Historical Ceramics, Archaeology, and Working-Class Families at 20SC179

Below is a study that I did last semester for a class that I took on archaeological field and lab methods. I analysed an assemblage of historical ceramics from a site in southeast Michigan, as a preliminary study; I intend on doing more work on the ceramics at the site for my MA thesis. I did not answer all of the questions that I really wanted to answer, due to time constraints; but, I have some time to continue the study in a broader sense. Remember, please do not site without permission.

Working-Class Families and Ceramics: Socioeconomic Standing and Market Access at the Fort Gratoit Site in St. Clair County, Michigan

Julia R. Joblinski

ANT 540

9 December 2019

Introduction

            The main goal of this preliminary analysis was to use of two methods of ceramic analysis – sherd counts and estimation of the minimum number of vessels (MNVs) – to assess socioeconomic standing and market access at the Fort Gratiot site in St. Clair County, Michigan. Due to its location and history, the Fort Gratiot site (20SC179) represents a unique context in which working-class families lived during the 19^th-20^th centuries. Differences in artifact deposition between those at the Fort Gratiot site and other rural sites in Michigan are also noted. Therefore, another goal of this research is to evaluate the usefulness of the ceramic analysis methods mentioned previously for the site’s depositions of historical artifacts.

Ceramic assemblages can be used to help archaeologists better interpret historic archaeological sites. The most common methods include sherd counts and determining the minimum number of vessels present. However, there is much debate over which method is most efficient and better for different site conditions. Additionally, the uses of these methods for interpreting socioeconomic access has been contested. Another goal of this study was to compare MNVs and sherd counts against each other for comparing their usefulness on interpreting the lighthouse keepers’ and their families’ market/economic access. Lastly, this study focuses on just one excavation unit from 20SC179 – H-900 – and future analyses of ceramics from the entire site is necessary in order to gain a fuller understanding of market access of those living on the site.

Background

20SC179

            The Fort Gratiot Light Station (Michigan Archaeological Site Number 20SC179) began operation circa. 1825, and is still used as a Lake Huron lighthouse today. It was the first lighthouse in the state of Michigan, and it was named after the Fort Gratiot which was just 500 yards away from the lighthouse. However, the structure was poorly built, and the lighthouse keeper mentioned that it was dangerous to climb. Over time, the structure began to crack and crumble and during a storm in 1828, the tower was completely destroyed. It was rebuilt in a different location in 1829 (Port Huron Museum 2018). Lighthouse keepers and their families would live on-site in houses adjacent to the lighthouse tower. They kept domesticated farm animals (pigs, chickens, sheep, etc.) and gardens on the site, as well as hunted local game for sustenance. According to the Port Huron Museum (2018), lighthouse keepers did not make much money (see also Surface-Evans et al. 2016).

            In October of 2016, Dr. Surface-Evans and Central Michigan University (CMU) students conducted subsurface and geophysical testing on the Fort Gratiot Light Station grounds with the goal of reconstructing the landscape and figuring out where four buildings – two privies, a shed, and the original lighthouse keeper’s house – were originally located. Archival research, oral histories, shovel test investigations, and geophysical surveys were all used as lines of evidence for locating the old locations of those structures. As a result, the locations of the two privies and original keeper’s dwelling were reconstructed (Surface-Evans et al. 2016). During the 2019 summer field season, Dr. Surface-Evans and CMU students opened test units in those locations. Units that are possible Keeper’s Dwelling locations are H-900 and H-901. Privy units were P-902, P-903, and P-904. The analysis presented here focuses only on Unit H-900, due to time constraints.

Methodological Discourse

            Although analyses of historic ceramics using sherd counts and MNVs are supposedly helpful and produce meaningful results, many have argued for the use of one method over the other (Hull 2007; Orton et al. 1993). Certainly, neither method is completely free of issues. Sherd counts, for example, rely on the types of vessels, which are typically based upon paste and decoration. Depositional and post-depositional processes at the site can cause more sherds to break; the smaller the sherds, the more difficult it is to determine their stylistic attributes. Therefore, if one area of the site has more disturbance, sherd counts can overestimate the amount of certain ceramic wares at sites (Rice 1987; Orton et al. 1993). However, Hull (2007) has demonstrated that using sherd counts were useful for estimating social and economic position at historic rural sites in Ireland.

Calculating MNVs have their downsides as well. MNVs can be strictly quantitative or qualitative, or a blend of both, depending on the nature of the ceramic assemblages in question. MNVs are determined by assuming that sherds sharing the same combinations of attributes belong to the same vessel. However, defining attributes for each MNV grouping must be explicitly clear; otherwise, the whole process of calculation can cause inter-observer error, and the process will not be replicable for future studies (Voss and Allen 2010). Depending on the type of pot and the skill of the analyst, it can be difficult to tell whether two sherds which do not join come from the same pot or not (Orton et al. 1993). Still, Voss and Allen (2010) maintain that MNVs are useful for quantification of mass-produced ceramics, which is useful for estimating socioeconomic position. MNVs also contribute information for depositional and post-depositional site processes.

This study of historic ceramics aims to address and evaluate these methodological issues, while also seeking to learn more about the lives of the people who lived at rural historic sites in southeast Michigan. Analysis of sherd counts and MNVs can provide information about the individuals who inhabited historic sites, especially in rural areas where other historical data are lacking. According to Hull (2007), North American archaeologists working in a wide range of contexts have found a positive correlation between the counts and percentages of flatware in tableware assemblages and economic status. Conversely, a negative correlation between bowls and economic status has been found (Hull 2007). The forms of tableware are, as suggested by the literature, linked to socioeconomic position and dietary trends. Individuals in a higher socioeconomic standing had more access to solid, more nutritious foodstuffs, whereas those in a lower socioeconomic standing likely had access to cheaper, liquid-based foods, which necessitated bowl-like vessels (Hull 2007; Rice 1987). Additionally, archaeologists have demonstrated that the comparative presence, absence, and frequency of expensive or matching ceramic tableware sets within archaeological assemblages can be a means of interpreting socioeconomic position at historic sites (Hull 2007; Majewski and O’Brien 1987). To learn more about the families who lived at 20SC179, both sherd counts and MNVs can be used to analyze the frequencies and absence or presence of matching tableware sets, in addition to certain forms of tableware associated with different socioeconomic situations (Hull 2007; Voss and Allen 2010).

Ceramic Definitions

Since this study relies heavily on the definitions of paste, decoration, and wares, explanations of the differences of each are warranted. “Paste” refers to the composition of the ceramic clay body. The ware categories within “refined earthenware” include whitewares, creamwares, yellowwares, ironstone, and hard- and soft-paste porcelains. These ware categories are defined by their paste (clay) composition, color, and firing temperatures. For example, the overall appearance of hard-paste porcelain is significantly more glass-like as compared to ironstone, which has more of a grainy appearance. Porcelains are made of vitreous clay, meaning that they are fired at significantly higher temperatures than their whiteware counterparts, for example. Whitewares, creamwares, and yellowwares – which are fired at similar temperatures from non-vitreous clays – are usually defined by the color of their pastes. Refined earthenwares were largely used for table service and kitchen use. Some, such as ironstone and yellowware, had more utilitarian household uses, such as food preparation and in bathroom fixtures. Some refined earthenware – porcelain especially – was used for decorative household items as well.

“Unrefined earthenwares” largely include stonewares and redwares. The paste appearances of unrefined earthenwares mostly depend on the material used in the clay; firing processes can also alter color. Their stony or earthy paste compositions set them apart from their refined earthenware counterparts. For example, redwares were typically made with a reddish-brown or buff-colored clay body. Stonewares, as implied by the name, contain heavy temper and have a greyish-brown clay bodies.

The decorative elements of the ceramics found at 20SC179 will be defined in greater detail in the Results section of this study. However, it will be noted here that according to Miller (1980; 1991), different modes of decorations were more expensive than others over time. For example, gilded ceramic wares (or “gilt”, in which gold plating was applied to the glazed surface) was more expensive than sponge-painted ceramic wares in the mid-1800s (Miller 1980). Conversely, the decorative elements of utilitarian ceramics (e.g. wide-mouthed stoneware jars and yellowware mixing bowls) varied much less in cost when it came to their differences in decoration (Miller 1980).

Methods

Since one purpose of this analysis was to compare sherd counts with MNVs, both methods were used. For calculating sherd counts, the following ware categories were recorded systematically: stoneware, redware, creamware, whiteware, yellowware, ironstone, and soft- and hard-paste porcelain. Within each paste category, the following decorative attributes were recorded: undecorated/plain, transfer-printed, hand-painted, molded, sponge-painted, gilt, decal, colored glazed, salt glazed, and tin glazed. Counts of sherds that do not fit any of these categories or attributes were placed in a section titled “other” and were elaborated upon in the form’s comments section. If distinct trade patterns could be seen on the sherds, those were noted as well. Each unit level and the presence of burning were recorded for the purpose of depositional analyses and the assessment of site conditions.

To insure consistency with the recording process for determining the minimal number of vessels present, two different forms were devised. In Voss and Allen’s (2010) synthesis of MNVs, they state that individual sherds may be unique from any other MNV groupings; therefore, just one specific sherd may represent a vessel (2010). It was found that using a separate form for vessels represented by single sherds made the recording process more efficient, and it was useful to separate those from the other MNV counts. Ware categories were recorded, along with the modes of decoration that were mentioned previously. Metrics such as minimum/maximum thickness and rim and base diameter were also recorded, along with maker’s marks, evident burning, and provenience information. If any cross-mended or otherwise affiliated sherds were present from other intraunit proveniences, they were recorded on the form. Fragments that exhibited non-descript or common attributes that could fit into more than one MNV grouping were excluded from the MNV recordation.

When inferring the socioeconomic situation at each of 20SC179, a few methods were employed. According to Hull (2007), the overall number of vessels and the frequency of expensive ceramic types can reflect economic and social status at historic sites. Therefore, the frequencies of expensive or non-expensive ceramics were calculated. Additionally, the forms of specific pottery wares may be linked to socioeconomic status (Hull 2007; Rice 1987). When possible, the forms of sherds were recorded during analysis; therefore, the frequencies of flatware and bowls were taken into account as well. When applicable, sherds exhibiting observable maker’s marks were traced to the original manufacturer using one of the following: Godden’s (1964) Encyclopedia of British Pottery and Porcelain Marks, Kowalsky and Kowalsky’s Encyclopedia of Marks on American, English, and European Earthenware, Ironstone, and Stoneware (1999), and Lehner’s Encyclopedia of U.S. Marks on Pottery, Porcelain and Clay (1988).

Results

Sherd Counts

Table 1: Total Sherd Counts by Ware Category
	Soft-Paste Porcelain	Whiteware	Creamware	Yellowware	Redware	Ironstone	Stoneware	Total
N	3	642	6	3	36	1	4	692
%	0.4	92.8	0.9	0.4	5.2	0.1	0.6	100

            Several observations about the quantities of cost-varying ware categories can be made. The total number of sherds from Unit H-900 is 692. Of those, 642 (92.8%) are whitewares; 36 (5.2%) are redwares; six (0.9%) are creamwares; four (0.6%) are stonewares; three are soft-paste porcelains (0.43%); three are yellowwares (0.4%); and one is ironstone (0.1%) (Table 1). The ceramics in H-900 imply a dearth in the most expensive ceramic materials at 20SC179. For example, Miller (1980) shows that porcelains were among the most expensive ceramic wares that were available during the 1800s. Furthermore, soft-paste porcelains were of lesser quality and cost compared to hard-paste porcelains. Quantities of lower-cost refined earthenwares such as yellowware are relatively low as well as utilitarian stoneware. Due to the high numbers of whiteware sherds in comparison to other ware categories, a study of the different cost-varying modes of decoration on whitewares through sherd counts is warranted.

Table 2: Total Whiteware Counts by Decorative Category
	Undecorated	Hand-Painted	Transfer-Print	Dipped / Annular	Sponge-Painted	Molded	Blue Shell Edge	Total
N	319	53	132	31	54	12	6	642
%	49.7	8.3	20.6	4.8	8.4	1.9	0.9	100

            As already indicated above, 642 whiteware sherds were recovered in Unit H-900. Of those, the following decorative elements were analyzed: 319 (49.7%) non-descript/undecorated sherds; 132 (20.6%) were transfer-printed sherds; 54 (8.4%) were sponge-painted sherds; 53 (8.3%) were hand-painted sherds; 31 (4.8%) were dipped or annular decorated sherds; 12 (1.9%) were molded sherds; and six (0.9%) sherds exhibited blue shell edge decoration. According to Miller (1980), gilt and hand-painted motifs were among the more expensive decorations on whiteware vessels. Alternatively, sponge-painted and plain wares were on the low end of cost; transfer-printed and dipped/annular-slipped wares were mid-ranged in price (Miller 1980). Just through considering the vast number of undecorated whitewares compared to the total numbers of other wares categories, one might make the interpretation that the lighthouse keeper families had better access to ceramics that were lower in cost, since plain wares cost less to manufacture (Miller 1980). However, due to the amount of breakage of these ceramic sherds, the disproportionately high number of undecorated whitewares is most likely representative of the portions of vessels that are simply not decorated with anything.

Minimum Number of Vessels

           

Table 3: Vessel Quantities by Ware Category
	Soft-Paste Porcelain	Whiteware	Creamware	Yellowware	Redware	Ironstone	Stoneware	Total
N	3	52	1	2	7	1	2	68
%	4.4	76.5	1.5	3	10.3	1.5	3	100

The total estimation of vessels in Unit H-900 is 68. Of those, 52 (76.5%) are whitewares; one (1.5%) is creamware; two (3%) are yellowware; seven (10.3%) are redwares; one (1.5%) is ironstone; and two (3%) are stonewares (Table 3). Like the sherd counts, there are a significantly larger amounts of whiteware vessels when compared to the other ware categories. Therefore, due to these large amounts, counts of the different types of decoration on whiteware vessels were sorted out (Table 4).

Table 4: Whiteware Vessel Quantities by Decorative Category
	Undecorated	Hand-Painted	Sponge-Painted	Flow Blue	Blue Shell Edge	Transfer-Printed	Dipped / Annular	Molded	Other	Total
N	4	6	4	1	6	17	2	9	3	52
%	7.7	11.5	7.7	1.9	11.5	32.7	3.8	17.3	5.8	100

As indicated by the table above, undecorated whitewares make up 7.7% of all whiteware vessels; there are six (11.5%) hand-painted vessels (one of which was burnt); four (7.7%) sponge-painted vessels; one (1.9%) flow blue vessel, which was also heavily burnt; six (11.5%) blue shell edge flatware vessels; 17 (32.7%) were transfer-printed vessels; two (3.8%) were dipped/annular wares; nine (17.3%) were molded vessels; and three (5.8%) vessels were placed into an “Other” category. The “Other” category consisted of vessels that combined decorative attributes together in more obvious ways. These consisted of a molded transfer printed vessel, a molded vessel with colored glaze, and a molded vessel with dipped decoration.

Unlike the sherd counts, the estimation of MNVs helped make the presence of matching sets more obvious. For example, two black transfer printed vessels – a plate and a shallow bowl – were pieced together during the MNV recordation process (Figure 1 and Figure 2). This black transfer printed pattern is called “Rhone Scenery” and was manufactured by T. J. & J. Mayer circa. 1845-1850 (Kowalsky and Kowalsky 1999).  Other matching sets were recorded, and consisted of hand-painted saucers and bowls, sponge-painted bowls and teacups, and dipped/annular slipware. According to Hull (2007), the presence of matching sets may indicate that households had enough means to purchase more than one type of ceramic vessel of the same type in a single event. Since there is evidence for this at 20SC179, it is suggested here that the lighthouse keepers’ families at least generated enough income to purchase multiple matching sets for household use.

"Rhone Scenery" plate

"Rhone Scenery" shallow bowl

T.J. & J. Mayer maker's mark

Conclusions and Future Analyses

It is apparent that in both the sherd counts and MNV estimations that transfer-printed ceramics are the most abundant ceramic category within Unit H-900. Interestingly, both the sherd counts and MNVs indicated relatively similar counts for other ware categories when comparing the two methods against one another. Additionally, effects of post-depositional processes – i.e. excessive sherd breakage – slightly impeded the sherd counts, due to a slight bias in the counts of undecorated sherds. The presence of matching sets – which were noted during the MNV process – was useful, since it implies that the families at 20SC179 had at least enough economic means to purchase ceramics in matching sets. Therefore, MNVs were more useful for interpreting socioeconomic standing at 20SC179.

The results of this preliminary study seem to suggest that the lighthouse keepers and their families had access to ceramics that were lower- to mid-ranged in cost. These loose estimations were made using sherd counts and MNVs as presented here, and the economic scaling of ceramics as outlined in works by Miller (1980; 1991). Interestingly, both the sherd counts and revealed that whiteware ceramics – particularly transfer-printed whiteware vessels – were most abundant at the site.

            Additional work needs to be done to make this study more complete. In order to properly interpret the socioeconomic standing of the lighthouse keepers’ families through time at 20SC179 via ceramic artifacts, for example, the analysis of ceramic assemblages from the other units (H-901, P-902, P-903 and P-904) need to be analyzed. Furthermore, a better understanding of which local markets were available (both economically and location-wise) to the Fort Gratiot Light Station families is necessary. Lastly, archival research and knowing more about the economic situation of those living on-site will help interpret how these working-class families lived.

REFERENCES CITED

Godden, Geoffrey A.
1964.   Encyclopedia of British Pottery and Porcelain Marks. Bonanza Books, New York, NY.

Hull, K. L.
2007.   Beyond the Mean Ceramic Date: the Interpretive Potential of Historic Ceramics in Cultural Resource Management. Ontario Archaeology, 83/84: 80-91.

Kowalsky, A. A. and D. E. Kowalsky
1999.    Encyclopedia of Marks on American, English, and European Earthenware, Ironstone, and Stoneware (1780-1880). Schiffer Publishing LTD., Atglen, PA.

Lehner, L.
1988.   U.S. Marks on Pottery, Porcelain & Clay. Collector Books, Paducah, KY.

Majewski, Teresita and Michael J. O’Brien.
1987.   The Use and Misuse of Nineteenth-Century English and American Ceramics in Archaeological Analysis. In Advances in Archaeological Method and Theory, Vol. 11. Academic Press, Inc., Cambridge, UK.

Miller, George L.
1980.   Classification and Economic Scaling of 19^th Century Ceramics. Historical Archaeology, 14: 1-40.

Miller, George L.
1991.   A Revised Set of CC Index Values for Classification and Economic Scaling of English Ceramics from 1787 to 1880. Historical Archaeology, 25 No. 1: 1-25.

Orton, C., P. Tylers, and A. Vince.
1993.   Pottery in Archaeology. Cambridge University Press, Cambridge, UK.

Rice, Prudence M.
1987.   Pottery Analysis: A Sourcebook. University of Chicago Press, Chicago, IL.

Surface-Evans, S. L., J. Bentley, J. Cunningham, M. Cyrus, A. Halula, C. Herron, K. McDonald, L. Nelson, A. Puskas, H. Quinn, E. Reardon, G. Swallow, R. Williamson.
2016.   Comprehensive Cultural Resources Management Report for the Fort Gratiot Lighthouse Park & Museum. Cultural resource management report submitted to the St. Clair County Parks & Recreation and Port Huron Museum of Arts and History.

Voss, B. L. and R. Allen.
2010.   Guide to Ceramic MNV Calculation Qualitative and Quantitative Analysis. Technical Briefs in Historical Archaeology, 5: 1-9.