Interpreting the Coarse Earthenware Assemblage at Building o: The Importance of Reanalyzing Legacy Data

April 1, 2025

Paper presented at the Middle Atlantic Archaeological Conference in Gettysburg, PA, March 20-23

This work builds on a previous paper from 2024. Check it out here: Devine 2024. 

Until recently, historical archaeologists viewed coarse earthenwares as having limited diagnostic value. Many simply categorized them under the general term of “redware” and assumed they were utilitarian vessels. Our approach to coarse earthenwares has begun to change, however, as a result of a growing body of historical research and new analytical tools. Dr. Lindsay Bloch and other scholars have shown how scientific methods such as mass spectrometry combined with documentary research on 18th- and 19th-century pottery workshops can be used to identify specific regional types and trace broader trends in the distribution and consumption of coarse earthenwares (see Figure 1; Magid & Means, 2003; Bloch, 2015; Bloch & Heindl, 2015; Bloch and Bollwerk: 2024: 1, 6-8).

Figure 1. Left: The Pottery Shop, unknown artist, c. 18th-early 19th century, British (source) Right: Colonial Potter, A colonial American potter assisted by indentured servants, c. late 18th century (source).

This paper is a case study showing how these new methods of analysis were used to reassess American and European coarse earthenwares from Building o, an 18th- to early 19th-century quarter site at Monticello (see DAACS page). The study does not include colonoware or native American ware since these represent different pottery traditions. This paper describes the methods used to reassess the collection and compares the new data with previously recorded information in the DAACS database. This study sought to answer 4 research questions:

1. Does the original information for Building o coarse earthenwares change as a result of reanalysis?
2. How does this reanalysis affect our current interpretation of the assemblage or site?
3. How can this data be used to understand broader trends in coarse earthenware use and consumption?
4. Does data change significantly enough to justify the time and effort involved in reassessing archived collections?

The paper concludes with a summary of how reanalysis affected both the data and our interpretation of the Building o assemblage. It also discusses the implications of using new data to better understand broader trends in coarse earthenware use at Monticello and other 18th- and 19th-century sites.

Background

The decision to reassess the Building o assemblage developed out of another research project comparing the coarse earthenwares from our recent excavations of Site 30, a Revolutionary-period quarter site, to assemblages from Building o and four other sites at Monticello (Figure 2). One of the goals of this project was to compare the relative frequencies of American-made and imported coarse earthenwares from the different sites and identify any temporal and spatial variation.

Figure 2. Left: Location of Building o site along Mulberry Row and an aerial view of the structure foundations. Right: Location of Home Farm Quarter Site 30 within the larger Plantation Survey project and an overview of a recently excavated unidentified feature (photos and maps courtesy of the Monticello Department of Archaeology).

Once I ran my data, however, I quickly realized that the artifact records from Building o and several other sites lacked the same level of detailed information as Site 30 (Figures 3-4). The difference in the level of analysis limited my ability to compare certain types of information. The primary reason for this discrepancy had to do with new fields added to the database. From its inception, one of the main goals of the DAACS project was to standardize terms and methods of recording site information to facilitate comparative studies. While basic cataloging protocols have remained constant over time, the database has undergone several updates over the last twenty-four years. The DAACS project added new classification and measurement fields to record physical attributes of coarse earthenwares. One of the most significant updates to the database occurred in 2015 with the introduction of a coarse earthenware typology based on regional ware types. As a result of these changes, older projects like Building o lacked the same fine-grained analysis and CEW type information compared to more recently excavated sites like Site 30. These inconsistencies in data make it difficult to do in-depth, comparative studies across sites (Galle 2018; Bloch 2015).

Figure 3. Coarse earthenware types from Building o (n=178) as cataloged into DAACS in 2001. CEW = coarse earthenware. CEW UnID = coarse earthenware that is not assigned to a manufacturing location. Building o was excavated in 1979-1982 (image by author).

Figure 4. Coarse earthenware types from Site 30 (n=72) as cataloged into DAACS in 2024. CEW = coarse earthenware. CEW UnID = coarse earthenware that is not assigned to a manufacturing location. Site 30 excavations began in 2022 and are ongoing (image by author).

After discussing this problem with my colleagues, we agreed that renanalyzing one of the legacy assemblages would be a useful study to see how the data changed and whether these changes impacted our interpretation of the assemblage and site. I chose the Building o project for several reasons. First, the site had a larger coarse earthenware assemblage than some of the other sites in my initial study. I wanted to use a robust sample size to be able to draw valid conclusions from the new data. I also decided to focus on Building o since it was the first site cataloged into the DAACS database in 2001, which meant analysis of the assemblage occurred before later updates to the database. This reanalysis also ties into a larger reassessment of the site’s chronology being undertaken by members of the Monticello Archaeology Department (see Wheeler and Sattes 2024 blog post).

Figure 5. Graphic illustrating the "constellation of attributes" that impact identifying coarse earthenware types and forms (image by author).

The first step in reanalyzing the Building o coarse earthenwares involved pulling the assemblage and cataloging it using current DAACS cataloging protocols. This involves identifying and recording various physical attributes such as paste color, inclusion types and density, decoration, surface treatment, and vessel form. Along with recording physical attributes, these sherds also were compared to coarse earthenware sherds from Monticello that underwent elemental analysis by Dr. Lindsay Bloch. Dr. Bloch’s 2015 and 2016 elemental studies of American and European coarse earthenwares included samples from several Monticello sites. These samples form the basis of our coarse earthenware type collection. Looking at this is critical to our efforts to identify where a vessel was made (see Figure 5; Bloch:2015; 2016).

Figure 6. Sample of coarse earthenware sherds from Building o (photo by author).

The Building o coarse earthenwares were only assigned a type if they had three of the known physical attributes of a particular region. Sherds that could not be assigned a type were cataloged as “Coarse earthenware, unid.,” which meant that they could be American, English, or European. Sherds were assigned a vessel form based documentary research of the wares produced by known potters as well as key sherd attributes such as rim and base diameter, sherd thickness, rim and body shape, the presence or absence of interior and exterior glaze, and decoration. Mended sherds also proved helpful in identifying form.

Findings

Comparing the original data with the updated information shows significant changes in our interpretation of the assemblage (Figures 3 and 7). One basic change in the overall dataset was the total number of coarse earthenwares in the assemblage. In the course of reanalyzing the assemblage, I discovered that several sherds had been misidentified as coarse earthenwares. There also were several coarse earthenware sherds in the assemblage that came from flowerpots and thus were not included in this study.

Figure 7. Coarse earthenware types from Building o (n=178) as re-cataloged into DAACS in 2025. CEW = coarse earthenware. CEW UnID = coarse earthenware that is not assigned to a manufacturing location. Building o was excavated in 1979-1982 (image by author).

The graphs also shows that the data on the proportion of imported, American, and unidentifiable coarse earthenwares changed significantly as a result of recataloging. The most obvious change was a significant decrease in the number of unidentified coarse earthenwares. Unidentified sherds composed 80% of the assemblage in the initial data; reanalysis resulted in that percentage decreasing to 10%. This major shift can be explained by the significant increase in the proportion of the assemblage identified as American-made. The slight decrease in the number of English and European sherds also reflects the reassignment of sherds to another coarse earthenware type.

Figure 8.Sample of American-made coarse earthenwares and redwares from Building o (photo by author).

This major shift in the number of American-made coarse earthenwares takes on greater meaning when doing a further breakdown of the assemblage into regional types (Figure 9). The new data provides a more nuanced interpretation of the assemblage when compared with the previous study. Western Virginia coarse earthenwares composed almost 2/5 of all the locally-made sherds while 3% of the American sherds were produced in the Philadelphia, Alexandria, Baltimore region. Overall, 83% of the assemblage was American-made while only 7% was imported from England or Europe.

Figure 9. Regional coarse earthenware types from Building o. Left: results from 2001 DAACS cataloging. Right: results from 2025 DAACS re-cataloging. CEW = coarse earthenware; UnID = unidentifiable regional type; Am. = American. The oval is highlighting sherds with unidentifiable regional types; the square from 2025 is highlighting a significant increase in more identifiable types with re-analysis. (image by author).

The assignment of regional types to the Building o assemblage has important implications for our interpretation of the site and the role of American-made coarse earthenwares on 18th- and 19th-century Chesapeake sites (see Figures 10-12). The dominance of these locally-made coarse earthenwares at Building o supports other studies showing that American coarse earthenwares tend to far exceed the number of English and European types during this time period (Bloch: 2015, 9; Bloch and Bollwerk: 2024, 6-7). The prominence of American coarse earthenwares beginning in the 18th century was directly related to disruptions in the Anglo-American trade, a growing demand for goods among American consumers, and an increase in the number of local pottery workshops in early America. Potters who immigrated to both urban and rural areas of the Mid-Atlantic brought with them pottery traditions from England, Germany, and other parts of Europe that they attempted to replicate in their new homeland. The various forms and types of decoration found on these locally produced wares often reflected imported pottery traditions as well as the demands of the market (Comstock: 1994; Magid and Means: 2003).