phase forward to about AD 700 to 800. He also thought the “Bradley Spike type projectile points as well as a tear-drop shaped point similar to a Nodena point illustrated by Cambron and Hulse’s updated handbook [1975 A:21] … would be recognized as characteristic artifacts associated with Dead River ceramics.” Importantly, no one would consider a Bradley Spike or a Nodena-like point to be “Hamilton-like,” which suggests that either the Dead River phase has a diverse lithic assemblage or one of these two authorities is incorrect in describing the lithics of this phase.

The general description of the Bradley Spike provided by Cambron and Hulse (1975:19) is “a small to medium-sized, spike-shaped stemmed point.” This type is not discussed by Justice (1987) or McGahey (2000). The “Nodena-like” point type discussed by Chase as being characteristic of the Dead River phase is not particularly surprising since it is generally similar in form to the Bradley Spike. The Nodena point is described by Cambron and Hulse (1975:97) as “a small to medium-sized lanceolate point with rounded base.” However, nothing that fits the Bradley Spike or Nodena descriptions was recovered from Madison Park.

Hope Hull Phase

Walthall (1980:177) considered the Hope Hull phase to date from AD 700 to 900 and again noted that “Hamilton-like small, triangular arrow points have been frequently reported from … Hope Hull components.” In his retrospective discussion of Hope Hull, Chase (1998:71) stated, in regard to 1MT49, the type-site locale, that “Projectile points appeared from their description to be small triangles of the Hamilton type.” He further described Hope Hull, lithics based on his work at twenty-four sites, in the following way.

Associated lithics include a very small variety of the Hamilton Triangular point. These points rarely exceed 25 mm in length. Curiously, they do not occur with any degree of frequency even on the most productive sites. These are made either of quartz or of a black form of chert which is found in pebble form in the river bed. It is probably a transported alluvial material and may relate to the Knox formation in northwestern Georgia.

The general description of the Hamilton given by Cambron and Hulse (1975:64) is “a small triangular point with incurvate blade and incurvate base.” A “Hamilton Incurvate” type is described by Justice (1987:229-230) as part of his Late Woodland/Mississippian Triangular Cluster. He describes the type as:

“isosceles triangular forms with characteristic incurvate blades…. Basal edges of these points include straight and excurvate variations. Extreme delicacy and symmetry are characteristic of the type, resulting from refined flaking techniques. The blade edges frequently exhibit fine serrations and the tip and basal margins of these points are acute and needle-like in degree of refinement. The faces are shaped by broad, shallow flake scars resulting in a flattened cross section with fine retouch along the blade margins. The basal edge flaking characteristics differ. However, basal thinning scars are mostly restricted close to the lateral margins of the base.”

The distribution of this type is restricted by Justice (1987: Map 100) to north Alabama, citing the discussion by Cambron and Hulse. Justice (1987:256) provides metric data for Hamilton Incurvate points from Hiwassee Island, Tennessee (length 35-45 mm, thickness rarely exceeding 3 mm) and from southwest Virginia (length 24-35 mm, width 12-23 mm, and thickness 3-6 mm). The use of “small” in the description by Walthall and emphasized by Chase generally corresponds with these data.

While Justice (1987) does not extend the distribution of the Hamilton Incurvate to central Alabama, his Madison type, also part of the Late Woodland/Mississippian Triangular Cluster, is distributed across the Eastern Woodlands, including the entire state of Alabama. He suggests a date range of “AD 800 to the beginning of the historic period,” and describes the Madison point as “basically straight sided isosceles triangular arrowhead with excurvate-bladed variants…straight and slightly concave bases are both characteristic of the type” (Justice 1987:224-227). Cambron and Hulse describe the type as a “small, thin, triangular point that has been referred to as Mississippi Triangular,” but cites Jenkins and Nielsen (1974) as finding the type in “Late Woodland-Early Mississippian sites (West Jefferson Phase) within the Warrior Drainage” (Cambron and Hulse 1975:84). The Madison type is also discussed by McGahey (2000), who shows it having a wide distribution in Mississippi and a chronological placement from AD 500 to 1700. Interestingly, McGahey (2000:200) remarks that “an unusually large percentage of performs of this type were abandoned before they yielded completed points” which could be the “Nodena-like” points described by Chase (1998).

A final lithic artifact type, the greenstone celt, deserves discussion under the Hope Hull phase. While it cannot be said that either the use of greenstone or greenstone celts is indicative of a Hope Hull assemblage, greenstone celts are commonly discussed in association with Hope Hull. Chase (1998:72) specifically mentions “very well made greenstone celts which have been found in Hope Hull site association in Elmore as well as in Montgomery County.”

Other Woodland Diagnostics Relevant to the Madison Park site

Five other point types with suggested Woodland associations are relevant to the discussion of the Madison Park site: Camp Creek, Ebenezer, Hernando, Baker’s Creek, and Gary. The Camp Creek point type is another that would fit the description of “small Hamilton-like, triangular arrow points.” The general description of the Camp Creek type provided by Cambron and Hulse is “a small to medium, triangular point with incurvate base” (Cambron and Hulse 1975:22). As evident from the illustrated example, the base of a Camp Creek point has parallel sides before the blade tapers to a point; that is, there is a distinct hafting area. While Cambron and Hulse discuss the Camp Creek type as dating to the Early to Middle Woodland, Justice (1987:229) lists it as a morphological correlate of his Late Woodland/Mississippian Triangular Cluster.                       

The Ebenezer point type differs morphologically from the small, triangular points previously discussed; it is generally described as a small, short-stemmed point with rounded stem and excurvate blade edges (Cambron and Hulse 1975:42). However, Cambron and Hulse (1975:22) consider it possibly a part of the Greeneville Complex in association with Camp Creek point, and date it to the Early to Late Woodland.

Hernando points, which were used throughout southeastern North America during the Early and Middle Woodland periods, are triangular with basal notching that creates a “U” or “V” shape (Perino 1985:180).

The Bakers Creek point is certainly not Hamilton-like. Cambron and Hulse (1975:8) describe this type as “a medium-sized, expanded stem point” and date it to the Early to Middle Woodland, as they suggested for the Camp Creek type. Justice (1987) discusses the Bakers Creek type as part of the Lowe Cluster and considers it diagnostic of the middle and terminal Middle Woodland (Justice 1987:211).

Interestingly, McGahey suggests “Bakers Creek points are thought, primarily on the basis of surface associations, to be a Middle to Late Woodland type in Mississippi” (McGahey 2000:189). While surface associations are tenuous and there is geographic distance, the possibility that the temporal range of Bakers Creek points extends into the Late Woodland is intriguing.

The Gary point is the earliest of the temporal types discussed here, but this “medium-sized point with contracted stem also has the longest period of utilization as it is associated with Late Archaic to Middle Woodland time periods” (Cambron and Hulse 1975:57). Justice (1987:189) places the Gary Contracted Stem point in the Dickson Cluster and describes it as a Late Archaic to Middle Woodland type. In Mississippi, McGahey (2000:144) considers the Gary type to date to the Late Archaic.

This brief overview of diagnostic bifaces dating primarily to the Late Woodland demonstrates the chaotic nature of our current understanding. While it is difficult to argue with Walthall’s statement that Late Woodland sites in central Alabama are generally associated with “Hamilton-like small, triangular arrow points,” this general statement masks the potential variability that could be present in a given assemblage. The nature of the excavation at the Madison Park site allows some assessment of the current state of knowledge concerning lithic culture historical types from the region. On the other hand, it is also important to keep in mind that this is only one site.

Diagnostic Bifaces from the Madison Park site
A total of 121 hafted bifaces were recovered from Madison Park during surface collection and excavations of units and features (Table 6-1). Seven point types were identified: Madison (n=53; Figure 6-2), Hamilton Incurvate (n=37; Figure 6-3), Camp Creek (n=15; Figure 6-4), Baker’s Creek (n=3; Figure 6-5), Gary (n=3; Figure 6-6), Hernando (n=1; Figure 6-7), and Ebenezer (n=1; Figure 6-8). Eight of the hafted bifaces lack diagnostic attributes and could not be classified to any existing culture historical type.

Table 6-1. Point types by level and features.

Point Type

Level

Features/
Surface

Total

1

2

3

4

5

6

   n

%

n

%

n

%

n

%

n

%

n

%

n

%

n

%

Madison

15

28.3

11

20.8

8

15.1

5

9.4

3

5.7

2

3.8

9

17.0

53

100.0

Hamilton

21

56.8

5

13.5

6

16.2

1

2.7

-

-

1

2.7

3

8.1

37

100.0

Camp Creek

2

13.3

4

26.7

1

6.7

3

20.0

-

-

-

-

5

33.3

15

100.0

Hernando

-

-

-

-

-

-

-

-

-

-

-

-

1

100.0

1

100.0

Gary

-

-

-

-

1

33.3

-

-

-

-

-

-

2

66.7

3

100.0

Ebenezer

-

-

-

-

1

100.0

-

-

-

-

-

-

-

 

1

100.0

Baker’s Creek

2

66.7

-

-

-

-

-

-

-

-

-

-

1

33.3

3

100.0

Indeterminate

2

25.0

2

25.0

1

12.5

-

-

1

12.5

-

-

2

25.0

8

100.0


Figure 6-2. Typical Madison points from the Madison Park site (note minimal retouch on specimens c and d): (a) Knox chert, FS 179; (b) Knox chert, FS 601; (c) Knox chert, FS 183; (d) heat-treated Coastal Plain chert, FS 580.


Figure 6-3. Typical Hamilton points from the Madison Park site (note minimal retouch on specimen e): (a) Knox chert, FS 528; (b) Knox chert, FS 572; (c) Knox chert, FS 257; (d) Knox chert, FS 251; (e) Knox chert, FS 569.

Figure 6-4. Typical Camp Creek points from the Madison Park site: (a) quartz, FS 571; (b) Knox chert, FS 405; (c) Knox chert, FS 355; (d) Knox chert, general surface collection.

Figure 6-5. Typical Baker’s Creek points from the Madison Park site: (a) quartz, FS 503; (b) quartz, general surface collection.

1
Figure 6-6. Typical Gary points from the Madison Park site: (a) quartz, FS 333; (b) quartz, general surface collection; (c) quartz, general surface collection.


Figure 6-7. The Hernando point from the Madison Park site: quartz, FS 251.

 


Figure 6-8. The Ebenezer point from the Madison Park site: quartz, FS 601.

If one considers Madison, Hamilton Incurvate, and Camp Creek points to fit Walthall’s general description of small, Hamilton-like triangular arrowpoints, then the majority (n=105; 93%) of diagnostic bifaces are of this “type,” which Justice (1987) terms the Late Woodland/Mississippian Triangular Cluster. The majority (n=79, 87.7%) of the Madison and Hamilton Incurvate points found at Madison Park were knapped from a flake blank, and this is true for more than half of the Camp Creek points (n=8, 53.3%). The 18 remaining points of these three types are classified as indeterminate blank types, but could have been derived from flakes as well. Similar techniques evidently were used in the manufacture of these three types. Some Madison and Hamilton points (see Figures 6-2 and 6-3) exhibit minimal bifacial edges (n=4) and others are only unifacially retouched (n=6). However, none of the Camp Creek points exhibit this minimal manufacturing investiture so there is some variability as well.

Vertical distribution of the diagnostic projectile points indicates that Late Woodland hafted bifaces appear in all levels. At least one of the Late Woodland/Mississippian Triangular Cluster types was found in each level (see Table 6-1). Excluding features, the majority of Madison (n=34), Hamilton (n=32) and Camp Creek (n=7) projectile points were recovered from the upper three levels of the excavations (approximately 20-35 cm below ground surface).

Three points from the Madison Park site were recovered from contexts dated by radiocarbon assay. A radiocarbon date of AD 960-1040 is associated with a Hamilton point from the third level of the midden (approximately 10-15 cm bs). A Madison point was recovered in Feature 9, dated AD 1010-1180, and another was recovered from Feature 111, dated AD 890-1020. These three dates overlap, indicating that these projectile point types, although previously thought by some to be temporally separated, are in fact contemporaneous. These data further suggest that Hamilton and Madison projectile points may not be useful in distinguishing the cultural phases established for central Alabama, but are only indicative of the Late Woodland period in this region.

As is so often the case with scientific endeavor, new data raise as many questions as answers. At the Madison Park site, several different biface types apparently were used at the same time. Madison, Hamilton, and Camp Creek projectile points are suggested to belong to a Late Woodland cluster of projectile points. This raises the question, why were such distinctive projectile point types all used at the same time? Of course, these point types are only distinctive if one focuses on the differences, but it is hard for someone trained to look for such differences to ignore them. Viewed from a culture-historical paradigm, these different styles of arrow points would represent different cultures. Following this line of reasoning, three different peoples, each with a different mental template for the correct way to manufacture arrow points, came together at the Madison Park site. However, other cultural-behavioral answers to this question—such as a functional explanation for three different projectile point forms—deserve consideration as part of the investigation of cultural lifeways.

Questions concerning the culture history of the site are also raised by the presence of hafted bifaces that do not fit a Late Woodland Triangular cluster. Do these hafted bifaces represent an earlier occupation? This would be somewhat surprising given their wide distribution across the site and in features. Another possible explanation is that these tools were scavenged from other earlier sites. The counts are relatively small, but are there so few that this suggestion is tenable? Perhaps these points represent spears or knives in contemporary use with triangular projectile points. Additional explanations will be considered in our discussions of cultural lifeways and tool function.


Copyright © 2013 by The University of South Alabama
Last Updated:
Wednesday, August 14, 2013 3:02 PM