In 1993, the Colonial Williamsburg Foundation’s (CWF) Department of Archaeological Research (DAR) conducted an archaeological assessment of a tract of land slated for a housing development known as Holly Hills (Muraca 1993; McFaden and Muraca 1994). The Holly Hills tract was once part of Middle Plantation, the 17th-century settlement which was the precursor to Williamsburg. The shovel-test survey was part of a larger DAR project focused on the 17th-century component of Rich Neck plantation (site 44WB52; McFaden et al. 1999). Rich Neck was once the homestead of Virginia’s most prominent colonists, including Secretary of the State Thomas Ludwell. The 1993 survey revealed an 18th-century site consisting of artifact scatters, a series of below-ground features, and the remains of a brick chimney foundation approximately 100 feet north of the Kemp/Lunsford/Ludwell plantation home site. This site pre-dated the probable, 19th-century slave-related site excavated by the DAR in 1990 less than 300 feet away (Samford 1991).
Excavations of the Rich Neck Slave Quarter (RNSQ) by the DAR took place between June 1994 and April 1995. The site remains suggested a two-room, impermanent dwelling with ground-laid sills and a central H-shaped hearth and chimney. Nearly all of the major features associated with the site consisted of sub-floor pits, or “root cellars”, within the confines of the dwelling. The archaeological and historical evidence indicated that the residence once housed members of the enslaved community who worked as field hands on the Ludwell plantation from c. 1740s to 1778. A 1994 shovel-test survey led to the discovery of an additional enslaved domestic site (68AP), which was occupied between c. 1710 to 1740s (Agbe-Davies 1999).
The DAR projects are directed by Principal Investigator Marley R. Brown III. On-site excavations at RNSQ from 1994-95 were supervised by Maria Franklin, with the assistance of staff archaeologists Ywone Edwards-Ingram and Anna Agbe-Davies.
68AL Documentary Evidence
Given the prominence of the Ludwell family and its connection to the Lee family of Virginia, a wealth of primary and secondary sources exist relating to the Ludwell-Lees. These include documents associated with the ownership and management of Rich Neck plantation and the enslaved families owned by the Ludwells.
There are at least two probate inventories associated with Rich Neck plantation. Phillip Ludwell III’s 1767 probate inventory (transcribed in VMHB 1913) and his will (transcribed in VHM 1911; McGhan 1982: 652-653) date to the period of occupation of the Rich Neck dwelling (c. 1740-1778). Ludwell III’s inventory accounts for nine working plantations, including Rich Neck and Greenspring (the home plantation), and 235 enslaved blacks (21 of whom resided at Rich Neck). Rich Neck was then inherited by Ludwell III’s daughter, Lucy Ludwell Paradise. Before she died in 1814, an inventory of her estate was taken in January of 1812 (VHS 1812). In addition to listing the contents of Paradise’s three properties (which included Rich Neck and Chippokes plantations, and a house in Williamsburg), the names of enslaved individuals as well as their ages are given.
A collection of letters penned by William Lee includes correspondence regarding the management of the Ludwell family plantations and the enslaved work force from 1766-1783 (Ford 1968). There are also two important collections of wills, letters, estate and legal documents, etc., pertaining to the Lee and Ludwell families at the Colonial Williamsburg Foundation Library (Ludwell Papers, 1678-1828), and the Virginia Historical Society (Lee Family Papers, 1638-1867).
There are two historic drawings of Rich Neck plantation dating to the third quarter of the 18th-century. The first is a land plat drawn in 1770 by surveyor Wm. Goodall (Goodall 1770). It indicates that the tract consisted of 3865 acres. The survey was likely commissioned by the trustees of Ludwell III’s property as a means to facilitate the execution of his will. The second document is a military map of Williamsburg (Desandrouins 1781) and the surrounding environ completed in 1781 by a French officer by the name of Desandrouins. Rich Neck is labeled simply “M Ludwell”. A path leading to Jamestown Road is lined with four to five structures. These were probably a combination of slave dwellings and storage buildings. “Ludwell’s mill”, located at the intersection of Jamestown Road and College Creek, was used either for timber or grain.
A number of individual and family biographies exist relating to the proprietors of Rich Neck (including the Ludwells) from the 17th- to the 19th-centuries (Jester 1987; Nugent 1979; WMQ 1910:208-214). In Archibald Bolling Shepperson’s (1942) book, “John Paradise and Lucy Ludwell”, the author discusses the Ludwell family history and their plantation affairs (including mention of Rich Neck) at length.
68AL Excavation History
During the 1993 field season, the DAR conducted a shovel-test survey which lead to the discovery of the Rich Neck Slave Quarter (McFaden and Muraca 1994). A portion of a feature was found and excavators shoveled the overlying plowzone to expose it. As additional features appeared, shovel-shaving continued following the features rather than the established gridlines. Once the footprint of the dwelling was revealed, two below-ground features were cross-sectioned (contexts 501 and 683). The site was backfilled until June 1994 when full-scale excavations proceeded. The 1993 backfill was first removed in quadrants (contexts 1-4) using shovels. Plowzone was removed using 1x2m (n=1), 1x1m (n=12), and 2x2m (n=15) excavation units. Most of the 1994 excavation units were dug in order to remove plowzone remaining in units from the 1993 season. All of the excavation units were shovel-shaved to subsoil and 100% of the dirt was dry-screened using ¼” screens.
In all, 32 features were discovered (see Table 1). The remains of a central H-shaped hearth and chimney included the bottom, mortared brick courses of two chimney cheeks (master context 23), and hearth fill and burnt subsoil (master context 12) where the fireboxes were once located. No other architectural features, including postholes, were discovered. Fifteen subterranean pits were exposed. The pits were mostly located close to the hearth and were probably used for food storage. Most of these features, or root cellars, were cut into and/or cut other pits. This was clear indication that the root cellars were in use at different times. The remainder of the features consisted of a robber’s trench (master context 24) where bricks were removed from the northwest chimney cheek, four tree holes, one animal burrow, and nine unidentified features.
Table 1: Rich Neck Slave Quarter features.
Feature Description Master Context Associated Contexts
61, 64, 90, 92, 97, 103, 105, 110, 111 (cut for 110), 112,113, 119, 117, 121, 122,134-137,145 (cut for 135), 149 (feature cut)
25, 63, 73, 99,101 ,501, 1241, 1243, 1245, 1247, 1249, 1251, 27 (feature cut)
126, 128, 141, 161, 129 (feature cut)
29, 30 (feature cut)
31, 32 (feature cut)
34, 36, 37, 109,123, 127,132,133, 35 (feature cut)
44, 53, 120, 683, 684 (feature cut)
49, 71, 158, 185 (feature cut)
66, 91, 100,102, 67 (feature cut)
83, 84 (feature cut)
49, 144, 168 (feature cut)
47, 51, 146, 52 (feature cut)
21, 22 (feature cut)
93, 95, 94 (feature cut)
42, 68, 86, 130, 162, 163, 164, 165, 166, 169-173, 176, 181, 184 (feature cut)
57, 59, 179, 196, 58 (feature cut)
125, 125, 197
55, 56 (feature cut)
14, 15 (feature cut)
16, 17 (feature cut)
40, 41 (feature cut)
74, 75 (feature cut)
118, 143 (feature cut)
152, 153 (feature cut)
156, 157 (feature cut)
174, 175, 186 (feature cut)
107, 108 (feature cut)
154, 155 (feature cut)
177, 178 (feature cut)
The research design included an intensive, flotation sampling program: all of the major pit features and the robber’s trench were hand-trowelled and 100% of the excavated fills were collected for flotation. Thousands of faunal fragments (Franklin 2001) were recovered as well as charred seeds and numerous small finds.
A shovel-test survey was also undertaken in 1994. The 1993 survey generally placed shovel tests at ten-meter intervals. The 1994 survey collapsed the intervals by placing shovel tests at five meter intervals. Thirty-nine shovel tests (contexts 6000-6039) were excavated and dry-screened. Results of SURFER analysis indicated that refuse was broadcast northwest and west of the dwelling, and may have been swept periodically away from the frontal approach of the house (Agbe-Davies 1994). These results were further supported by the higher artifact concentrations within excavation units in the same area relative to those both south and west of the house. A concentration of artifacts just east of the dwelling was further investigated by establishing a 5x5m unit (context 76) over the area and piece-plotting artifacts from the ground surface to 10cm below grade. Three 1x1m units (contexts 138-140) were then excavated within the 5x5m unit, but no features were discovered. One of the shovel tests (context 6025), however, did hit a feature. A 1x1m unit (context 160) was dug adjacent to the shovel test which led to the discovery of site 68AP (c. 1710-1740s).
68AL Summary of Research
Based on the archaeological remains, historical evidence, and comparisons with research findings of other 18th-century, Chesapeake slave housing, the RNSQ dwelling was likely a wooden, double-pen residence with an internal wall and dirt floor. It measured approximately 20’ wide and 30’ long. Given the gender and age demographics of the enslaved Rich Neck community (Ludwell probate inventory), two families probably shared the structure, each with its own residence.
Ceramic and tobacco pipe stem data helped to establish an occupation span of c. 1740s-1778. Virginia half-pennies minted in 1773 were recovered from several pits backfilled during the latter stage of occupation. Documentary sources reveal a possible explanation. In 1778, William Lee learned that the Paradises, who owned a large portion of Rich Neck, which likely included the quarter, had been abandoned by all of the enslaved, save for one (Shepperson 1942:136). Governor Dunmore had decreed that any slaves escaping to the British who were willing to fight the rebels would be freed (Kulikoff 1986:418–419; Mullin 1972:130–136). As many as 70 enslaved individuals fled Greenspring that same year as well.
RNSQ site phasing consisted of determining the life cycle of the house and its related features from initial construction to post-abandonment (see Table 2). Six phases were determined based upon the depositional histories of each feature (as determined using ceramic crossmend data, TPQs for each context, and the sequence of backfill for each feature) and the relationships between features.
Table 2: Rich Neck Slave Quarter site phasing. The TPQs given for Phases I-IV are based upon the most recent TPQ established for a particular context assigned for each phase.
Phase I: Initial Construction of House (c. 1740) Master Context
Context Numbers Associated with Phase I
57, 59, 179, 196, 58 (builder’s trench cut)
Phase II: Root Cellar Activities (tpq 1745) Master Context
Context Numbers Associated with Phase II
134, 135 (145 cut), 137, 149 (feature cut)
29, 30 (feature cut)
35 (feature cut)
71, 158, 185 (feature cut)
144, 186 (feature cut)
14 and 17
49 (redeposited sub backfill)
21, 22 (feature cut)
169, 170, 171, 172, 173, 176, 181, 184 (feature cut)
Phase III: Root Cellar Activities (tpq 1765) Master Context
Context Numbers Associated with Phase III
90, 92, 97, 105, 117, 103, 110 (111 cut), 112, 119, 121, 122, 136
100, 102, 67 (feature cut)
84 (feature cut)
Phase IV: Root Cellar Activities (tpq 1773) Master Context
Context Numbers Associated with Phase IV
27, 73, 99, 101
129 (feature cut)
32 (feature cut)
34, 36, 37, 109, 123, 127, 132
120, 684 (feature cut)
47, 51, 146, 52 (feature cut)
94 (feature cut)
42, 68, 86, 130, 162, 163, 164, 165, 166
Phase V: Final Backfilling of Features and House Abandonment (c. 1773-1778) Master Context
Context Numbers Associated with Phase V
25, 63, 501, 1241, 1243, 1245, 1247, 1249, 1251
126, 128, 141, 161
44, 53, 683
Phase VI: Post-Abandonment Activities (post-1778) Master Context
Context Numbers Associated with Phase VI
55, 56 (feature cut)
Most of the artifacts recovered from the sub-floor pits constituted secondary depositions. The ceramic crossmend evidence indicated that individuals collected backfill dirt largely from the midden northwest and west of the house (see above discussion under “Excavation history”). Since residents in both halves of the house probably used the same area for refuse disposal and the retrieval of backfill for defunct root cellars, it was impossible to compare the two households using archaeological evidence. However, given the depositional history of the root cellars, it is possible to analyze the data diachronically as representative of both households over time.
The majority of the historic artifacts constituted the architectural materials associated with the dwelling (nails, window glass, etc.) and domestic debris commonly found on household-related sites (ceramics, wine bottle glass, table glass, etc.). The artifacts represented not only household activities such as foodways and sewing, but objects related to clothing and personal adornment, home furnishings, hygiene practices, and medicinal usage were also present. A variety of woodworking tools such as hammers and chisels were recovered. Hilling hoes and sickles were token reminders of the daily toil of Rich Neck’s enslaved community.
Faunal analysis revealed that this community actively participated in subsistence practices to supplement their food rations (Franklin 1997; 2001). A wide variety of fish (including catfish, herring, and perch), wild mammals (including raccoon, rabbit, and opossum), birds (Canadian goose, chicken, and turkey), and oysters were consumed. Personal garden plots were probably allowed based upon the botanical evidence. Beans, cowpeas, and squash were identified, as well as various grains including barley, wheat, and rye. Foraging for foodstuffs was also indicated by the presence of honey locust, black walnut, acorn, and blackberry seeds.
In 1994, in the process of conducting a test pit survey on the periphery of the Rich Neck Slave Quarter (68AL), Colonial Williamsburg archaeologists Ywone Edwards, Maria Franklin, and Anna Agbe-Davies identified a large artifact-filled feature. Subsequent excavation revealed a small cellar and another associated feature that bridged the years between the 17th-century manor complex and the later 18th-century quarter. These features were designated 68AP, the “small cellar” site.
Analysis focused on the two primary artifact-bearing features, approximately 3 meters apart. The smaller pit, designated “M3,” was sub-rectangular and quite shallow with a carefully dug flat bottom. This pit contained only one depositional layer with a terminus post quem of 1755. The cellar, designated “feature 1,” was dug straight into the clay subsoil with no evidence of a lining or architectural elements that would hint at the nature of the superstructure above. Based on tpqs, the initial filling of the cellar dates no earlier than the 1730s, with the uppermost layers dating to 1745 or later.
The cellar is the only remaining evidence of the structure that stood in the early 18th-century. After abandonment, the empty cellar was exposed for a brief period of time, during which intermittent and random filling occurred. The second quarter of the 18th-century saw at least two episodes of rapid intentional filling, perhaps close together in time, but from different primary deposits. The filling dates overlap with the earlier end of the date range of the nearby Rich Neck Slave Quarter (68AL). Indeed, the filling of “feature 1,” may relate to the initial occupation and use of that later dwelling.
68AP Documentary Evidence
Very little documentation exists for the occupation represented by “feature 1” and “M3”. An inventory taken in 1767 lists occupants, tools, and livestock for a quarter on Rich Neck plantation (Virginia Magazine of History and Biography (VMHB) 1913). However, given the date, the inventory probably refers to the Rich Neck Slave Quarter (68AL).
The Rich Neck plantation ceased to be the Ludwell family seat in the first decade of the 18th-century. Early 18th-century owner Philip Ludwell II lived several miles away at Greenspring, as did his son Philip Ludwell III, who inherited his father’s holdings in 1727/6. There is no evidence that anyone other than an enslaved work force resided at Rich Neck after ca. 1706. All of the available references to plantation managers and occupants date to the years Philip III spent in England before his death sometime in the 1760s (Smith 1993). Although these texts refer to a time after the abandonment of the structure represented by “feature 1”, Agbe-Davies (1999:32) has speculated that the filling of “feature 1” may be part of a reorganization of Philip III’s holdings upon his departure for England prior to 1760. Please see the Rich Neck Plantation history for further information.
68AP Excavation History
The defining feature of this site—its small, unlined cellar—was first identified in 1994 during a phase II survey of the area surrounding the Rich Neck Slave Quarter. The close (5 meter) interval of the survey ensured that one of the 75 x 75 cm test units would overlay the small cellar.
In early 1995, the area around “feature 1,” was machine-stripped, exposing the approximately 4 x 4 meter extent of the cellar, and another feature (M3), that was, in size (1.2 x 1.6 meters) and shape (sub-rectangular), not unlike the sub-floor pits identified at the Rich Neck Slave Quarter, roughly 15 meters to the northwest. At this time, archaeologist Dwayne Pickett and his crew sampled a 1 x 1 meter column of the cellar to a depth of 70 cm, using arbitrary levels and screening all deposits.
In the summer of 1995, a larger area (approximately 15 x 13 meters) was machine-stripped in an attempt to identify foundations or postholes associated with the cellar prior to full-scale excavation supervised by Anna Agbe-Davies. Further clearing of the plowzone revealed a sloping entrance to the cellar at its western edge. The plowzone was not screened, aside from the 2.25% sample recovered by the initial test pit survey.
All of the features uncovered were excavated using trowels and screened through ¼ inch mesh, with the exception of subsoil erosion layers in the cellar, which were not screened. Excavators collected flotation and chemical samples, including a 25 x 25 cm column sample through the fill layers of the cellar. This column sample was processed, but not analyzed. Other samples include oyster shell and brick.
None of the irregularly-shaped features appeared to be cultural, so excavation and analysis focused on the cellar, “feature 1”, and “M3.” The cellar was excavated in quadrants, leaving the southwest quadrant (largely obscured by the root system of a large tree) in place. The fill of the cellar extended 1.3 meters below subsoil at its deepest point, and was composed of four major depositional episodes interspersed with erosion layers of displaced subsoil, indicating exposure to the elements. A Harris Matrix illustrates the depositional sequence and correlates the deposits in the three quadrants.
68AP Summary of Research
In an analysis of the survey that revealed the small early- to mid-18th-century cellar at Rich Neck (feature 1) Agbe-Davies (1994) noted that plowzone artifact concentrations were not a reliable predictor of sub-surface features. A crossmending study confirmed that the major fill layers in “feature 1” were discrete depositional events (Agbe-Davies 1997). Crossmends also revealed only one direct link between sealed contexts from 68AP (master context M6) and a feature from the Rich Neck Slave Quarter, root cellar 5. Non-contiguous vessel fragments linked M4 with root cellars 10, 15, and 21, and reinforced the association between M6 and root cellar 5. All of these root cellar features are among the earliest subfloor pits from the Slave Quarter site (Franklin 1997).
The resulting ceramic vessels and their functions are discussed in detail in the site report by Agbe-Davies (1999). However, a comparison of vessel forms with contemporary sites (quarter and non-quarter) revealed no clear patterns.
The site report also includes a discussion of the faunal material recovered from the uppermost fill deposit from “feature 1”. Biomass measures reveal that over 3/4 of the potential meat at the site would have come from cattle, pigs, sheep/goat, and unidentifiable large and medium mammals, with cattle forming the largest share. Similar rankings were suggested by the meat weight calculations.
A detailed assessment of the site formation processes that contributed to the development of “feature 1” concluded that it did not represent the demolition of the overlying structure, but rather intermittent discard of household debris, probably from the Rich Neck Slave Quarter, interspersed with brief erosion episodes (Agbe-Davies 1999:32).