Montpelier receives National Geographic Grant to uncover larger Slave Community on Property
The proposed project involves archaeological study of a remote set of slave quarters and work sites scattered across the woodlots of the 2700-acre Montpelier property, located in Central Virginia. The sites being examined were once part of the larger Madison plantation and appear to date to the late-18th and early-19th century. We are interested in determining whether these sites date to a key change in the agricultural operations of the estate. Starting in the 1790s James Madison modernized the crop production at the plantation to include plowing, a variety of crop types, and mechanized processing of grains. Prior to this time, the agricultural base consisted of tobacco production. These changes did not just impact the economy of the plantation--it resulted in large-scale changes to the daily lives of the enslaved workers who had lived at Montpelier for over 60 years.
While we have located several potential sites dating to this period through metal detector surveys, we need to conduct more intensive archaeological testing at these sites to determine their occupation period and use. This proposed project will provide the data necessary to interpret these sites within a larger context--not only within the changes at Montpelier, but also within the broader region. The results from these investigations will shed light on one of the founding father's attempts to modernize his farm during the early Federal era. This information has the potential to provide insight into Madison's views on slavery and the role he saw it playing within the nation.
The proposed project will allow for an archaeological examination of the James Madison's Montpelier plantation farm located in the Virginia Piedmont. Previous archaeological and documentary research have allowed for the identification of the historic core of the Madison plantation including extensive research on the main house, quarters for house slaves and skilled artisans, and the work buildings and slave quarters for the main farm complex (Reeves 2010; Reeves and Greer 2012; Marshall 2011; Trickett 2013a and b). We have spent the past four years engaged in a intensive study of the plantations historic core including Madison-era slave quarters and work. This National Endowment for the Humanities-funded study has given us an excellent context for understanding the range of slave housing Madison had in the plantation core and his efforts to modernize the farm in the late 18th century and early 19th century. What remains to be explored over the estate's 2700 acre landmass are the dozens of sites that formed the quarters for field slaves, barns, work areas, and other building sites that made up the larger plantation for outlying fields. Similar studies conducted at Monticello have revealed slave quarters and work areas scattered across a range of topographic settings (Bon-Harper, et al. 2003). Recent archaeological surveys at Monteplier have located over a dozen Madison-era sites related to the agricultural operations of the larger estate (Reeves and Woehlke 2013) (Figure 1). The closer study of these sites is the topic for this proposed study.
Similar to other plantations in the Virginia Piedmont, the Montpelier property was used intensively throughout the 18th and early 19th centuries for a variety of agricultural activities. Evidence of intensive tobacco production, in particular, can be seen in the depleted soils in this area caused from erosion of exposed fields. In the 1790s, James Madison arranged for the plantation to change from a reliance on tobacco production to a more diversified crop system involving animal husbandry, plowing, and rotating crops, changing the plantation to a farm operation (Trickett 2013; Hicks 2012; Morgan 1998). The infrastructure for both periods would have consisted of not only fields, fencelines, and roads, but also work and living complexes for the enslaved work force who provided the labor for production. The sites included in this proposed study date to the late 18th/early 19th century when Madison was changing the agricultural base of his parent's plantation. What makes the archaeological study of the this plantation complex unique is the unique preservation of these sites following their abandonment in the 1830s.
Sometime in the late 1820s and 30s crop production at Montpelier was severely restricted, due to the combination of a stagnant agricultural economy and depleted soils. At that time, large portions of the property containing steeper slopes were abandoned for crop production and allowed to return to forest, leaving the sites undisturbed into the 20th century (Dierauff 2011; Reeves and Trickett 2009). Following these sites' abandonment, a key factor in the preservation of the sites at Montpelier was the 20th-century duPont ownership. William and Annie duPont, who purchased the property in 1901, had the financial resources to exist independently from any revenue that could be extracted from the property and they refrained from agricultural and timber activities. As a result, much of the land at Montpelier remained undisturbed for the next 80 years--during which time other sites in Virginia and the South witnessed massive disturbance from mechanized plowing and commercial-grade timber harvesting. These uniquely preserved sites at Montpelier make up for the destruction of many of the late 18th and early 19th century plantation documents that occurred in the 1850s following the deaths of Dolley and James Madison (Miller 2007).
Clues to the large agricultural operations for the Madison plantation farm come from surveys conducted by the Montpelier Archaeology Department over the past two field seasons. In these surveys, we located over a dozen late 18th to early-19th century historic sites related to the agricultural operations of the Madison farm (Woehlke and Reeves 2013). The survey work that we conducted during the 2012-2013 seasons resulted in the identification of eight potential slave quarters and seven work sites on two different portions of the property. These sites were identified through gridded metal detector survey. The grids used for site location were set on 20-meter intervals with the entire area swept for metallic signals. Isolated signals were excavated and within clusters enough signals were excavated to confirm presence of historic hits. The majority of the hits recovered were wrought nails and early cut nails that allow for a rudimentary dating of the sites to the late 18th and early 19th century (Reeves 2013). In addition, at many of the sites, the density of nails and presence of non-ferrous items (identified through signals and sampling) provide evidence for domestic sites (as indicated by the presence of buttons, cooking implements, etc.) as opposed to a barn or work area. During these surveys, the majority of the metallic hits were left for future surveys and excavations. In this manner our initial surveys were essentially remote sensing surveys with minimal ground truthing to provide historic context for the deposits (Reeves 2013).
The artifacts recovered during the course of the 2012-2013 surveys were from exclusively historic contexts; no modern debris (such as shot gun shells, wire fence, or 20th century trash) was present. Given the results of these surveys, we propose to start the next phase of survey: a more intensive gridded metal detector survey conducted on a 10-foot intervals grid. Given the fact that all of the hits sampled at these sites were historic, during the close-interval metal detector surveys, we will mark all hits within each 10-foot grid square to create a density plot of hits for each site. Prior to conducting the surveys, each site will be cleared of leaves and debris to facilitate identification of surface features and allow the metal detector coil closer contact with ground surface. The density plots will allow the exact limits of the site to be defined in addition to identifying artifact concentrations that indicate the position of structures and trash deposits. The metal detector surveys also have the advantage of allowing definition of the site with no ground disturbance, thus preserving the integrity of the data for future archaeological excavation. These metal detector surveys will be conducted through our public programs that train metal detector enthusiasts the techniques of archaeological survey and importance of site preservation (Reeves and Clark 2013). Our survey techniques are part of a larger set of experiments in the archaeological discipline to locate smaller African-American sites missed through standard survey techniques (Espenshade and Severts 2013).
Once the sites are defined through close-interval metal detector surveys (used to define the main core of the site), we will excavate shovel test pits (stps) every 10 feet within the core of the site as defined by the 10-foot grid metal detector survey. Shovel test pits allow for the site stratigraphy to be defined (we cross-reference the strata found in each stp to obtain an overall understanding of site stratigraphy), allow for the identification of features (area of charred wood, ash, or deeper deposits) that can be further explored through units, and allow for the recovery of the full range of artifacts otherwise missed through metal detector survey (ceramics, glass, and bone). Once a feature or deposit of interest is encountered, excavation will be stopped to avoid damage to deposits. These particular spots will be noted for placement of excavation units.
The final phase of testing will entail the excavation of at least 2-3 5x5ft excavation units at selected sites where further definition of deposits is needed. No features will be excavated during this phase of survey, just presence noted. The sample of artifacts coming from excavation units will provide the assemblages necessary for making preliminary comparisons of sites encompassed by this study and with existing studies of slave quarters already excavated at Montpelier.
With the proposed sampling of the sites we have located, we expect to recover enough information from these sites to interpret the time period of occupation, function of the site, and definition of any surface features. During the course of the surveys we expect to sample at least 15 sites (eight potential slave quarters and seven potential work areas). Following the 10-foot grid metal detector and stp surveys, we will select 8-10 sites, based on density of deposits, to place 5x5ft excavation units. With this data, we will have a context for establishing the range of sites that made up the outlying farm as well as understanding how topography played a role in site usage at different times, especially for the abandonment of crop lands in the 1830s.
An important component of this project is enlisting the aide of experienced metal detectorists during survey. We have worked out a partnership with the leading manufacturer of metal detectors (Minelab Americas) to engage metal detectorists through involvement with archaeological surveys (see Reeves 2013 SHA blog). This partnership is already established through two seasons of field programs and provides potential for not only providing significant methodological contributions, but also tremendous public outreach between two groups that have been traditionally diametrically opposed.
What we hope to establish with the proposed study is a more holistic understanding of the Madison plantation/farm. We hope to see how the sites found in the peripheral areas compare with the sites located within the core of the Madison plantation. Our recently-completed NEH study of the plantation core indicates Madison was using the latest in agricultural innovations and housing design for the areas closer to the main house. The patterns derived from the NEH study will begin to provide a baseline for understanding whether these changes extended to the periphery of the plantation. What makes this work significant is there have been every few studies of peripheral plantation areas within such well-preserved contexts. The results of this study will provide the necessary site context to begin planning for larger studies of these unique sites. The discoveries made have the potential to provide insight into daily life for a group of people poorly recorded in the historical record. Their lives, so far from the main house, represent the experience of most enslaved individuals at larger estates such as Montpelier. The novel techniques utilizing metal detector surveys for defining site areas combined with unique contexts gives the opportunity for this study to make both methodological and substantive contributions in a range of areas.
Archaeology staff member and metal detectorist plotting sites in woods at Chicken Mountain.
Map showing Montpelier’s 2700 acre property. Area in green is wooded with the remainder in fields. The dashed lines denote areas we have surveyed at 100% coverage in metal detector surveys. The area in peach is the location for the plantation historic core where we have spent the past four years conducting an intensive study of the Madison-era slave quarters and work areas.