Submitted by Jannie Loubser and Tommy Hudson (jloubser@newsouthassoc.com)

Introduction

The current owner of a 700-acre ranch on Pigeon Mountain, Walker County, bought the property in 1948 from a Euro- American family that owned it since the late 1830s. This family had a tradition that there were already stone piles on the land when their ancestors first took possession of it. Taken at face value, this tradition implies that Cherokee Indians that lived in the area prior to removal could have piled the stones. This short article presents ethno-historic eyewitness accounts and tentative archaeological evidence that historic period Cherokee Indians could indeed have piled these stones.

Historic eyewitness accounts

In the mid-eighteenth century, the trader Adair (1930:193– 194) observed, “every Indian traveler as he passes that way throws a stone on the place…In the woods we often see innumerable heaps of small stones in those places, where according to tradition some of their distinguished people were either killed, or buried, till the bones could be gathered.” Adair’s statement that “they continue to raise and multiply heaps of stones” suggests that this practice was still in vogue as he traveled through the southeastern United States between 1735 and 1775. At a place along the trading path in North Carolina where General Middleton killed a great number of Cherokee warriors, the eighteenth century naturalist Bartram (1955:283) “observed on each side of the road many vast heaps of these stones.” From these two independent sources it would appear that stone piles occurred close to Indian trails.

In the light of these accounts, ascribing the stone piles on the Walker County property to the former Indian inhabitants of the area might seem reasonable were it not for the fact that Euro-American farmers are also known to have piled stones (e.g., Gresham 1990). Euro-Americans piled stones for a variety of reasons, chief among these being clearing land for agricultural fields or for stockpiling building material. Unless stone piles contain in situ Native American Indian artifacts and/or features, archaeological identification normally depends on dates; AD 1800 being a general dividing line. Unfortunately, severe fluctuations in atmospheric radioactive carbon—roughly between 1700 and the present—make this a difficult period to date by means of the radiocarbon dating method (i.e., when calibrated against the annual tree ring curve, the “raw” radiocarbon count typically intersects the curve at two or more points). In the absence of artifacts and/or features covered by the stone piles then, dating is at best only an inconclusive alternative, leaving us with equivocal results. Nonetheless, we thought it prudent to present the dating results here as part of a growing database on stone piles in Georgia.

The site, the excavations, and the results

Open hardwoods currently cover the steep slopes on which the 28 stone features (one is a short wall) occur. As can be seen on Tommy Hudson’s sketch map, most piles occur on a slope with a northeastern aspect. All the piles are down slope from two springheads. The biggest pile occurs on the edge of the landform where it drops down to the confluence of to creeks. An empty area (indicated as “void” on the sketch map) occurs on the steep slope within a ring of piles. Most of the piles have been disturbed in one way or another.

We decided to excavate a 2 by 1 meter trench through one of the more intact piles, known as Stone Pile 8 (sketch below). Even this pile, unfortunately, has been disturbed. A depression occurred near the center of the pile and many stones from the pile were scattered on its down slope side. Excavation proceeded by first carefully removing individual stones from the pile and placing them in a “mirror-like” fashion on a ground sheet next to the trench (after excavation these stones were carefully replaced to their original positions). Removal of soil occurred in 5-centimeter thick levels. All dirt was screened through a quarter- inch mesh to facilitate artifact and charcoal recovery.

Charcoal fragments from roughly 7 centimeters below the actual ground surface were collected for identification and dating (Table 1). Leslie Raymer, ethno-botanist at New South, identified both partially and fully carbonized pine and fully carbonized hardwood in the sample. Only the fully carbonized hardwood charcoal pieces were submitted to Beta-analytic for radiocarbon dating (the small sample was dated directly via Accelerator Mass Spectrometry, or AMS). The charcoal came from a reasonable context, considering that the sample moreor- less corresponded with the deepest stones of the pile that were embedded within the dirt (in fact the charcoal came from the transition between light brown clayey loam to light gray loamy clay). Two soils samples were also collected for Oxidizable Carbon Ratio (OCR) dating (one from 5 cm deep within the clayey loam and the other from 10 cm within the loamy clay).

In terms of depth below ground surface, the two OCR samples sandwiched the AMS sample. Given that this stratigraphic succession is not the result of disturbance, it can be expected that the lowest OCR date would be earlier than the AMS date and that the upper OCR date would be the most recent. As can seen from Table 1 this expectation has been met. From the south profile it looks as if the OCR samples come from well below a stone resting on the ground surface. It is worth mentioning that this stone fell off the main pile some time in the past and is not occupying its original location. When viewed in relation to the core of the pile, as shown in the west profile, the upper OCR sample is slightly higher than the lowest stones. In terms of stratigraphic association, the charcoal came from a context related to roughly when the pile was stacked (i.e., directly in line with the very lowest stones). For this reason the AMS date is probably the best estimate of when the stones were first piled. Although the intercept of the AMS date with the calibration curve is AD 1670, other possible dates within the 1 Sigma range are equally likely. Overall, however, given the evidence presented here, we propose that a pre-Euro-American date is most likely (i.e., between AD 1660 and 1800). This would place the stone piles roughly within the time period that Adair and Bartram saw similar piles during their travels through Georgia and neighboring states.

Implications

We submit that due to the vagaries of radiocarbon dating for the late-eighteenth and nineteenth centuries, the date ascribed to the stone pile is not conclusive. Moreover, the stone pile showed signs of disturbance and it is indeed possible that charred tree roots were dated that might not have much to do with the time when the stones were stacked. Unfortunately, in the absence of features and artifacts from the stone pile, assessment of its authorship remains equivocal. There is nevertheless the tantalizing possibility that these piles indeed date to a time when they were ethnographically linked to the Southeastern Indians. Similar to the eyewitness accounts that suggest stone piles occurred next to Indian trails, the stone pile concentration recently investigated in Walker County occur on an old trail that runs through a nearby gap through Pigeon Mountain. This trail links a big Woodland and Mississippian period site at the eastern base of the limestone ridge with a desolate plateau to the west. The trail passes a prehistoric pecked rock art site in cliffs overlooking the gap and the stone pile site reported here. In other parts of Georgia, prehistoric pecked rock art and stone pile sites also occur in or near mountain gaps through which old trails passed. Viewed in terms of Southeastern Indian beliefs, such locales are very likely transition points on the landscape between the world of the living to the east and the world of the dead to the west (e.g., Mooney 1982:18).

References cited

Adair, J.
1930 Adair’s History of the American Indians. Edited by S.C. Williams. Promontory Press, New York.

Bartram, W.
1955 Travels of William Bartram. Edited by M. Van Doren. Dover Publications, Inc. New York.

Gresham, T.
1990 Historic Patterns of Rock Piling and the Rock Pile Problem. Early Georgia 18(1-2):1-40.

Mooney, J.
1982 The Cherokee Ball Play. The Journal of Cherokee Studies 17:10-24.

Acknowledgements

We thank the current ranch managers for allowing us to dig on their land. We deliberately left out their names and the location of their property in order to help preserve the stone piles and rock art on the ranch. Joe Joseph, President of New South Associates, Inc., is thanked for allowing Jannie Loubser the time to conduct the excavations and write-up. Marilyn Moore and Tommy Hudson from Geo-Enviro Engineering, Inc. initiated the work, provided accommodation, and paid for the dates. Leslie Raymer is thanked for identifying the charcoal samples. Jack Wynn is thanked for his badgering and Brian Thomas for his editing.

Posted online on Wednesday, June 1st, 2005