April 2, 2026: Intersite Variation of Coarse Earthenwares

By Dr. Christine S. Devine

April 2, 2026

Paper presented at the Middle Atlantic Archaeological Conference in Gettysburg, VA, March 11-15.

For the last several years, a major research focus of the Monticello archaeology department has been the reanalysis of coarse earthenware vessels from different sites at Monticello (Figures 1-6). A major goal of this research is to understand the role of coarse earthenwares in different household assemblages and identify where these goods were being produced. The first step in this project involved reexamining these artifacts using new scientific methods of identification and standardized terminology to allow for more accurate cross-site comparison. (Devine 2024, 2025).

This paper marks the beginning of the next phase  in our study of coarse earthenwares at Monticello.   It compares assemblages from four Monticello sites and identifies changing patterns of coarse earthenware use and sources over time. It also explores other possible factors that might account for variation, such as access to different markets, the introduction of new ware types, and consumer choice. These preliminary findings serve as a starting point to begin reconstructing local trade patterns and better understand the acquisition of coarse earthenwares at Monticello and other plantation sites.

Research Questions:

1. Do the frequencies of coarse earthenwares compared to other ware types change over time?
2. Does the proportion of imported and American-made coarse earthenwares change over time?
3. What types of regional wares appear in the assemblages, and do any patterns emerge in the frequency of certain regional types?
4. What factors might account for changes in coarse earthenware use and sources?

Site Backgrounds

The coarse earthenware assemblages in this study are from four sites discovered at Monticello, the home farm at the core of Thomas Jefferson’s 5,000 acre plantation in Albemarle County, Virginia. Two of the sites in this study, Building o and Jefferson’s house, are located at the top of the mountain and represent two very different households (Figure 7, right).

Building o Excavation

Dr. William Kelso led the excavation of the Building o site in 1979 and the early 1980s (Figure 8). The excavation was part of larger project to locate a series of workshops and living quarters for enslaved and white workers along a major thoroughfare on the mountain referred to as Mulberry Row. Based on Thomas Jefferson’s 1796 Mutual Assurance map and the artifact assemblage, Kelso concluded that Building o was a domestic dwelling occupied by enslaved domestic laborers between 1770 and 1800 and represented one building event  (Hill 2002, Kelso 1982). Later studies of the site suggest that there were likely two structures possibly representing two distinct building events (Schumate 1992; Wheeler and Sattes 2024).

Mansion Perimeter Excavation

In the early 1990s, Dr. Barbara Heath and the Monticello archaeology department conducted a mitigation project around the perimeter of Thomas Jefferson’s house (Figure 9). The project was part of a large-scale renovation of the mansion roof and drainage system that required excavation along the house foundations. Although much of the area had been disturbed as a result of modern utility lines and landscaping, Heath discovered numerous architectural features associated with the house’s initial construction in the 1770s and Jefferson’s post-1796 massive remodeling of the mansion. She also discovered Jefferson-period deposits containing artifacts that likely came from the Jefferson household (Heath 1990-1991). In 2019-2020, the Archaeology Department digitized all of Heath’s records and cataloged the collection into the DAACS database.

Home Farm Quarter Sites

The two other sites in this study, Site 17 and Site 30, were identified during the Plantation Archaeology Survey (Figure 10). The survey project is part of a research initiative that began in 1997 to locate potential sites on the 2,500 acres of Jefferson’s land owned by the Thomas Jefferson Foundation. Both Site 30 and Site 17 are located on the south slope of the mountain where Jefferson placed his agricultural fields. These sites and their artifact assemblages represent households from different periods in the economic development of the plantation.

Site 30

Site 30 was a Revolutionary-war period site occupied by enslaved field laborers. This period of occupation is referred to as the “tobacco period” at Monticello since tobacco was the primary commercial crop being produced on the plantation. Crystal O’Connor led the excavation of Site 30 from 2022-2025, which resulted in the discovery of several features, including a sub-floor pit and one large, unexcavated area to the south of the pit (Figure 11). The sub-floor pit and high artifact concentrations in three areas suggested that there were two and possibly three dwellings on the site. The proximity of these dwellings to the agricultural fields, short-term occupation of the site, and relatively small artifact assemblage indicated Site 30 was occupied by enslaved laborers during the last quarter of the 18th century (Devine 2025).

Site 17

Written sources identified Site 17 as a domestic dwelling occupied by Edmund Bacon, a white overseer and farm manager at Monticello from 1806 to 1822 (Figure 12c). Bacon oversaw both agricultural and commercial operations at Monticello after Jefferson decided to diversify his plantation and shift from tobacco to wheat production and small-scale manufacturing in the 1790s. Monticello archaeologists and Washington and Lee University students excavated the site in 2009 and 2010 (Figure 12). The artifact assemblage indicated that Bacon and his family lived in a modest house made of wood with a stone foundation or chimney and owned few personal items and mostly outdated ceramics. In their study of the Bacon assemblage and documentary sources, Allison Bell, Donald Gaylord, and Kristen Sharman concluded that the absence of costly goods at the Bacon site was not necessarily a result of limited resources but rather an intentional decision by Bacon to save his earnings, which he used to buy horses and invest in land and slaves after he left Monticello and moved to Kentucky (Bell et al. 2019). Bell’s, Gaylord’s, and Sharman’s (2019) findings are a good reminder that it is important to refrain from interpreting assemblages before examining multiple sources of information and exploring different theories that might explain an assemblage.

Methods of Analysis

The first step in doing a comparative analysis of coarse earthenwares from these sites was to make sure the assemblages were catalogued using the most up-to-date research and coarse earthenware cataloging protocols developed by the DAACS (Digital Archaeological Archive of Comparative Slavery) project (Bloch 2015, 2016a, 2016b, 2016c; Bloch and Bollwerk 2024). These protocols include measuring and recording physical characteristics or attributes such as paste color, types of inclusions, inclusion density, decoration and surface treatment, vessel shape, and form for each sherd (see Figure 13). This attribute information along with elemental studies of the clays used in imported and American-made coarse earthenwares can be used to identify the region where the vessel was produced if there is enough evidence.

After updating the data, the next step involved filtering out undiagnostic or ambiguous data that might skew the analysis: 

1. Contexts that lacked spatial or temporal information
2. Unglazed coarse earthenware flowerpot sherds (difficult to date)
3. Tin-enameled wares (Very broad manufacturing date ranges)
4. Unprovenanced contexts
5. Ware types without known manufacturing dates
6. Contexts with less than 5 ceramic sherds

Standardizing the information on coarse earthenwares and eliminating nondiagnostic data allows for a more accurate comparison of assemblages and determine if variation is related to time. At Monticello, archaeologists use a combination of statistical methods including frequency seriation, adjusted MCDs and TPQs, and correspondence analysis to do site chronologies and identify occupational phases. This statistical approach was developed by the DAACS project and Drs. Jillian Galle and Fraser Neiman. Correspondence analysis is a key statistical method which helps identify patterns of similarity or dissimilarity among ceramic assemblages by assigning contexts a position or score along a gradient or dimension based on their ware type frequencies. (Wheeler and Sattes 2024; O’Connor and Neiman 2024; Bates, et al 2020). This method allows us to see if factors such as time contribute to variation in ceramic assemblages. It also shows which ware types influence the placement of contexts within the broader chronological order.

Correspondence analysis of ware type data from the four sites shows that contexts from Site 30 are clustered together and spatially distant from the Mansion Perimeter contexts, which are also grouped together (Figure 14). The spatial location of these two groups correlates with important differences in the ware type assemblages (Figure 15). Earlier ware types are seen primarily in the Site 30 contexts while later types are associated with the Mansion Perimeter contexts. Most of Building o and Site 17 contexts are clustered in between the early and later contexts from Site 30 and Mansion Perimeter and somewhat overlap. This spatial patterning indicates that the variation in the assemblages is related to time.

This correspondence analysis of contexts and ware types can be used to identify chronological phases. We can plot the time-related dimension one scores for all the contexts on a histogram. In Figure 16, the bars represent clusters of contexts with similar ware type frequencies. The higher bars are representative of “peaks of different chronological phases.” (Wheeler and Sattes 2024). The beginning and end of a phase is marked by dotted lines. Each phase is assigned a beginning and mid-point date based on the BlueMCD (Best linear unbiased estimate mean ceramic date) and TPQ90 (90th percentile of latest occurring ceramic ware type) for that phase. We use a weighted MCD and TPQ in order to give greater weight to ware types with tighter date ranges and eliminate outliers in the data that may skew the chronology. It is important to remember that phases do not represent a single site but rather contexts from different sites that share similar ware type frequencies.

Findings

The assignment of phases provides the chronological framework to compare the site assemblages at the context level and see how coarse earthenware use may have changed over time. Figure 17 shows proportions of different ceramic materials across the four phases and indicates several broad trends. Refined earthenwares compose the greatest proportion of ceramics in all phases, with their relative frequency staying at the same level over time (between 61 and 67%). The frequency of porcelain steadily increases over time (going from < 10% to almost 30% of all ware types by the fourth phase). The only ware types that decline are coarse earthenwares and stonewares, with the greatest decrease occurring between the first and second periods. The dominance of refined earthenwares and steady increase in porcelains over time suggest that households regardless of their socioeconomic background chose to invest in more costly ware types rather than coarse earthenwares and stonewares.

What is more difficult to interpret is the significant decline in coarse earthenwares between the first and second phases. One explanation is that the contexts from Site 30 are heavily influencing the trend. Figure 18 shows that Site 30 contexts with earlier ware types and American Redwares are dominant in the first period. For this reason, it is difficult to ascertain whether the declining proportion of coarse earthenwares is site specific or if it reflects a broader trend. Expanding this study to include more coarse earthenware data from other sites at Monticello would help answer this question. 

Another possible explanation for the decline in coarse earthenware use may have to do with the increasing availability and acquisition of a new or more desirable ware type. Some scholars have attributed the decline of American coarse earthenwares to the introduction of the more highly fired and durable American stonewares (Myers 1977; Russ et al, 2013). One way to test this theory is by comparing proportions of imported and American-made coarse earthenwares and stonewares in each phase. Figure 19 shows that the proportion of American stonewares begins to increase at the same time there is a decline American coarse earthenwares and imported stonewares. The increase in American-made stonewares in these assemblages coincides with the growth of the American stoneware industry in the second half of the 18th-century. In Virginia, stoneware production grew significantly in the Shenandoah Valley region and lower James River Valley between the late 18th and early 19th centuries (Russ, et al 2013; Comstock 1994, pp. 3-18, 64-84). One way to further test this theory would be to include more data on coarse earthenware and stoneware assemblages from other sites. 

The final research question in this study sought to identify regional sources of coarse earthenwares and changing patterns of use. A breakdown in Figure 20 of the coarse earthenware data into regional types shows that imported wares came primarily from well-known pottery regions in England, including Buckley, Coal Measures, and Staffordshire (see Figures 21-24). These types are present in small proportions in the first and second phase but do not appear in the two later periods.

The majority of coarse earthenwares across all phases are American-made (Figures 25-28). While many of the American wares could not be attributed to a specific region, the data does show the presence of two regional types, Western Virginia (Figures 29-32) and PHAB wares (Figure 33).

Of the two types, Western Virginia coarse earthenwares occur in greater proportions than the PHAB wares, especially in the second and third phases. It’s interesting to note that the second chronological period has the greatest variety of both imported and American wares. 

The regional type frequencies shown in Figure 20 also indicate a significant increase in the number of unidentified coarse earthenwares during the last two periods (green bar). There are two possible explanations for this pattern. One possible reason for this increase is that some of these sherds came from flower pots but appear in the data because their form could not be definitively identified. The only two projects in this study that contained identifiable flowerpot sherds were the Mansion Perimeter and Building o (Figure 34). Correspondence analysis (see Figure 18) shows that many of the Mansion Perimeter contexts fall within the last two chronological phases which also is where the greatest frequencies of unidentified coarse earthenware sherds occur.

The uptick in unidentified coarse earthenwares also may indicate the introduction of a new coarse earthenware type. In her elemental studies of coarse earthenwares at Monticello, Dr. Lindsay Bloch identified a group of coarse earthenwares whose elemental signature suggested they were produced somewhere in central Virginia. She was unable, however, to verify at the time, that this represented a new regional zone of production (Bloch 2016a, 2016b, 2016c). Further research on local kiln sites and elemental testing of more sherds will expand our ability to identify new ware types and add to our understanding of how local production and markets impacted coarse earthenware trends (see Figures 35-38 for examples of so-far unidentified coarse earthenwares).

Conclusions and Future Research

Comparison of ceramic assemblages from four different households at Monticello shows several trends in coarse earthenware use and acquisition over time. When compared to other ware types, the frequency of coarse earthenwares and stonewares declined at the same time refined earthenwares remained proportionally high and the number of porcelains increased. This finding suggests that households on the plantation, regardless of their socioeconomic background, had access to and acquired ceramics considered to be more costly and desirable than coarse earthenwares and stonewares. 

A further breakdown of coarse earthenware and stoneware data into imported and American-made categories also revealed that while stonewares in general declined, this decrease reflected a significant drop in the number of imported stonewares. American-made stonewares, on the other hand, actually increased in frequency. One possible theory explaining this decline in coarse earthenwares and imported stonewares is that households at Monticello were gradually turning to newer, more durable American-made stonewares as they became more available in the later 18the and early 19th-centuries. This study also showed that Monticello households acquired coarse earthenwares from a wide range of imported and American sources. American wares consistently occurred in greater proportions than English wares, suggesting that American potteries were an important source of coarse earthenwares. While our ability to identify trends in regional sources of coarse earthenwares is somewhat limited by the high frequencies of unidentified sherds, the data clearly shows that residents at Monticello had access to regional sources of these goods. Future studies of coarse earthewares at Monticello will hopefully build on these findings to include data from other site assemblages. 

Future studies of coarse earthenwares also should include documentary research to identify local potters and the various ways in which households at Monticello accessed these goods (Figure 39). This type of information would not only deepen our understanding of the importance of locally made coarse earthenwares but also put a human face on the data. 

Acknowledgements

Special thanks to Corey Sattes and Fraser Neiman for their infinite patience in helping with R code and statistical methods. I would also like to thank fellow analysts Iris Puryear and Miranda LeClerc for their suggestions, support, and collaboration in all lab endeavors.

References

Bates, Lynsey A., Jillian E. Galle, Fraser D. Neiman. 2020. “Building an Archaeological Chronology for Morne Patate. In Archaeology in Dominica: Everyday Ecologies and Economies at Morne Patate, edited by Mark Hauser and Diane Wallman, pp. 64-87. University Press of Florida, Gainesville.

Bell, Allison, Donald Gaylord, and Kristin Sharman. 2019. “’All My Little Might of Money’: Signaling, Structure, and Mobility among the Middling in Nineteenth-Century Virginia and Kentucky,” Historical Archaeology, June 2019, v. 53, No. 2, pp. 372-392.

Bloch, Lindsay. 2011. “An Archaeological Study of Common Coarse Earthenware in the Eighteenth-Century Chesapeake,” MA thesis, University of North Carolina, Chapel Hill. https://www.floridamuseum.ufl.edu/wp-content/uploads/sites/34/2020/09/LBloch-MA.pdf

Bloch, Lindsay. 2015. “Made in America? Ceramics, Credit, and Exchange on Chesapeake Plantations,” Unpub. Ph.D. Dissertation, University of North Carolina.

Bloch, Lindsay. 2016a. “An Elemental Approach to the Distribution of Lead-Glazed Coarse Earthenware in Eighteenth-century Chesapeake,” American Antiquity, April 2016, v. 81, no. 2, 231-252.

Bloch, Lindsay. 2016b. “Reanalysis of Coarse Earthenwares from Williamsburg-Area DAACS Sites.” Unpub. report on file with the Digital Archaeological Archive of Comparative Slavery. 

Bloch, Lindsay. 2016c. “XRF Analysis of Monticello Study Collection,” Unpublished report on file with Monticello Department of Archaeology.

Bloch, Lindsay and Elizabeth Bollwerk. 2024. “Coarse Earthenwares at Flowerdew Hundred: Tools of Colonization in the 17th-Century Chesapeake,” Paper presented at the Southeastern Archaeological Conference, Nov. 13-16, 2024, Williamsburg, VA.

Bloch, Lindsay and Brenda Hornsby Heindl. 2015. “In One Kiln: Technologically Revising Research on Earthenware and Stoneware Production in Virginia,” Paper presented at the Council for Northeast Archaeology annual meeting, Nov. 2015.

Comstock, H.E. 1994. The Pottery of the Shenandoah Valley Region.  The Museum of Southern Decorative Arts. 

Devine, Christine Styrna. 2024. “Using American Coarse Earthenware Types as a Tool for Site Interpretation and Comparison at Monticello,” Paper given at the Southeastern Archaeological Conference, Nov. 13-16, 2024, Williamsburg, VA.

Devine, Christine Styrna. 2025. “Interpreting Coarse Earthenwares at Building o: The Importance of Reanalyzing Legacy Data,” Paper given at the Middle Atlantic Archaeological Conference, March 19-23, 2025, Gettyburg, PA. 

Hill, Martha. 2003. Building o. Background. Digital Archive of Comparative Slavery. https://www.daacs.org/sites/building-o/ Accessed 2/22/26.

Myers, Susan H. 1977. “A Survey of Traditional Pottery Manufacture in the Mid-Atlantic and Northeastern United States,” Northeast Historical Archaeology, Journals for the Council of Northeast Historical Archaeology, v. 6, nos. 1-2, Spring 1977, pp. 1-13.

Neiman, Fraser. 2022. “Site 30: Updates from 2022, https://www.monticello.org/monticello-archaeology-news/site-30-updates, Accessed 2/22/2026

Russ, Kurt C., Robert Hunter, Oliver Mueller Heubach, and Marshall Goodman. 2013. “The Remarkable 19th-Century Stoneware of Virginia’s Lower James River Valley,” Ceramics in America, ed. Robert Hunter, pp. 201-258. 

Wheeler, Derek, and Corey Sattes. 2024. “Reassessing the Chronological Boundaries of Monticello’s Mulberry Row, Charlottesville, Virgina,” Paper presented at the Southeastern Archaeological Conference in Williamsburg, VA, November 13-16.