Virginia Ceramic Studies, A Brief Overview
Archaeologists have long been fond of ceramics because of their abundance in the archaeological record. Unlike their other artifact class cousins, ceramics exhibit a certain tenacity when confronted with the ravages of time; despite the fact that a whole pot may only last a short while before it's dropped and shattered, its pieces can survive thousands of years. Therefore, ceramics have often become diagnostic artifacts for many culture groups, indeed, many ancient cultures are known only by their ceramic assemblages. Because of their durability and abundance ceramics are key pieces in archaeological analysis.
Ceramic analysis begins with basic description. Some of the initial categories include color, temper, or the objects added to the clay, and portion. These categories can then be broken down into more detailed discussions of temper, such as size, and more exact morphological descriptions. When whole pots are recovered then a description of the pots' shape can be added. Other initial categories are various measurements, such as thickness, length and curve of the walls. One can also examine the activity on the surfaces of the sherds or vessel. Surfaces can be treated with objects such as, in the case of Virginia, nets, fabrics or burnishing, which leave impressions on the surface. Or, they can be painted.
From these basic descriptive categories one can then proceed to more sophisticated analyses. The first of these is the construction of typologies. Archaeologists, who are rather fond of typologies, group ceramics exhibiting similar attributes into sets, or wares, and can describe relationships among groups of people based upon their ceramic differences and similarities. Typologies coupled with stratigraphic analysis and scientific dating can then provide insights about the changes in ceramic technology and also the movements and interactions of people. In the past, archaeological analysis of ceramics has relied heavily on typological schemes that assemble ceramics into groups based on characteristics and geographic area. However, more recent work has begun to show the value of attribute analysis as a better alternative. When whole pots are available functional analysis is possible, in which the various shapes can be connected with their specific uses in the domestic or ceremonial contexts, for example. When only sherds are available, which is usually the case, the curvature measurements and residue analysis can also indicate vessel morphology. Surface treatment and decorative analysis can be among some of the most complicated elements of ceramic analysis. Examinations of shapes and patterns can lead to the development of styles and motifs. These styles are then used to discern movements of peoples, reconstructing trade networks and change over time. These sorts of analytical categories are then incorporated into larger questions about settlement patterns, ceremonial analysis, foodways reconstruction and social interaction spheres. Ceramics play an important part of larger questions because of their durability and the potential for highly sophisticated analyses.
The appearance of ceramics around 1000 BC has long been one of the indicators of the Woodland Period in the Mid-Atlantic region. Before the development of ceramic technology, containers were constructed of a variety of materials, such as wood and skins, which are far more vulnerable to the ravages of time. Vessels of steatite and fashioned into rectangular shapes with lug handles survive in far larger numbers. The relative scarcity of these vessels relative to later prolific ceramic numbers has led some researchers to link steatite bowls to ritual feasting and long-distance exchange relations dominated by higher-status individuals (Hantman & Gold 2002: 278). The earliest ceramic forms, such as Marcey Creek Ware (950±95 B.C., uncallibrated) and Bushnell Ware (1110±60 B.C., uncallibrated), exhibited a steatite temper and were constructed in very similar shapes. This shift has been characterized by Jeffrey Hantman and Debra Gold (2002) has a democratization of the steatite trade. They suggest that peoples with easy access to soapstone undercut the trade and elitism by constructing vessels from clay using the steatite as a temper, thus giving greater access to the same shaped vessels to all levels of society (279-280). The quadrangle vessels were replaced with connoidal shapes constructed using coil and scrape methods. During the Early (1000 B.C.-500 B.C.) and Middle (500 B.C. - A.D. 800) Woodland periods widespread experimentation appears to have occurred wit surface treatments and tempers. Surface treatments included net impression, cord marking and smoothing. Sand and lithic materials were the most common temper added to these early ceramics. Clifford Evans (1955) attempted to systematize the observed variations of ceramic types, which was later refined. Evans also defined later ceramics consisting of a shell temper, which were codified into Mockley Ware (about 200 AD), which was present in coastal and interior Virginia and up into New Jersey and Delaware.
Unlike their earlier steatite cousins, these ceramic forms were not reserved for special events or person but were rather used and manipulated in everyday contexts. Of course, variation existed in these ceramic traditions that indicated vessels associated with exceptional circumstances. This shift has been linked with the significant movements in social development discerned in the Mid-Atlantic region. Hantman and Gold link this transition from steatite to clay vessels to the cyclical nature observed in status objects associated with mortuary ritual (2002: 287-289). The tradition of shell tempered wares continued into the Late Woodland, with some exceptions, with noted decreasing in sherd thickness over time. Changes in surface treatment were also noted, moving from cord and net impressions to fabric and simple-stamped paddle markings, or Townsend and Roanoke Wares. The development of these codified ceramic traditions began not in Virginia, but in Delaware.
Perhaps the most critical contribution to the development of ceramic traditions in the Mid-Atlantic region were the two publications by Margaret Blaker (1950, 1963). Her first contribution was a brief article concerning the ceramics recovered from excavations at the Townsend site in Delaware. During her analysis she had discerned a homogeneous ceramic tradition that she dubbed Townsend Ware and subdivided it into five classes. These initial five classes were denoted by their surface treatment and decorative elements. They included: Townsend herringbone, corded horizontal, Townsend incised band, Rappahannock incised and Rappahannock fabric impression. Blaker dated these ceramics from the late prehistoric to the historic and suggested an Iroquoian influence. Further analysis of these ceramics (Blaker 1963) produced more detailed descriptions of the classes. She concluded that she could not determine the chronology of the classes from the stratigraphic record of the Townsend site, however, she suggested that the Rappahannock incised and fabric impressed, being the most frequently occurring, had the longest lifespan of the classes.
Not long after Blaker's initial publication, Clifford Evans responded to what he saw as a paucity of prehistoric consideration in the archaeological musings of the state of the Virginia (1955). Drawing from a total of ninety-six sites, Evans embarked on an extensive and highly detailed analysis of the Native ceramic traditions of Virginia with the goal of reaching beyond basic description and building ceramic series and complexes for the state. He began by sorting each site's ceramics by temper, followed by temper characteristics, firing features and finally by surface treatment. Each series he discerned from the mountains of sherds was meticulously documented and described in his text. However, his devotion to detail and intricacies resulted in so many diverse ceramic series, almost all defined by the river systems, that synthesizing his work is a rather daunting task. While there is something to be said about paying strict attention to the nuances of ceramic attributes, his analysis resulted in a "Virginia ceramic tradition" that was disjointed and overly meticulous. Evans' series and complexes would later be revised and reworked into a more manageable and comprehensive typology.
One of the more significant series for Evans was the Chickahominy Series, which he separated into eight classes based on the surface treatments. Including in the series were the ceramics recovered from the Potts site, significant for its fantastic stratigraphic attributes. It was from this site, and hence from the Chickahominy Series, that Evans formed his chronology of Virginia ceramic traditions. Evans concluded that net impressed and roughened surfaces were the earliest manifestations of surface treatment, which subsequently declined over time, giving way to plain and cord-marked surfaces, the latter of which declining with the onset of fabric impression. Similarly, Evans noted a decline in gravel temper as shell became more popular (1955: 93-94, 97). Unfortunately, current theoretical trends at the time of Evans' publication did not allow for the types of studies that would later emerge in the Middle Atlantic, specifically in Delaware.
Evans also proposed several hypotheses for connecting Virginia ceramics to those of nearby states. His review of archaeological literature indicated that the spread of design ideas was suspected to have originated in the middle Delaware River Valley, disseminating northward towards New York and Connecticut and southward to Virginia. Evans likened early Virginia ceramics to those of New Jersey and Maryland through the attributes he assigned to the Chickahominy Series. These sherds were similar in their basic attributes as well as the incised "V" designs and cord-wrapped dowel impressions Evans saw in the Chickahominy, the main difference being the color of the Maryland sherds. Evans even went so far as to suggest that all shell tempered varieties conformed to Chickahominy Series attributes. Drawing on a rather cursory survey of Maryland studies, Evans concluded that the Virginia coastal tradition extended northward into Maryland and Delaware (1955: 113-114, 117, 120-121). Subsequent research would reveal this to be the case.
While studies of this nature in Virginia were limited to Evans' thorough contribution, research in Delaware continued in which numerous ceramic studies were produced. The first of these significant contributions was by Daniel Griffith (1977). Much of Griffith's discussion of Delaware ceramics is bogged down in explicating the importance and applicability of processual archaeology. Hence, his emphasis on his analysis was on developing a chronology of the ceramic traditions. He places great emphasis on tight data control, especially upon context. He found Blaker's original typology to be lacking, and sought to alleviate the earlier lack of sophistication. While much of his analysis is overburdened with processual dogma, his methodology of ceramic analysis has proved invaluable. Griffith's conclusions about the Townsend chronology have proved to be applicable to those of Virginia and his analysis of the stylistic trends a source of comparative consideration.
As previously stated, one of the key goals of Griffith's thesis was to develop a chronology of Townsend ceramics and corresponding motifs for Delaware. Griffith (1982), as well as Evans (1955), noted that prior to the 1948 excavations at the Townsend site ceramic analysis was confined to pure attribute description, often incomplete. Blaker's analysis was the first step towards a more temporally-conscious analysis. Subsequent analysis of her conclusions and other site analyses refined her early assessments with increasing sophistication and expanded them to include a larger geographic area. Unfortunately, Griffith noticed that this attention to the development of the Townsend typology created a bias against non-Townsend ceramics. With a nod to himself, Griffith stated that this had begun to change, beginning with his thesis and his work with Richard E. Artusy (1977).
Other significant contributions to the Delaware ceramic traditions were produced by Jay F. Custer. In one of his early works Custer suggested that plotting stylistic attributes over areas and across time would better outline prehistoric social interaction spheres (1985). Later collaboration with Griffith led to the conclusion that style had the greatest potential of all ceramic attributes to reveal changes of peoples and cultures through time. Using Griffith's methods from his thesis they examined Townsend and Minguannan ceramics. They noted that between these two ceramic traditions of Delaware there were marked similarities in the observed motifs, despite the possibilities of technical variation. They then turned to Overpeck and Bowmans Brook ceramic types also noting significant similarities. Their comparison culminated in a call for an increase in regional comparisons of ceramic designs: "ceramic designs, when analyzed at this level, are not 'badges' of ethnic groups or common 'traditions,' They are simply stylistic attributes that are sensitive to intergroup interaction" (Griffith & Custer 1985: 18).
These significant Delaware contributions brought into sharp focus the need for comparative consideration, and while this was significant for the state of Delaware, it left Virginia in the dark. Prehistoric cultures, obviously, did not obey modern state boundaries. Ergo, if numerous similarities were noted among the Townsend ceramics in Delaware, then conceivably some discernable stylistic relationship would be manifest with those ceramics in Virginia. This is not to suggest that there was a consistent ethnic identity along the Mid-Atlantic coast, but as Custer and Griffith suggested, an analysis of the technical and stylistic attributes from both states contain the great potential for further explication and elaboration of the social and political relationships of Middle-Atlantic peoples, especially in light of the rise of the Powhatan chiefdom. Presumably, if the technical ceramic traditions can be traced from Delaware to Southern Virginia, then certain stylistic elements would travel as well.
The work of Evans in Virginia long stood as the source for ceramic analysis. However, his exhaustive descriptions of ceramic types were refined in 1982 by Keith Egloff and Stephen Potter. While they concentrated on the coastal plain of Virginia, their descriptions and accompanying photographs combined many of Evans' series into more manageable classes. Their analysis in affect "cleaned-up" the rather unwieldy typology which differed for each river. This resolved the fractious tone of Evans' study and created a more cohesive picture of Virginia Native ceramics.
While work in Delaware appears to have reached a state in which comparison and conclusions regarding the social interaction spheres as well as ethnic and social boundaries could be discerned from the ceramic typological evidence, as it stands now, this may not be the case in Virginia. While a great deal is known about some of the attributes, such as surface treatment, motifs have been limited to attribute description. Current work in Virginia has focused on elucidating social interaction spheres as they relate to ceramic types, not specifically motifs, but focusing on attributes rather than types. Michael Klein's (1994) dissertation employed an absolute seriation method in order to more accurately define ceramic typology. Work conducted by Klein (see also 1997), Hantman and Gold (2002) and Gallivan (2003) have highlighted the fluctuating social hierarchies of Virginia Indians using vessels and their attributes to explicate elite activity. As previously mentioned at the outset of this review, steatite vessels were linked to specialized activities and elite exchange. In looking at the rise of social inequality in the James River system, Gallivan concluded that increased stylistic variation in the Woodland period indicated small social networks. The increase of permanent settlements led to stylistic exchange on the local level to express shifting inequalities. He also found that social interaction between the smaller, discrete units, was becoming more unrestricted. The subsequent increased social heterogeneity necessitated more prestigious goods (Gallivan 2003: 127-142, 151, 175).
Research currently being conducted at the College of William and Mary will undoubtedly contribute significantly to the ceramic knowledge of Virginia. The first of these is a sizeable collection of artifacts resulting from the survey of the Chickahominy River conducted by William and Mary Professor Norman Barka in the late 1960s and 1970s. While final analysis of the ceramics contained therein has not yet been completed, the corpus of sherds suggests the possibility for significant contribution. Current studies in progress include detail morphological and stylistic ceramic analyses as well as the development of a comprehensive GIS database of the survey's results. Also currently under consideration are the ceramics unearthed after the first field seasons at the Werowocomoco site. Among the artifacts resulting from 2003 excavations were numerous examples of Townsend and Roanoke Simple-Stamped Wares, as well as a check-stamped sherd, a rare find in this area of the east coast.
Due to the unfortunate weather occurrences caused by hurricane Isabel in September of 2003, the Jamestown collection of Native ceramics is not currently available for study. However, preliminary perusing of some of the collections and the results of the Jamestown Archaeological Assessment (Blanton, Kandle and Downing 2000) indicate an abundance of shell tempered sherds, most with unidentifiable surface treatments, as well as a few grit and sand tempered pieces (Appendix 1). A brief perusal of some of the sherds contained within the collection before the onset of the storm yielded mostly shell tempered sherds. Surface treatments were either completely eroded, not present or the sherds were too small in size to reach any definitive conclusions. A few cases of Townsend Ware were noted. It is suspected that additional examination of the collection will yield more detailed information. Further conclusions will have to remain in abeyance during the hurricane recovery project.
It appears then that ceramic studies in Virginia have begun to move away from typological construction and into attribute based approaches. The developments of these typologies were a key step in analytical processes, and have been refined to a point that they are logical and wieldy enough to be applied. These typologies have been employed in discerning settlement fluctuations and ceremonial analysis. With the additions of the rather substantial Chickahominy collection and the possibility of a rather significant corpus from the Werowocomoco site, Virginia ceramic analysis will be able to move beyond the typological discussions and apply the extant literature utilizing these typologies and synthesize the information and provide even greater insight into the cultures and practices of the people of the Virginia.
Blaker, Margaret C.
1963 "Aboriginal ceramics; the Townsend Site near Lewes, Delaware." The Archaeolog 15(1):14-29.
Blanton, Dennis, Patricia Kandle and Charles M. Downing
Custer, Jay F.
1987 "Late Woodland Ceramics and Social Boundaries in Southeastern Pennsylvania and the Northern Delmarva Peninsula." Archaeology of Eastern North America 15:13-27.
Egloff, Keith T.
1987 Ceramic Study of Woodland Occupation Along the Clinch and Powell Rivers in Southwest Virginia. Virginia Division of Historical Landmarks, Research Report Series 3:Richmond.
Egloff, Keith T. and Stephen R. Potter
Gallivan, Martin D.
Griffith, Daniel R.
Griffith, Daniel R. and Jay F. Custer
Griffith, Daniel R. and Richard E. Artusy
Hantman, Jeffrey L. and Debra L. Gold
Klein, Michael J.
Quimby, George T.
Rountree, Helen C. (ed.)
Rountree, Helen C.
1990 Pocahontas's People: The Powhatan Indians of Virginia through Four Centuries . Norman:University of Oklahoma Press.
1993 "The Powhatans and the English: A Case of Multiple Conflicting Agendas." Powhatan Foreign Relations 1500-1722. Helen C. Rountree, editor, 173-205. Charlottesville. University Press of Virginia
Rountree, Helen C. and E. Randolph Turner III
2002 Before and After Jamestown: Virginia's Powhatans and their Predecessors . Gainesville: University of Florida Press.
Turner, E. Randolph
1982 "The Virginia Coastal Plain during the Late Woodland Period." Middle and Late Woodland Research in Virginia: A Synthesis. Theodore R. Reinhart and Mary Ellen N. Hodges, editors, 97-136. Richmond: The Dietz Press.
1993 "Native American Protohistoric Interactions in the Powhatan Core Area." Powhatan Foreign Relations 1500-1722. Helen C. Rountree, editor, 76-93. Charlottesville: UniversityPress of Virginia.
Last Updated: 22-Nov-2006