Colonial Williamsburg Digital Library

Introduction

The Colonial Williamsburg Foundation Department of Archaeological Research conducted a Phase III archaeological investigation of site CG-3, located in James City County approximately seven miles east of the City of Williamsburg (Figure 1). CG-3 was assigned state designation 44JC633 by the Virginia Department of Historic Resources. This tract of land, associated with the property of Carter's Grove Plantation, has been deeded to James City County for the construction of a public school building. The purpose of this project was to thoroughly examine a relatively small but intact aboriginal site that would be impacted by this construction. Sites of this type are generally discounted as "ephemeral lithic scatters" since little information is known about their importance and relationship to regional settlement patterns (Blanton 1991).

At approximately 30 meters above mean sea level (amsl), site 44JC633 is situated atop the eastern half of an elevated landform or ridge located between two major ravines. The site is positioned just south of existing Route 60 in the Grove Community of James City County (Figure 2). A prehistoric site, 44JC633 measures approximately 100 square meters. Laboratory analysis of lithic distribution and site stratigraphy indicates a relatively intact site. Most evidence of plowing remains above the prehistoric layers. Other soil processes, like floralturbation have probably taken place in the area. Some basic patterning of lithic materials, however, revealed through the excavation of small units within a confined area, seems to suggest an intact site. The site is divided into two loci based on distinct lithic concentrations and are identified in this report as Locus-1 and Locus-2.

Little evidence of primary, final stage, or maintenance reduction activities were found during laboratory analysis. Secondary thinning flakes and flake fragments and shatter comprised most of the debitage. Additionally, tested cobbles, informal groundstones (hammerstones), an informal tool (utilized flake), hafted bifaces, bifaces, cores, fire-cracked rock, and miscellaneous unmodified stones were recovered from the excavations. Patterns of lithic reduction, material types, lack of features, extent of intra-site mending and refits, and site size and structure, clearly indicate very short-term episodes of occupation and possibly reoccupation through time. Specifically, this evidence seems to suggest two limited function encampments in which individuals engaged primarily in intermediate stage reduction of quartzite cores or bifaces (Locus-1) and more subsistence-oriented activities (Locus-2). A minimum of 19 distinct blocks of material were reduced in varying degrees at the site with 36.8% (n=7) at Locus-1 and 63.2% (n=12) at Locus-2.

The Colonial Williamsburg Foundation Department of Archaeological Research conducted Phase I and Phase II archaeological investigations of site 44JC633 in February, 1991 and from April, 1991 to May, 1991 respectively. The site was first identified during the Phase II cultural resource survey. Additional, extended investigations of the site were conducted from June, 1991 to September, 1991. These investigations were associated with the proposed school construction in accordance with an agreement with James City 2 FIGURE 1 Site Location on the Locust Grove Tract of Carter's Grove. County. The purpose of Phase I and II surveys and Phase III excavation at this site was to examine the structure and patterns of a small, intact prehistoric site on Virginia's Coastal Plain.

Other Archaic and Woodland Period aboriginal sites (Figure 3) are within close proximity to 44JC633, indicating the area was conducive to hunting and gathering, among other prehistoric activities. Seasonal occupancy and possibly base settlement of Native Americans in and around this upper part of the James-York Peninsula appears to have spanned many generations in time.

This project was directed by J. Michael Bradshaw whose responsibilities included fieldwork, artifact analysis, and report preparation. David Muraca, Steve Atkins, Beau Harbin, Gunnar Brockett, and Brad McDonald conducted Phase I and II testing of the site. Students from the College of William and Mary Summer Field School, Corey Judson, Audrey Horning-Kossler, Elizabeth Anderson, Mike Jarvis, and Dagmar Von Toal served as field supervisors and assistants. Mary Clemons, Carol O'Melia, Brian O'Melia, Sandy 3 FIGURE 2
Project Area and Environs.
(U.S.G.S. 7.5 Minute Yorktown Quadrangle, U.S.G.S. 7.5 Minute Hog Island Quadrangle) FIGURE 3
Prehistoric and Historic Site Locations in the Vicinity of CG-3 (44JC633). 4 Bradshaw, Missy Bradshaw, Joseph Robertson, Bill Kossler, Scott Hetrick, Christine Bowen, Shannon Bradshaw, Robin Fehnel, Dennis Cotner, and Martha Moore worked in a voluntary capacity as excavators. Laboratory processing of artifacts was supervised by William Pittman, Amy Kowalski, and Pegeen McLaughlin-Pullins. Analysis of artifacts was conducted by J. Michael Bradshaw, Pegeen McLaughlin-Pullins, Michael Collier, and Lisa Youngers. All phases of the archaeology were accomplished under the general supervision of Marley R. Brown, III, Director of Colonial Williamsburg's Department of Archaeological Research.

Environmental Setting

Carter's Grove plantation comprises approximately 800 acres situated on the north bank of the James River, between Grice's Run (which borders the property to the southeast and Wareham's Run, a little over one mile to the northwest. Carter's Grove centers on a neck of high ground, separated from the adjoining countryside by ravines and swamps-characteristic of land well-suited for agricultural activity. The distinct high-ground topographic feature known as the Kingsmill scarp is an ancient beach head, trending east-west and fronting onto the James River. This high ground gently slopes to an ancient river terrace approximately thirty feet above sea level. Most of the direct river frontage from that point consists of steep, eroded bluffs.

Historically, this particular area appears to have been mostly open farmland, cultivated over the past several hundred years. Little or no agricultural activity is being conducted in this area at present. The wooded area in the southern half of the project area is covered by mixed deciduous and conifer forests of loblolly pine, oak, spruce, fir, cedar and holly, with an understory of dense honeysuckle, various ivies, and greenbrier. A large logging project has just been completed to the south of the project area.

The climate is typical of the east-central portion of the Virginia Coastal Plain, where the average winter temperature is 41 degrees Fahrenheit and the average spring/summer temperature 76 degrees. The average relative humidity ranges from 80% or less in the morning to 60% or less in the afternoon. Prevailing winds are generally strong (over fifteen mph), mostly originating from the southwest.

Erosion appears to have been a significant process in the development of the present- day topography, with slopes of the scarps receiving the greatest impact. Coupled with annual plowing, from preceding years, no archaeological material is expected to survive on any slopes. Natural erosion, particularly in areas along rivers and streams, appears more substantial than in other regions. Plowed fields left exposed to wind and water processes were generally devastated by these erosional elements. The severe storms characteristic of the James River area are probably the most significant natural earth-moving factors.

Specific Morphology

The topography of southeastern Virginia is characterized by a succession of coastal and riverine scarps and terraces. The terraces are emergent plains formed under stream, estuarine, bay, swamp and marsh conditions during the late Pliocene and Pleistocene; they decrease in elevation seaward and toward the major rivers. The scarps, which were cut by shoreline erosion, maintain remarkably uniform elevations all along the Atlantic Coastal Plain (Cooke 1931). The number, origin and age of the terraces has been the subject of much controversy (Cooke 1931; Flint 1940; Oaks and Coch 1973).

The Surry scarp (Flint 1940), which passes north-south through Williamsburg, marks the boundary between the Middle and Lower Coastal Plain. The toe of the scarp is at an elevation of about 29 meters and is cut into Pliocene-age strata. The Lackey plain (Johnson 1972) extends from the Surry scarp eastward to the Ruthville scarp and trends northeast- southwest across the innermost Lower Coastal Plain. The plain, which reaches its maximum elevation of 27 meters against the Surry scarp, slopes eastward to about 25 meters at the crest of the Ruthville scarp. On the James-York Peninsula and near major rivers elsewhere, the plain is extensively dissected. The Surry scarp was a fastland beach and the Lackey plain was covered by open bay when the Windsor Formation was being deposited during the Early Pleistocene.

A low, subdued scarp—the Ruthville—forms the boundary between the higher plains in southeastern Virginia and the Grove plain. The Grove plain, on which Carter's Grove mansion is constructed, extends from the base of the Ruthville scarp seaward to the Lee Hall or lower scarps and varies in elevation from 22 to 24 meters. The Grove plain was formed as an estuarine-bay plain during the Early Pleistocene.

The Kingsmill scarp is the most continuous and, in many places, the most prominent scarp along major rivers in the Coastal Plain. It trends east-west and forms the declivity between the upland upon which the Carter's Grove mansion is built and the flat below. The base of the scarp ranges in elevation from 13 to 14 meters. The Huntington flat (Coch 1971) is bounded landward by the Kingsmill scarp and by the James River at Carter's Grove, and ranges in elevation from about 10.5 to 14 meters. The plain was formed during the late Middle Pleistocene by the ancestral James River estuary and adjoining ancestral- Chesapeake Bay. Severe wind and water erosion along the James River created the existing bluffs and beach. An underlying ancient bed of water-deposited chert and quartzite 7 cobbles is currently eroding from the bluffs. The proximity of this layer to the site suggests this was one of several sources of lithic materials for tool making by the Native Americans.

Previous Archaeology

Both prehistoric and historic sites have been discovered in the area of Carter's Grove Plantation. Fourteen state site survey forms from this area are on file with the Virginia Department of Historic Resources in Richmond. Archaeological examinations started with William Kelso's intensive survey of the property in 1971. The subsequent excavations have been led by various people under the direction of Ivor Noël Hume from 1971 to 1984, and by Marley R. Brown III since 1984. The records generated by the pre-1984 excavations are still being analyzed by Noël Hume. The lack of completed reports or other documentation pertaining to these sites means that they can be described in only the most general of terms.

The Carter's Grove mansion itself and its associated archaeological remains are a major eighteenth-century archaeological resource. Built in the 1750s, this colonial estate was constructed by Carter Burwell and occupied by his descendants until 1838. Several outbuildings which were associated with this dwelling survive archaeologically and were assigned site survey number 44JC109. The features uncovered by William Kelso's 1971 survey include fencelines, ditches, terraces, formal gardens, paths, a possible dairy, another dwelling, and an icehouse. All features dated to the eighteenth century. Large portions of this site were investigated using a combination of machine trenches and hand excavations. The formal gardens, the dairy, the late eighteenth-century dwelling, and the ice pits were completely excavated.

Several features that were at the time thought to be tanning pits were found and excavated at this time as Site 44JC110. Subsequent excavations in this region have revealed that these pits actually represent root cellars associated with colonial slave quarters. More than twenty late eighteenth-century root cellars were identified. Some pits were wood- lined, and all contained domestic refuse. The entire area around the pits was excavated, except for a ravine located nearby that may still contain domestic refuse from this occupation.

To the southeast of the mansion, five areas of large burnt subsoil were uncovered and interpreted as brick clamps (Site 44JC111). One area contained datable artifacts that suggested a 1740-1750 date range. Several features including clay pits and post holes were found in association with these burned areas. Two wells were also excavated from this area. The entire complex was fully excavated in 1971 by William Kelso.

Several first- and second-quarter seventeenth-century sites were uncovered and excavated in the 1970s. Sites A, C, D, E, and H were part of a large early settlement known as Martin's Hundred and are described in detail in Noël Hume's 1982 book Martin's Hundred. The remains of a fortified early administrative complex, known as Wolstenholme Town, as well as several outlying dwellings, were examined. This book includes a description of the everyday life of the colonists, as well as the evidence of a violent conflict between Native Americans and the European settlers in 1622. The sites that have been 8 assigned state site numbers include 44JC120, a seventeenth-century site of unknown function located northwest of Wolstenholme Town; Site D (44JC114), a small post structure; Site A (44JC116), a structural complex that contained buildings, fences, and graves; Site C (44JC115), the company compound and fort; and Site E (44JC117), a single seventeenth- century structure located to the north of Wolstenholme Town.

Prehistoric sites excavated in this area include a shell midden and Woodland ossuary (44JC119), and another shell midden (44JC118). Both are located on an ancient river bank near where the archaeological museum now stands. Several different individuals have examined this ridge, including the Virginia Research Center for Archaeology's Keith Bolt, who explored the western part of the ridge; Norman Barka of the College of William and Mary, who excavated one small square in the ossuary; and David Muraca of Colonial Williamsburg's Department of Archaeological Research, who in 1987 excavated the eastern part of the Woodland site where the archaeological museum now stands. In addition to excavating this prehistoric complex, this last excavation uncovered two colonial barns. Several other shell middens have been identified on this property, including site 44JC130 on the bank of the James River. Erosion threatened, and may have since destroyed, this site. A large Middle Woodland site (44JC158) was excavated where the Carter's Grove Visitor Center now stands, as well as a large colonial post structure interpreted as a barn, located under the parking lot for the Visitors Center. This excavation was directed by Robin Duffy under the supervision of Ivor Noël Hume.

Paleo-Indian Period

Paleo-Indian populations were organized into bands of about 200 individuals. These people were highly mobile hunters and gatherers who depended on deer, elk and moose, as well as small mammals and fish for subsistence (Custer 1990). The flexibility of these bands 12 allowed for adaptation to seasonal fluctuations as well as environmental changes over longer periods of time (Moodey 1992).

The Paleo-Indian toolkit consisted mainly of implements made from microcrystalline or cryptocrystalline silicate lithic materials, particularly chert and jasper (Gardner 1989; Reinhart 1989). Materials of this type are conducive to the high quality of workmanship characteristic of points and tools from this period (Gardner 1986). Consequently, cherts, jaspers, and related lithic materials were highly prized and became the focus of extensive procurement excursions. It is also assumed that this material became an important item of trade (Reinhart 1989). The lithic technology that produced the noted fluted projectile points and associated debitage is the basis for establishing the Paleo division of prehistoric time (Johnson 1989).

The paleoenvironment of the Middle Atlantic region consisted of complex boreal forests (Custer 1990). Some have argued, however, that this region was a complex mixture of grasslands and boreal and deciduous woodlands (Boyd 1989; Gardner 1989).⁴

Settlement patterns of Paleo-Indians have been somewhat difficult to establish because of the rarity of sites. A single, fluted projectile point usually constitutes a site. Approximately 50-75 Paleo-Indian sites have been identified in Virginia, with perhaps ten located in the Piedmont and Coastal Plain provinces (Turner 1989). Gardner (1981), proposed a model of settlement patterning comprised of five functionally distinct site types: 1) quarries; 2) quarry reduction stations; 3) quarry-related based camps; 4) base camp maintenance stations; and 5) outlying hunting sites (Turner 1989).

Archaic Period

Indians of the Archaic Period were also mobile hunters and gatherers with seasonal treks to specific areas for certain foods and lithic material. It is often debated that the earliest part of Archaic time was actually a continuation of the Late Paleo-Indian Period (Gardner 1989). Although a certain amount of Paleo-Indian culture probably extended into the Archaic Period, there were also important episodes of culture change (Custer 1990).⁵

Likewise, environmental change occurred during this time but it is unlikely that the transition was as dramatic as once believed. Because of a poorly developed data base, however, the Archaic Period of Virginia's prehistory is the time range where the least is known (Custer 1989).

The term Archaic was first used by William Ritchie in describing cultural periods of prehistory in the State of New York (Ritchie 1932; Gardner 1986). In Virginia, Archaic denotes both a time frame and a stage in cultural development. As previously stated, the Archaic Period was a time of mobile hunters and gatherers. Therefore, most characterizations 13 of Archaic time rely on a cultural perspective void of settled villages, agriculture, and ceramic technology.

Floral remains both in the archaeological and geological records have contributed extensively to the interpretation of Archaic environments and available animal species (Graham and Mead 1987; Custer 1990; Brush, Lenk, and Smith 1980; Watts 1979; Davis 1983; Gadreau and Webb 1985). Custer (1990), suggests that animal species of the Archaic parallel those of today; however, a variance of species density and distribution occurred in time and space.

The rate of environmental change is very slow and spans generations of human existence. Given the fact that residual changes in conditions included a more temperate climate with increased moisture, a greater variance in seasons, and a rise in sea level (Belknap and Kraft 1977), Early and Middle Archaic environments paralleled that of the Paleo-Indian Period (Custer 1990). The environmental transition, then, as it affected prehistoric lifeways and adaptations was more a cumulative effect with the resulting cultural change more distinct in later Archaic time (Gardner 1986).

Very little information is available on settlement patterns and site adaptations of the Archaic Period, especially in the Tidewater region of Virginia's Coastal Plain. Most archaeological data derives from surveys and excavations conducted in the western part of Virginia, specifically those in the Flint Run Complex in the Shenandoah Valley (Carr 1986; Gardner 1989; Custer 1990).

The settlement strategies for sites above the Fall Line Zone are commonly linked to raw lithic material resources and their procurement (Gardner 1989; Custer 1990). In contrast, studies of Archaic sites below the Fall Line Zone focus on their proximity to swamps, bogs, freshwater springs and other wetland areas (Custer 1982; Wanser 1982; Rappleye and Gardner 1979; Custer and Bachman 1986).

Gardner (1977) proposed models for five Archaic site types which have been developed into two models of settlement systems (Custer, Cavallo, and Stewart 1983). These are:

• 1. Quarry Sites-An area for the initial procurement of lithic materials.
• 2.Quarry Reduction Sites-An area where bifacial preforms, flakes, and cores were readied for transport.
• 3. Base Camps-Primary habitation site.
• 4. Base Camp Maintenance Stations-Resource procurement sites within close proximity to the base camp. These areas are revisited by inhabitants of the base camp.
• 5. Outlying Hunting Sites-Resource procurement sites farther away from the base camp.

In the two settlement models, Archaic groups are placed in either a cyclical pattern of movement based on lithic resources and seasonality or in a serial model where movement does not depend on raw lithic material resources (Figures 5 and 6).

FIGURE 5
Cyclical Settlement Model.⁶

FIGURE 6
Serial Settlement Model.⁷

Woodland Period

The Woodland Period is always associated with the development of pottery and the practice of horticulture. Villages were also constructed. A greater movement of other cultural groups into Virginia occurred during this time. Specifically, contact took place between the Mississippian, Adena and Hopewell cultures. Once contact was made with the Europeans, metal tools and other European trade goods entered the Native American culture.

Stone tools of this period took on new characteristics with the triangle type being the most prominent. Arrows made their appearance about five hundred years prior to European contact. Compared with other time periods, Custer (1989) suggests "more varied projectile point forms" derive from the Woodland Period. Numerous projectile point typologies have been developed for Woodland Period tools but not, however, without argument.⁸

With the advent of horticulture, lifestyles of the Woodland peoples became more sedentary. Social systems grew in complexity as populations increased and cultural exchange developed. Custer (1989) also characterizes the Woodland Period as a time that saw "less portable storage technologies and non-transportable facilities ... stratified societies, elaborate exchange systems, and complex burial patterns."

Lithic Scatter Research: An Overview⁹

Until recently, archaeologists have avoided exploration of small, prehistoric sites on the Coastal Plain of Virginia. Sites of this type are generally categorized as "lithic scatters" and deemed insignificant. Neglecting to conduct archaeological excavations and analysis of such sites has, over a period of time, created a gap in the archaeological record. Site 44JC633 is significant in that the information derived will help to bridge that gap in Virginia prehistoric archaeology.

Lithic scatters are quite abundant and have been considered the most common type of "site" that relates to human habitation in prehistoric time. Blanton et al. (1991) describes lithic scatters as being a low density of prehistoric artifacts, dispersed over an area and consisting mostly, if not all, of lithic material. Caldwell (1954) and Coe (1964) noted that these sites exist and provided a general description as being Archaic and derived from short-term occupancy. Otherwise, they indicate these sites are of little importance and use in the interpretation of the prehistoric record. This was to change, however, with the advent of mandated cultural resource management (CRM) studies. Canouts and Goodyear (1985) thoroughly document this trend and go on to describe how resultant documentation of such sites by the hundreds compelled archaeologists to consider them more carefully. This was particularly true in the Piedmont region and ultimately models accounting for these sites were developed (House and Ballenger 1976, Goodyear, House, and Ackerly 1979) and even site-specific studies were conducted (House and Wogaman 1978; Goodyear, House and Ackerly 1979:77-88).

Particularly exemplary of earlier areal studies involving these sites are the highway corridor surveys in South Carolina conducted by House and Ballenger (1976) and Goodyear, House and Ackerly (1979). In large measure these studies developed and explored tenets of a riverine-interriverine settlement model. In this context the numerous, small interior sites were expected to represent short-term procurement camps, either for the extraction of nuts or white-tailed deer; larger sites adjacent to larger streams were recognized as likely habitation/base camp sites. Criteria such as artifact density, assemblage diversity, site size/complexity were utilized in evaluating the function of individual sites. Ultimately, however, site-specific investigations were called for to better address such issues.

Still prominent among the few site-specific studies of lithic scatters were the investigations of the Windy Ridge (House and Wogaman 1978) and Stoddard (Goodyear, House, and Ackerly 1979) sites which grew out of those survey projects. At the Windy Ridge Site (38FA118) a block excavation was opened to evaluate the potential for preserved activity 18 areas. Contrasting distributions of raw materials were documented and spatial congruence between Morrow Mountain hafted bifaces and quartz debitage was observed. These results established that such sites can retain some internal structure, but in not all cases will the ;significance of the patterns be readily apparent. Based on assemblage composition and setting, the site was interpreted as the locus of repeated hunting camps for the taking of white-tailed deer.

At the Stoddard Site (38LU42-Locus 3), a stratified unaligned surface sample was obtained. While the overall sample was small, the results of the project were nonetheless promising. What they indicated is that within an area defined as a single site, the distribution of artifacts, including diagnostic items, was not uniform or random. Instead, Middle and Late Archaic material was centered in different areas of the site. As a result it is apparent that even cultivated and eroded upland sites can retain at least vestiges of their structure and cannot be dismissed as having no information potential.

Both Windy Ridge and Stoddard were reoccupied locations comprised of multiple loci, and in the case of Windy Ridge only one of these loci was excavated. Such has been the norm in small site studies to date, in the sense that the richest, most obvious "lithic scatters" are chosen for investigation beyond the survey level. A hidden limitation, however, is that the complexity of sites created by multiple occupations acts to obscure patterns and require tentative interpretations. Lacking have been investigations of the smallest, lower density sites or loci representing a single occupation, which potentially stand to yield the most useful, uncluttered information.

Archaeological Character of Lithic Scatters

Despite the progress that has been made over the past two decades in understanding the role of these sites, it is still appropriate, albeit cliche, to declare that there is still much to know. Before addressing the topics which merit focused attention, it will be helpful to summarize the current state of knowledge regarding "lithic scatters."

If nothing else it has been sufficiently documented that sites of this type are, indeed, very common and seemingly ubiquitous regardless of physiographic region. Attention was first drawn to them via investigations in the Piedmont but studies in both the Coastal Plain and mountainous regions have identified many such sites as well. Also, while within these regions lithic scatters are proportionally more common in the interriverine areas relative to large, "habitation" or "base camp" sites, it has become abundantly clear that they frequently occur in the vicinity of major drainages. In effect, we can assert unequivocally that these sites are the most common manifestation of prehistoric activity.

Through review of the literature it is clear that "lithic scatters" are not everywhere alike and cannot be pigeon-holed, and subsequently dealt with, in a monolithic fashion. For the purposes of discussion these sites can be characterized under two general headings. The first includes lithic scatters which exist in the most literal or classic sense. Such sites are not extensive but occur as relatively isolated, discrete loci of moderate to low artifact density. These characteristics imply that occupation occurred only once or at least a minimal number times, more or less precisely in the same location. In their purest form they would 19 represent a single occupation/component. Certainly though in this category the sites will vary with respect to artifact density, size, setting, and number of components.

Another type of site often characterized as a lithic scatter is more extensive. Such sites can cover relatively large areas but still consist of only low to moderate density scatters of primarily lithic artifacts. An essential characteristic of this group is that they actually represent accretional deposits composed of overlapping and/or contiguous, smaller loci of the type just described. At first glance they give the impression of more permanence or larger group size, when in most cases they are more the result of frequent reoccupation, often for millennia. This kind of site is most common in places which seem to have had "favored location" status, examples of which include locations near major streams, springs, or raw material sources. CG-3 fits well into this scenario. It should be added, however, and it is often difficult to discern, that some such sites may, indeed, be "aggregation" sites at which the various loci represent activity/occupation areas of distinct social units which occupied the site at the same time.

An issue that has arisen in the context of lithic scatter studies is assemblage composition, namely whether it is variable from one of these small sites to another, perhaps in different settings or time periods. Focused studies of this issue are not common but one, again the Carolina Piedmont, can be cited as representative of the conventional wisdom. Sassaman (1983) found that Middle and Late Archaic assemblages at South Carolina lithic scatters were essentially redundant. In other words, the range of artifacts represented at these sites showed no significant differences regardless of setting as measured by distance to water, elevation, or landform. This is not to say that the site locations are completely random, but rather that choice of settings is widespread and redundant, especially where the topography and environment is homogenous in character.

Intra-site structure has been documented at some small sites, but primarily this has been restricted to the discovery of "structure" in the form of overlapping or contiguous loci at more extensive sites. This was the case at Site 44PW308 where one locus of an extensive ridgeline site was isolated (Blanton and Robinson 1990). The significant, early investigation at Windy Ridge also achieved similar results (House and Wogaman 1978). Controlled surface collections have also yielded information supportive of the accretional structure of some sites (Goodyear, House, and Ackerly 1979:77-88). The net effect of these studies is that the presence of gross structure has been confirmed at multicomponent, reoccupied sites, and that relatively discrete loci are present.

Finally, the prevailing opinion is that most of these sites are Archaic in origin. While this has often been borne out by extensive surveys it is also clear that later, Woodland scatters also exist. These are more prevalent prior to the Late Woodland in most areas, but are known for all periods. These later small sites have received even less attention than their Archaic counterparts and little about them is reported other than intersite, distributional data.

Ethnoarchaeological Contributions to Lithic Scatter Studies

Much of the inspiration for the study of lithic scatter sites in recent decades derives from results of ethnoarchaeological research among modem hunter-gatherers such as the Australian 20 Aborigines, Kalahari Bushmen, and Inuit Eskimo. By observing these groups researchers have recorded the formation and/or use of small, ephemeral sites and regularly compared them to archaeological lithic scatter sites. The dynamics of these behaviors and processes have been examined and described for an archaeological audience with the goal of improving prehistoric hunter-gatherer studies.

These contributions will be reviewed in the remaining portion of this section. Before beginning, however, some qualifying comments are in order. It has been observed more than once that ethnoarchaeological studies all too often generate "cautionary tales" stressing the limitations of archaeological methods and data. These concerns generally are linked to the complex interrelationships between material culture and behavior. At the same time positive results have emerged, but often as only general or simple linkages between behavior and material culture. Another shortcoming is the fact that recently extant hunter- gatherer groups were restricted largely to marginal or extreme environments. Consequently it has been cautioned that argument by analogy with regard to prehistoric contexts must be limited and that both cultural and environmental contexts be seriously considered. In spite of these limitations the results have ultimately proven useful and continue to enhance archaeological interpretations of small sites; it will be in this vein that the results of these studies are applied here.

Among the earlier studies of this nature is that by Yellen (1977) with the ¡Kung in southern Africa. To describe intra-camp patterning he developed a model which segregates a camp into two concentric rings. The Inner Ring is described as the focus of virtually all communal activities, and includes shelters, hearths, and their related scatters of debris. A direct correlation is proposed between the area of the inner ring and group size. The Outer Ring lies immediately beyond the inner ring and encompasses the space devoted to more specialized and ephemeral activities such as butchering and hide dressing. A like correlation is proposed between the extent of the outer ring and duration of occupation. Within each of the rings can occur smaller, fairly discrete "limited areas of scatter," and these can be subdivided into either "nuclear area scatters" which are linked to family unit activities near a hearth/hut, versus "special activity areas" elsewhere. The former are "richer" or exhibit greater diversity of materials than the latter, owing to their repeated use and status as a focal point of many daily activities.

Other of Yellen's more specific observations are important to consider. Of primary importance is the fact that the smallest camps tend to be less regular in their arrangement and often conform to natural features such as vegetation. Sites in this category are the most temporary and ephemeral and should have their correlates in the archaeological record. His observation that the hearth areas are the focus of repeated and diverse activities is significant. Among these are manufacturing as well as cooking and socializing. As a result by-products of both subsistence and manufacturing activities are regularly intermixed. Hut areas are not marked by debris scatters since they serve almost exclusively for sleeping. While specialized activities take place beyond the inner circle and their location is not random, their occurrence is often dependent on less concrete factors such as where shade can be found.

Archaeological applications of this model are dependent on a few factors. One is that naturally constrained areas such as rock shelters do no permit such an arrangement of activities. A second is that all or virtually all of a site must be excavated or assessed in order to observe such configurations. Finally, at sites (groups) surpassing a certain size the pattern may break down as alternative arrangements are adopted. Otherwise application of the model may have some utility, namely for the interpretation of site structure. It remains to be established, however, in a variety of settings and environments, whether such patterning is present or observable. In this respect, intensive studies should make this a priority issue. Eventually regionally and temporally distinct patterns may emerge with respect to infra-site structure, even to the point that such patterns will be as diagnostic as formal artifacts can be. Also potentially discernible with sufficient samples will be group size, duration of occupation, and site function.

Both Gould (1980) and Binford (1983) have conducted ethnoarchaeological research among Australian Aborigine groups in desert regions of western and central Australia. Gould has elucidated important patterns related to the procurement and use of lithic materials, but these relate more to inter-site differences as opposed to infra-site patterning and are pertinent to all sites (Gould 1978, 1980; Gould and Saggers 1985). More relevant to this discussion, he does recognize two basic categories of sites and contrasts their structure and assemblage composition (Gould 1980).

Habitation-base camps are those sites which are occupied by the largest group sizes comprised of all ages and sexes. They occur in locations continuously reoccupied due to their proximity to water and other significant resources. Such sites are more extensive and the array of activities, and consequently discards, is very diverse. Internally they are often comprised of distinct occupation and activity areas but he is quick to note that these become quickly obscured by the effects of intensive and repeated use. Only the largest features and the latest activities are potentially discernible in the general scatter. Specialized activity sites on the other hand are much smaller and retain more integrity. They serve as the loci of specific resource extraction such as quarries, as hunting blinds, or as overnight camps. Most often they are occupied by only one age and sex group and for seldom longer ,than one overnight. Consequently the residues which remain tend to stay where they were originally left and retain high potential for ethnoarchaeological or archaeological interpretation.

In practice, this simple dichotomy of settlement types is a mainstay of current hunter- gatherer settlement studies. What is lacking, however, are better models of intra-site organization. Clearly the smallest, most ephemeral sites stand to yield more readily definable patterns. At the larger, reoccupied sites the potential for confident interpretation is linked directly to the intensity of reuse. In less constrained, temperate environments the degree of reoccupation may be relatively less intensive and, thus, more conducive to pattern analysis. A potential avenue for study, based on the assumption that the equivalent of these smaller sites are the basic building blocks or components of larger camps or reoccupied sites, is to focus on the structure of small, well-preserved scatters in order to elucidate or infer the basic structure of larger, more complex sites of the same period.

Lewis Binford has based much of his model-building on the results of ethnoarchaeological observations. While he, too, has focused more often on inter-site, settlement issues, some of his infra-site studies are important to this discussion. To begin, Binford (1983:139-149) refines the understanding of internal site organization beyond Yellen by reporting other distinctive but culturally-dictated arrangements. He notes that in contrast to the ring pattern, semicircular, linear, and dispersed arrangements of huts or minimal group spaces are reported within individual camps. It is further noted that in the most dispersed camps, the individual scatters may wrongly be defined as separate, non-contemporary sites in an archaeological context.

At a more refined level, Binford also acknowledges that hearths are the focus of an array of activities. Working with Eskimo hunters, he observed two general but distinctive areas of discard around a hearth. Closest to the hearth in the seating area is what he terms a "drop zone." It is here that the smallest, usually incidental by-products of activities fall and are left. Just beyond the drop zone, further from the hearth, is a "toss zone." Larger pieces of debris are intentionally removed or literally tossed into this perimeter as a form of hearth area maintenance. Interspersed in these zones, but more often beyond the drop zone, can be small "dumps" where homogenous collections of debris from specific activities are discarded. Examples would include debitage from stone working, bone scraps from butchering, or even broth from cooking. Once established, such areas can become a "magnet" for subsequent dumpings, eventually creating a larger, mixed disposal area.

The extent and complexity of these patterns are clearly contingent on group size and the length of occupation. Other complicating factors include the bulk of discards and weather. Activities resulting in large quantities of debris will be carried out farther from the central activity area. These decisions are also based on the length of the occupation, relating to whether the debris will eventually interfere with other activities later. In the same way that Yellen makes note of shade as a factor contributing to selection of activity areas, Binford observes that wind can effect the ultimate configuration of a site. Shifting winds can cause hearths and activity areas to be moved for continued comfort.

Binford concludes that many hunting and gathering groups organize their space similarly within sites, most often in a circular or semi-circular pattern around hearth areas. Other patterns are recorded but occur less universally. The regularities he documents are encouraging for archaeological evaluations of small lithic scatters. While the pitfalls of direct analogy must be consciously avoided, his models are useful starting points for interpretation.

Proposed Research Approaches for the Treatment of Lithic Scatters

At this point the potential for future lithic scatter studies can be outlined. Progress has been made but it will be argued that their research potential is far from exhausted and serious consideration, combined with creative approaches to study, can yield significant results. These suggestions will not include a discussion of such sites from the standpoint of settlement studies. This does not imply that this area warrants no further inquiry, but simply that 23 this topic has been more typical of lithic scatter treatments to date and that more pressing is the need to assess the sites individually. In so doing a more accurate appraisal of the function, duration of occupation, and variability of scatters can be attained.

Having documented that intra-site structure exists at least on a gross scale in the form of loci, a crucial next step is to investigate structure within individual loci. The principle goals of this would be to examine the organization of activities and patterns of discard within locus boundaries. Based on distributions of artifacts, including refitting/conjoining results, it may be possible to infer where activities did and did not take place and in turn estimate group size and duration of occupation. Features in the traditional sense should not necessarily be expected but areas of relative artifact concentration can be treated similarly. The resulting distributional information can potentially be interpreted in terms Binford's (1983) toss and drop zone, and other ethnoarchaeologically-derived models.

Not all lithic scatters will be conducive to focused study.Archaeologically the sites of this type most likely to yield significant information are 1) those that retain sufficient integrity as indicated by spatial patterning, and 2) those that represent single or a minimal number of occupations. In the first instance, well preserved examples are rarely located. The ephemeral nature of many such sites makes them the most subject to damage from any number of sources. Severe impacts commonly result from plowing/land clearing/logging and subsequent erosion. This is, a problem particularly in the Piedmont where shallow soils and considerable relief result in many deflated and mixed deposits. Similar conditions also occur in sections of other physiographic provinces, including the Coastal Plain. An unfortunate reality is that many locations where better-preserved scatters can be expected are in marginal settings, which are not necessarily representative of an area at large. For instance site locations often coincide with the areas of extensive arable soils. Since these areas have the longest history of clearing and cultivation, however, they cannot be expected to contain well preserved sites. Instead, sites with integrity are more apt to occur where favored, arable soils are limited in extent, such as in poorly drained sections of the Coastal Plain or on restricted landforms in the Piedmont.

Pattern analysis will be most productive at sites representing single or very few episodes of occupation. Otherwise, the clutter of overlapping occupations is almost impossible to dissect and interpret. Again, reoccupied sites are common and are not random occurrences. Certain localities were favored for various reasons which we should attempt to identify. A goal should be to seek sites of this type with minimal reoccupations, or where the component loci are spaced horizontally such that overlap between them is minimized. In these cases there remains potential for fruitful analysis.

Ultimately, an appreciation for the nature of lithic scatters will depend on a carefully selected sample of sites investigated in similar fashion. It is safe to say that sites of this nature are routinely "written off' as insignificant or as having no research potential beyond a record of their location for the purposes of settlement analysis. Through this and other studies it is hoped that this view can be challenged and the potential of well preserved lithic scatter sites demonstrated. A key element in this program will be to examine several sites to allow for comparisons.

Minimally the sample should be comprised of at least one site from a representative range of settings within each of the major physiographic provinces. For instance, in the various provinces lithic scatters from major floodplains, along interior drainages, and in upland settings should be excavated. Perhaps these categories could be further refined to include examples representing different time periods. In addition, while it is true that the best preserved examples should be given priority, it would be instructive to also investigate others which have been plowed or otherwise disturbed to assess the negative effects or limitations they have.

Sites of this class with the potential to yield significant information can only do so with intensive investigation. Standard survey or testing methods consisting of interval testing using shovel tests and/or test units, or grab samples in surface contexts, will serve only to identify these sites. In the case of some of the smallest examples, only one or two positive tests may mark their existence. Therefore it is imperative to open a series of contiguous units, or conduct controlled surface collections, within the site area to discern patterns and accurately define boundaries. Only under an intensive excavation regime can subtle but meaningful variations in artifact density and distribution be recorded. Certainly analyses involving refitting and density plots will benefit from or be impossible without this approach.

Recovered data should be subjected to careful spatial analysis. Both the limited archaeological results available and ethnoarchaeological studies indicate that internal structure, manifest as variations in artifact density and distributions, should be present at these sites. With the benefit of intensive sampling results, recovered material can be plotted along a multitude of dimensions in the first step of pattern analysis. Different gross classes of material can first be plotted for clues to the location of activity or discard areas. When such areas are indicated, more specific categories of material can be plotted to evaluate the composition and, thus, the origin and significance of a concentration. It will be from these results that such questions as group size, duration of occupation, and site function can be addressed.

Spatial analyses can be enhanced by refitting or conjoining studies. Some of the most irrefutable support for pattern analysis can be derived from the refitting or conjoining of artifacts. The number and relationship of refits can act as a strong measure of the complexity of a deposit or assemblage. Under the most ideal circumstances it will be possible to estimate such things as the minimum number of bifaces, cores or other tools represented. This information can be applied in inferring group size and duration of occupation. Likewise, the interrelationships of crossmends and will make interpretations of activity areas and the number of occupations/components more realistic. For an example of the application of this method, articles by Cahen, Keeley, and Van Noten (1979) and Cahen and Keeley (1980) should be consulted.

The contribution of well-designed lithics analysis in this endeavor can be significant, owing in large measure to the fact that lithic artifacts comprise all or the better part of most assemblages. From this perspective it actually is imperative to maximize the interpretive potential of lithic material, but at the same time to integrate its study with that of other classes of information. Guiding this work are a set of basic research questions that concern 25 problems of intra-site organization and function, as well as inter-site or regional settlement dynamics.

Based on the results of both archaeological and ethnoarchaeological studies certain current understandings and expectations about lithic assemblages from small sites can be outlined. At the most general level these are that:

• 1. Late stage reduction and maintenance will be indicated; key indicators are rare cortical debitage and an abundance of thinning and retouch flakes.
• 2. A limited array of tools will be present due to the specialized nature of the sites.
• 3. Artifact densities will be relatively low due to the short occupation span and small group size.
• 4. The ratio of debitage:tools will be low since tool production was not common. Cores and early stage bifaces will also be rare for the same reason.
• 5. A large proportion of the tools recovered will be either expedient or broken since still-functional items will be curated and removed for use at other locations.

Some component-specific expectations are that:

• 1.The degree of variability between assemblages at different sites of the same period may change through time and indicate different settlement/subsistence strategies. In other words, in periods marked by logistical subsistence strategies (such as Paleo-Indian) or semi-sedentary settlement (such as Late Woodland) the assemblages at different sites are potentially more variable (i.e., between base camps, seasonal camps, and procurement sites). Inter-site/assemblage distinctions are minimized in foraging systems (such as in the Middle Archaic).
• 2. A corollary of the preceding statement is that as subsistence/settlement strategies vary through time, so does technological organization in general. In basic terms, curated/standardized/formalized technologies and toolkits are more characteristic of logistically organized systems. Expedient/generalized technologies are more often associated with foraging systems. These examples are restricted in large measure to hunter-gatherer adaptations, however. Extending the argument to more sedentary systems, research has indicated that in general lithic technology is very unstructured and expedient, although some specialized but limited production of status/exchange items can occur.

To test and further refine these expected patterns, a series of basic objectives and approaches is outlined below. These are designed to be applicable to a variety of sites in a variety of settings.

Among the more basic issues is patterns of lithic raw material procurement and use. In other words, what materials were utilized, what can their distributions tell us, how did this change through time, and how is the factor of availability manifested? Baseline information about the "lithic landscape" of the area must be compiled at the outset. This will involve consultation with area geologists, -field checks of sources, and a literature review. As the analysis proceeds, the raw material represented will be recorded for all classes of material. 26 This will reveal what the favored raw materials were. It can also evidence the "mobility scale" and/or patterns of inter-group contact. With these complementary sets of information the questions posed about procurement and use patterns can be addressed.

Another fundamental issue concerns the temporal dimension or cultural history. It is important to determine the period(s) of site use vis-a-vis current taxonomies and sequences, but also to refine the regional temporal sequence if possible. All formal artifacts will be typed according to current descriptions and metric attributes will be recorded.

A related but less common feature of chronological studies is identification of temporal signatures among other classes of artifacts such as debitage. In short, it may be possible to identify attributes of reduction strategies that are no less temporally sensitive than so- called "diagnostic" items. For instance, the frequency of bipolar, structured biface, expedient, and other reduction strategies can vary through time and be positively linked to certain periods.

Where patterns of reduction are observed that do not have temporal significance, they should be evaluated for functional significance instead. The same patterns of reduction cited above are also linked as well to different patterns of adaptation and settlement. For instance, the degree of mobility and/or the type of subsistence organization can be manifest by specific reduction strategies.

Lithic artifacts should be considered in studies of site structure. Distributions of different classes of lithic artifacts within a site can be important indicators of activity and discard areas and, perhaps, group size or composition. Such should include a refitting study to gauge the contemporaneity and integration of activity areas. Both vertical and horizontal distributions can provide clues to formation processes as well.

Every effort should be made to characterize the technology represented. This can entail evaluation of core and biface reduction strategies, what stages are represented, how they vary through time, and what their implications are for inferring site function. Of special interest is how the lithic reduction-maintenance continuum is manifest at these small, interior sites. Did production or early stage reduction occur elsewhere or did this change through time? Are these patterns consistent from site to site and among different raw materials? The patterns identified should provide indirect measures of duration of occupation and patterns of movement.

The functions of the artifacts recovered and what they reveal about site function is an important consideration. Without use-wear studies the function of different tools must be inferred from tool form. This usually provides only general information. However, and time permitting, breakage patterns, gross edge morphology, and macroscopic damage can be examined to refine the evaluations. A product of this analysis will be to produce ratios of different tool types represented at different sites and components. These will provide standard comparative indices from which site function can be assessed.

Ultimately a goal of the study will be to characterize the toolkits represented at these sites and, of course, what they reveal about site function, group mobility, and subsistence activities. Here again, ratios of different tool types represented at different sites will be the basis for this assessment. An important feature of this and other aspects of the analysis will 27 be to compare the 44JC633 findings with those from other areas and sites, but especially larger estuarine/riverine sites that presumably represent macro-social aggregation or base camp locations.

Field Methods

The purpose of the field investigation of site 44JC633 was to assess the importance of such areas of Native American habitation in the interpretation of the prehistoric archaeological record in the Coastal Plain of Virginia and its relationship to regional site models. The approach used in the field encouraged a complete and comprehensive recovery of artifacts. Laboratory analysis and an extensive review of the archaeological literature complemented the evaluation process.

The Department of Archaeological Research uses an excavation strategy called "openarea" excavation, where large horizontal areas are cleared and dug without intervening balks of standing soil. Profiles of soil stratigraphy are drawn of the most representative face and later compiled into a composite section drawing, while plan drawings are made of fully-exposed features. In a slight modification of this technique, excavations at 44JC633 were performed in contiguous one-by-one meter units; in a few cases isolated one-by-one meter units were placed in other areas.

Horizontal position was noted in reference to a systematic coordinate grid established from a north-south baseline previously marked during the excavation of adjacent site 44JC647 and was recorded vertically by shooting the top and bottom of the context with a laser theodolite. For recording and laboratory purposes, each excavated layer or feature, called a context, was given a unique context number and recorded using a standardized context record form (Figure 7).

Prior to mechanical stripping of the plowzone, seven arbitrary 1 x 1 m units were excavated. Five were completely excavated to subsoil level with the remaining two units only partially excavated. Although placement of each unit was initially arbitrary, they were within an area defined during Phase II testing.

In an attempt to test artifact distribution in the plowzone, determine the extent of the soil layer underlying the plowzone, and to locate certain artifactual "hotspots," fifty-three shovel tests were randomly placed within the suggested site parameters (Figure 8). Soils from both the 1 x 1 m units and the shovel tests were screened through ¼" and 1/8" mesh.¹⁰ Soil samples were taken from each unit for flotation and/or chemical analysis (see Appendix B). All elevations were recorded from a single transit station.

After mechanically stripping an 898 square meter area with a Drott 40 backhoe, the site was shovel-shaved to expose the intact layer of soil. The grid established during the initial testing of the area was then expanded over the entire stripped portion, dividing the complete site into 1 x 1 m units. Units were excavated using hand trowels and all artifacts were recovered by context. Each artifact was first "piece-plotted" using the laser theodolite, 30 FIGURE 7
Standardized Archaeological Context Record Form. FIGURE 8
Plowzone Shovel Test Distribution. 31 which recorded the horizontal and vertical position of the find. A plastic artifact bag was labelled with the "tag number" corresponding to the data recorded in the theodolite. This data was later integrated into the database containing the artifact and its other attributes. Artifactual and contextual information was put into GEOBASE format.¹¹

Once each unit was excavated to subsoil level, profile drawings were made and all cultural details were recorded.¹² No photographs or slides were made of any part of 44JC633.

Soil samples for water-screening were taken from several units. Each stratum of those units sampled were arbitrarily divided and labelled Layer A, Layer B, etc. The volume of the samples was 100% of both the horizontal plane and vertical dimension of each layer.

Both the initial testing and extended phase of excavation focused on establishing site limits, structure, integrity and cultural affiliation, as well as recovery of as much artifactual material as possible. The most minute lithic reduction fragment is important in the interpretation of site function. Therefore, an attempt was made to make as complete of a recovery of artifacts as possible.

Laboratory Methods

All recovered artifacts were bagged, labelled and turned in to the laboratory of the Department of Archaeological Research. There, the artifacts were washed, numbered, identified, and catalogued. An artifact inventory was compiled using a standard descriptive typology for prehistoric artifacts (Blanton 1991).

Patterns of manufacturing technology, temporal frames of the artifact assemblage, raw material types and cortex percentages were the basic foci of analysis. All debitage, tools and miscellaneous lithic artifacts were categorized based on standard guidelines for lithic artifact description and identification. Individual blocks of lithic material from the debitage was also identified.¹³ From this type of assemblage, estimates can be made of the minimum numbers of tools manufactured and the technology used to make them. To accomplish this, all lithic artifacts were openly placed on a countertop in the laboratory and then sorted into distinct groups of material. Refitting or conjoining of lithic material was attempted. ¹⁴

Introduction

The basis for the evaluation process of- Site 44JC633 derives from the research goals and field methods previously described. Extensive laboratory analysis was made possible from the systematic and comprehensive excavation and recovery of lithic artifacts from the site.

Results of Fieldwork

Results of Artifact Analysis

Lithic materials recovered from 44JC633 were extensively examined in the laboratory. As an initial step in sorting materials, a "lithic landscape" was recreated by arranging all stone artifacts in a grid pattern - the same as was used for the excavation. As this arrangement proceeded, it was obvious that a minimal number of original blocks of raw material were present at the site, and that sorting of debitage into their original constituent blocks was possible. Additionally, the possibility of refitting a number of these pieces became evident.

In Locus-1, six quartzite blocks and one quartz block were defined (see Table 1). Eleven quartzite blocks and one quartz block were defined in Locus-2 (see Table 3).

These should be viewed as a minimum number of blocks. It should be noted that because the examination and sorting of the debitage was strictly visual, some error and misclassification may exist. Natural variation within the blocks, as well as the fact that some had been altered by heat, made classifications at times difficult. Micro-debitage and those pieces that could not be associated with any particular block were not considered in the individual block analysis.

Data initially gathered on the individual blocks concerned quantities, weights, refitting within blocks, and the presence of cortex on individual pieces of debitage. In consideration of these factors over the horizontal scale of site distribution, the lithic materials were examined to detect patterns of site functions and activities.

In any quantitative study, the size of the sample is always an important factor. Therefore, all blocks in both loci were used as the focus of this study. Blocks of debitage in Locus-1 and Locus-2 account for 23.8% (n=334) and 76.2% (n=1072) of the debitage recovered from the site respectively.

FIGURE 9
Tree Hole Features.

FIGURE 10
Initial Test Units for CG-3.

38

Locus-1:

The relative frequencies of the six quartzite blocks and the one quartz block are represented in excavation units on a line graph (Figure 11). In terms of relative highs and lows, each block exhibits a close correlation in patterning. This close proximity of lithic block distribution could suggest a single occupational episode.

The primary depositional focus at Locus-1 is quite clear in Figure 11. The prominent peaks are those areas where lithic materials from most or all of the blocks were deposited. Units 401, 402, 403, 600, 601, and 602 are contiguous units that correspond with the most densely saturated areas in the overall distribution of debitage.

Regardless of the uniformity of structure and composition of the block debitage distribution patterns and the few tree hole disturbances previously described, site structure could still be detected. Figure 11 also illustrates some important patterns of distributional variation. Blocks D and E have peaks at Unit 603. Block B, in addition to peaks noted in other units, peaks again in Unit 604. Block A has slightly prominent peaks along the same plane in Units 618 and 619. Unit 619 also shows distinct concentrations of Blocks B and C. Figures 12, 13, 14, 15, 16, and 17 illustrate these differences as well as the patterns of deposition for individual blocks.

The raw quantities of debitage in each unit are illustrated in Figures 12 and 13. The high density areas for each block (except Block E), are generally near the south-southwest corner of Locus-1, surrounded by an almost circular pattern of diminishing artifact concentrations. No "dead zone" seems to exist, with the exception of Block E. Variance in size and location of the high density focus is minimal. This variance, more specifically, is characterized per block in Table 2.

Four general areas of focused deposition are described by these quantitative distributions. These overlapping areas suggest that lithic reduction of Blocks A, B, C, D, and F

TABLE 1.
Block Classifications By Unit Locus-1

Unit	A	B	C	D	E	F	G [Quartz]
222	5	4	3	7	0	1	1
401	19	17	18	15	0	0	1
402	21	4	1	16	1	5	0
403	9	4	6	8	0	1	0
404	2	1	0	0	0	1	0
405	2	2	0	0	2	0	0
600	8	11	5	12	0	3	0
601	5	6	9	9	0	1	0
602	13	9	4	8	0	0	1
603	2	2	0	6	3	0	0
604	2	6	0	1	0	0	0
618	5	2	3	3	0	2	1
619	5	2	2	3	0	0	1
620	0	0	0	1	0	0	0
621	1	0	0	0	0	0	0
Totals	99	70	51	89	6	14	5

39 FIGURE 11
Block Debitage Distribution FIGURE 12
Distribution of Quartzite
Blocks A, B, C, D, E, and F by Quantity
(Locus-1). 40 FIGURE 13
Distribution of Quartz Block G
and Fire-Cracked Rock (FCR) by Quantity
(Locus-1). FIGURE 14
Distribution of Quartzite Blocks
A, B, C, D, E, and F by Mean Weight Per Flake
(Locus-1). 41 FIGURE 15
Distribution of Quartz Block G
by Mean Weight Per Flake
(Locus-1). FIGURE 16 Distribution of Quartzite Blocks A, B, C, D, E, and F by Weight (Locus-1). 42 FIGURE 17
Distribution of Quartz Block G by Weight
(Locus-1).

TABLE 2.
Areas of Depositional Focus for Individual Blocks
(Locus-1)

	SIZE OF FOCUS		AREA OF FOCUS
BLOCK	SMALL	WIDE	SW	SC	CENTER	EC	NWC
A		x	x
B	x		x
C	x		x
D		x	x
E		x				x
F	x			x
Quartz		x					x

SW =	Southwest central portion of the site
SC =	South central portion of the site
CENTER =	Central portion of the site
EC =	East central portion of the site
NWC =	Northwest central portion of the site

43 occurred at four locations in or near the southwest corner of the site. The highest artifact_ concentrations occurred in the southwest area with a thinner scattering around the seating area. Block E shows a small concentration in the east central portion of the site and Block G (Quartz) debitage is concentrated in the northwest central portion.

Figure 13, the distribution of fire-cracked rock (FCR), corroborates the established central focus of the site. Twenty-one to thirty-one pieces of FCR were concentrated in the southwest corner of Locus-l, with a fanning out of FCR in lighter concentrations. It is interesting to note that higher concentrations occur in the northeast corner as well.

Implications and Significance of the Results

Technological Patterns

Another important part of site function, in addition to site structure, is the character of the reduction sequences for each lithic block.

Summary

It is doubtful that the primary function of this site was lithic reduction for tool-making. The raw material recovered as debitage, fire-cracked rock, and miscellaneous unmodified stone is quite informative. The abundance of large intermediate stage debitage instead of smaller, late-stage maintenance debris was somewhat unexpected when considering "typical" lithic assemblages from similar sites. The absence of stone tools coupled with the predominance of intermediate stage debitage lends to the interpretation that other activities were the primary reason for occupation of this site. The debitage reflects only the by- products of intermediate stage reduction activities incidental to other activities during the period of time the site was inhabited. What activities were the focus of the stay is virtually impossible to determine. Many scenarios can be developed, such as:

A small hunting party established a campsite. Since the adjacent ravine yields quartz and quartzite cobbles, some were collected for toolmaking during time when other activities have halted.²⁰ The party had planned to join another small group expected to arrive within a day or two. Inclement weather prevented a timely meeting, causing the first party to remain at the site longer than expected. More cobbles were collected to occupy the individuals during their wait.

Large secondary thinning flakes and other debitage with limited cortex implies the blocks of material were initially reduced elsewhere and then carried to the site as either flake blanks or early to mid-stage bifaces. At Locus-1, of the quartzite debitage recovered, 95.7% (n=1593) was non-cortical. Quartz yielded 90.9% non-cortical debitage (n=20). Locus-2 is similar, in that 94.1%, of the quartzite was non-cortical debitage and 90.9% of the quartz debris was non-cortical. It also appears that any tools started at this site were not completed here. The quantity of late stage debitage was minimal as were discarded or broken tools.

Based on cross-mends and debitage size, large, hafted biface preforms were probably the products of the reduction processes at CG-3. This would be characteristic of Late Archaic stone tool technology. Thirteen hafted bifaces were recovered during Phase II testing (n=7) and Phase III excavation (n=6). Of the six recovered during Phase III, 3 56 were located within the boundaries of Locus-2 (Contexts 777, 778, and 801) in an intact layer; one in the plowzone directly above the southwest corner of Locus-2 (Context 330); one between Locus-1 and Locus-2 (Context 609); and one completely outside of the site boundaries (Context 535):²¹

Since the design of this analysis focused more on site structure and function, only a few aspects of the lithic material (distribution, block type, and mean weight per flake). A more intensive examination of the artifacts recovered from CG-3 would be required to answer questions about behavior. The future research potential is considerable.

Image not available

FIGURE 25
Hafted Bifaces Recovered from CG-3/Phase III.