Lithic (stone artifacts and waste products) analysis in the southeastern United States tends to focus on descriptions of form, distribution, and raw material. These limited observations lead to considerations of simple patterns related to raw material procurement, manufacture techniques/stages, and mobility. Issues rarely addressed include bias in recovery methods, maintainability or reliability of tool design, and craft specialization. In this chapter, we take ideas and inferences proposed by Sam Brookes as inspiration when considering the lithic assemblage from the John Forrest site (22CB623) in Claiborne County, Mississippi. Brookes (1997, 1999b, 2004; Johnson and Brookes 1989) has conducted extensive research into Middle Archaic-period materials of the western Southeast and has made a case for craft specialization during this time based in part on the production of beads (see also Brookes, this volume; Crawford, this volume). Other researchers differ in interpretations of evidence regarding, for example, the presence/absence of craft specialists and/or characterization of specialists. For the John Forrest assemblage, we use an organization of technology model to investigate the production of lapidary items. Expansion of this model from previous applications is accomplished through a consideration of craft specialization as a socioeconomic strategy with the addition of ideology. Finally, conclusions are drawn regarding whether specialists crafted chert beads and other stone artifacts during the Middle Archaic period.
The John Forrest site dates to the Middle Archaic period based on diagnostic hafted bifaces. Interest in this time period has dramatically increased over the last few decades, especially as evidenced by coverage in edited volumes (e.g., Anderson and Sassaman 1996; Emerson et al. 2009; Gibson and Carr 2004). In the Lower Mississippi Valley, or LMV, this interest was sparked by the discovery that many of the earthen mounds predate the Late Archaic period Poverty Point site (16WC5), once believed to represent the earliest intentional mound construction in North America (Russo 1994; J. Saunders et al. 1994, 2005; R. Saunders 1994). Additionally, studies of rare artifact types, including stone beads at both mound and non-mound Middle Archaic sites (Crawford 2003, this volume; Hadley 2003; Johnson 2000; McGahey 2005), oversize bifaces (Brookes 2004; Johnson and Brookes 1989), fired earthen objects (J. Saunders et al. 1998; see also Hays et al., this volume), and well-made ground stone objects (Johnson 2000) persist. These investigations have generated an increasing number of hypotheses regarding the lifeways of Middle Archaic peoples. For example, while Joe Saunders (2004) sees no evidence for social inequality at three Middle Archaic mound sites other than the mounds themselves, others interpret the data to indicate the formation of unilineal kin groups (Widmer 2004) and tribes (Anderson 2004). Clearly, economic, social, and/or religious changes were occurring in the LMV during this time relative to what is known about the archaeology of the Early Archaic period. Craft specialization has long been considered by some researchers (e.g., Clark and Parry 1990; Price and Brown 1985) to have a close correlation with increasing cultural complexity, and we suggest that chert bead manufacture has economic, social, and religious implications.
Research in Mississippi and Louisiana has revealed Middle Archaic sites with assemblages generally similar to John Forrest with respect to beads, and these assemblages are influential in shaping the current debate regarding craft specialists. Loosa Yokena (22WR691), Watson Brake (16OU175), and the Keenan Cache are three such examples (Figure 6.1). In fact, McGahey (2005:16) describes the John Forrest lithic assemblage “as virtually identical to Loosa Yokena as far as the lapidary industry is concerned and [it] in fact may represent the activities of a single group of lapidarians who moved from one of the sites to the other, perhaps because of a river channel change.” Loosa Yokena is a non-mound site in Warren County, Mississippi. Watson Brake, in Louisiana, is a mound site where analysis of the chipped chert beads resulted in defined production stages (Johnson 2000) used for the analysis of the John Forrest assemblage. The Keenan Cache is a collection of 469 stone objects found by a farmer in Jefferson Davis County, Mississippi, in 1876 (Connaway 1981; Crawford, this volume). The collection contains an array of bead preforms and tubular and zoomorphic beads comparable to those from Middle Archaic contexts. Previous considerations of chert bead assemblages (Connaway 1981; Crawford 2003; Johnson 2000; McGahey 2005), while insightful, have not followed an organization of technology model to examine systematically how this technology was organized.
Figure 6.1. Middle Archaic sites discussed in the text.
Our approach has its roots in the ethnoarchaeological work of Lewis Binford. Viewing material culture within a living cultural system afforded Binford a unique perspective (e.g., Binford 1962), which eventually led to the development of an organization of technology framework. Binford and Binford (1968:2) observed that, “Since artifacts are cultural data and since they once functioned as elements of a cultural system, many of the explanations we might offer for observations made on the archaeological record will refer to organizational features of past cultural systems.”
Michael Schiffer (1972:157) expanded on this, stating that to use “archaeological data to infer past activities or organization,” archaeologists need a testable framework. Schiffer’s proposed framework was a behavioral system in which the entire “life cycle” of an artifact was considered within its living, systemic context before entering the archaeological context. The multiple processes that make up the systemic context were defined as storage, transport, procurement, manufacture, use, maintenance, recycling, and discard (Schiffer 1972:158). An organization of technology approach incorporates the systemic context or life history framework (Bradley 1975; Collins 1975; Sheets 1975; see discussion by Carr and Bradbury 2011) in an attempt to infer social and economic strategies (Binford 1979; Carr 1994; Nelson 1991).
Margaret Nelson (1991:59) created a diagram depicting levels of analysis that Carr and Bradbury (2011) recently modified as an organization of technology model (Figure 6.2). This new model emerged from application of the approach, consideration of entire lithic assemblages, and efforts to display the interaction between one level and another. As with the original diagram, the head of the model is the environment, followed by demographics, and the base is comprised of artifact form and artifact distribution. As cogently discussed by Nelson (1991), the bottom is situated in the archaeological record and the top is the most distant from that record with regard to making inferences from lithic assemblages. That is, an analyst would proceed through a number of successive inferences from artifact form and artifact distribution before making inferences concerning the environment. Past environments are reconstructed more directly through other specialties with their accompanying datasets, such as geoarchaeology, paleoethnobotany, and zooarchaeology (Carr et al. 2012). Thus, at a minimum a lithic analyst should have knowledge of environmental factors affecting raw material availability. Nonetheless, the impact of the environment on artifact form and distribution is mediated through a number of factors accounted for in the model (see counter argument by Andrefsky [1994] concerning the importance of raw material availability). That is, these are conditions managed through technological strategies when implementing social and economic strategies.
Figure 6.2. Organization of technology model developed by Carr and Bradbury (2011) originally adapted from Nelson (1991).
Until now, the impact of demographic factors on lithic assemblages has received little attention. While we see advantages to determining population size and other parameters through archaeological data other than lithics, demographics also impact the form and distribution of artifacts; hence these factors have a role to play in the organization of technology.
A major advantage of an organization of technology approach is the importance placed on social and economic strategies, and how technological strategy is shown as being responsive to these in a given environment. Such strategies can be highly variable, necessitating flexibility in the model. In previous organization of technology studies, assumed social and economic strategies have been derived from ethnographically documented hunter-gatherer lifeways, especially the reliance on mobility. However, “there is no evident theoretical conflict” preventing applying this approach to modeling a broad range of economic and social strategies (Carr 1994:2). For example, Cobb (2000) combines a political-economic framework and organization of technology to examine social asymmetry, interregional exchange with regard to influence by elites, and gendered labor divisions. Here, our focus is craft specialization and its specific place in the organization of technology model, as explored through an examination of chert bead manufacture in terms of whether Middle Archaic lapidarians were craft specialists. Sam Brookes (2004) and others (Connaway 1981; Hadley and Carr 2010; McGahey 2005) have suggested that craft specialists produced Middle Archaic chert beads. Other researchers disagree that specialization was necessary or is evident during this time (Johnson 2000; J. Saunders 2004). Craft specialization is one strategy with implications for the design of material culture and associated activities; these in turn influence artifact form and artifact distribution. However, defining specialization, identifying craft specialists using archaeological evidence, and incorporating craft specialization into a model of the organization of technology are not simple and straightforward tasks. The John Forrest lithic assemblage and related assemblages provide data relevant to these problems.
The John Forrest site in Claiborne County, southwestern Mississippi, lies between the Mississippi River and the Natchez Trace Parkway. Bordered on the west by a steep slope to the lowlands and to the north by James Creek, the site (an area approximately 420 m north-south by 245 m east-west) is located above the floodplain on Pleistocene uplands; the present Mississippi river channel is approximately three and a half kilometers to the west. The floodplain encompasses abandoned meander belts and active and abandoned river channels; the location of the site above the floodplain has ensured its preservation. We are uncertain of the exact Middle Archaic-period location of the river, but it would have been no more than 19 km to the west of the site, based on the current size of the river valley (Aslan and Autin 1999). The soil at the John Forrest site is a Memphis silt loam (0–2 percent slope) (Lane and Cole 1963) developed on a Pleistocene-aged loess (Aslan and Autin 1999:Figure 2, p. 803). The site location allows access to a diverse range of upland and lowland resources, which we assume to have been the situation in prehistory as well. A more detailed understanding of the paleoenvironment in the vicinity of the site would aid in making and testing cultural inferences.
A significant amount of surface collection, but only minimal subsurface testing, has been conducted at the site. Historically, the land was used as a hunting camp for many years, being periodically plowed and planted with deer feed, allowing the landowner to amass a large collection of lithic artifacts (n = 3,411) and identify concentration areas within the site. Sam McGahey, of the Mississippi Department of Archives and History, introduced us to the site, leading to the only formal investigations there in a joint field school between the University of Southern Mississippi and the University of South Alabama. Shovel tests were excavated on a five-meter grid at the southern end of the site where the landowner observed concentrations of beads and bifaces. A total of 534 shovel tests and seven 1-×-1-m units were excavated with the matrix being screened through one-eighth-inch mesh hardware cloth. Artifact recovery was restricted to the plowzone, as no intact deposits were identified at the site.
The large collection of artifacts from the site has allowed us to investigate the general nature of the site assemblage despite the lack of features and datable deposits. Following the field school, the landowner provided us with access to the surface collection for a detailed analysis. The chipped stone assemblage includes 1,283 bifaces, 368 hafted bifaces, 422 cores, 99 chipped beads, and 1,169 pieces of debitage. The hafted bifaces provide the best evidence for the date of the site, with the majority identifiable to culture-historical types associated with the Middle Archaic period. The majority of the identifiable hafted bifaces are Sykes-White Springs (n = 82). Interestingly, 81 percent of the non-hafted bifaces exhibit a highly polished surface visible without the aid of a microscope. Experimentation and microwear analysis, still relatively rare in the Southeast (Franklin et al. 2012), are needed to determine the exact source for the polish. Additionally, the collection has a variety of ground stone artifacts, including 90 modified greenstone pebbles, 4 greenstone pebble beads, 21 abraders, 5 axe fragments, 1 plummet preform, 15 atlatl-weight preforms, 3 hammerstones, 1 nutting stone, and 39 other ground stone objects of unknown function. Shovel test and unit excavations recovered a large chipped stone assemblage (n = 10,971), which, unlike the landowner’s collection, is unbiased by what is visible on the surface.
Figure 6.3. Typical blade cores in the John Forrest surface assemblage. Photo by Sarah Mattics.
These collections have been analyzed using standard methods. Artifacts were initially grouped by techno-functional categories as defined by Andrefsky (1998). A detailed analysis of each artifact was conducted, with specific attributes being defined using a combination of features noted by other researchers (e.g., Andrefsky 1998; Johnson 1979, 2000; McGahey 2000). While the John Forrest assemblage contains an interesting array of lithic artifacts (see Hadley 2003), we focus here on those associated with bead manufacture for comparison with Watson Brake, Loosa Yokena, and the Keenan Cache (Table 6.1).
Debitage totals 1,169 pieces from the surface collection and 10,867 flakes from shovel-test and unit excavations. The surface-collected debitage was not studied in depth due to collector bias; analysis of the surface collection only involved separating blades (n = 397) from nonblades (n = 772). By comparison, 4,173 blades were found at Loosa Yokena (McGahey 2005:Tables 1–11) and 70 at Watson Brake (Johnson 2000:99).
In the surface collection, 322 blade cores were identified (Figure 6.3). Attributes recorded included number of platforms, platform flake scars, complete blade scars, blade remnants, and other flake removals. Platforms were defined as prepared flaked surfaces creating an edge from which blades could be removed. The average number of platform facets was 1.8. Over half of the cores (n = 181) had one platform from which blades were removed in a single direction. Blade cores with removals from opposite directions, forming a bifacial edge, were less common (n = 85). These were not typed as bifaces because the two sides did not form flat, parallel faces and because the overall artifact morphology was roughly cubical or spherical. Conversely, Johnson (2000:96) types bifacial cores as bifaces for the Watson Brake assemblage. Two blade cores at John Forrest had no platform preparation. The remaining 54 blade cores did not fit into these categories and had a range of two to four platforms.
The blade cores from John Forrest are generally similar to those from Loosa Yokena (n = 442) and Watson Brake (n = 16), differences with the latter likely being a result of sample size. The average length of John Forrest cores is 3.8 cm, which is comparable to Watson Brake and Loosa Yokena blade cores (Johnson 2000; McGahey 2005). Those at Loosa Yokena and John Forrest have a range of one to four platforms while all but two blade cores at Watson Brake have a single platform (Johnson 2000; McGahey 2005). At John Forrest, true blade removals were counted separately from possible blade removal remnants and non-blade removals. Because remnant blade scars did not retain the original length, it was impossible to determine if they were technically the dimensions of a blade (i.e., twice as long as wide). However, based on the surrounding scars, it is plausible that these remnants were blade removals. The average number of blades taken from each blade core from the site (based on complete and remnant scars) is 3.8 removals, and the average of non-blade flake removals is 8.6. Johnson (2000:98) notes that the blade cores at Watson Brake were not intensively used based on the low blade scar average (3.6); blade cores apparently were more intensively used at John Forrest. The most probable use for the blades removed from the cores was in the manufacture of microdrills.
Although the excavated materials have only had a preliminary sort, a number of microdrills (n = 40) have been noted, and these are similar in morphology to those reported from Watson Brake and Loosa Yokena. There are 154 microdrills and 93 microdrill preforms from Watson Brake and 159 microdrills from Loosa Yokena (Johnson 2000:99; McGahey 2005:9). McGahey (2005:9) notes that most of the microdrills at the latter site “exhibit parallel blade scars on the dorsal surface indicating that they were struck from the platform.” Johnson (2000:99) was able to demonstrate that 30 percent of the preforms and 5 percent of the used microdrills from Loosa Yokena were certainly struck from blade cores.
Chipped stone beads recovered from the John Forrest Site represent one of the material culture classes that make the collection of particular interest. The 99 artifacts representing the bead production sequence make up less than 5 percent of the surface collection, but this is a significant amount compared to the seven such artifacts recovered from Watson Brake (Johnson 2000:100) and comparable to the 120 listed in tables from Loosa Yokena (McGahey 2005:Tables 1–11). The raw material used in the production of chipped stone beads is Citronelle Gravel, with heat-treating making the majority of the pertinent artifacts a red color. Local chert gravels were also the main source for chipped-stone bead material at both Loosa Yokena (McGahey 2005) and Watson Brake (Johnson 2000). Based on the Watson Brake artifacts, Johnson (2000:100) developed a bead production trajectory, essentially confirmed by McGahey for Loosa Yokena (Figure 6.4), which was employed to classify the John Forrest materials. The first step in bead manufacture was selection of appropriately sized gravels and knapping these into trifaces or quadfaces so that each face is roughly equal in size (Figure 6.5). Over half of the beads at John Forrest (66 percent) were classified as representing this first stage of production. The second stage involved grinding the chipped edges to transform the cross section from triangular/square to round and to smooth the ends. Evidence for grinding was found on 26 stage-two beads, and one specimen was completely ground. The final stage of manufacture was drilling, evidence of which was found on five artifacts. Only one complete chert bead was found during excavations at John Forrest.
Figure 6.4. Stages of chert bead manufacture as seen in the John Forrest surface assemblage. Photo by Sarah Mattics. Originally published in Peacock (2005).
Figure 6.5. Chert bead preforms from the John Forrest surface assemblage. Photo by Sarah Mattics.
Ground (as opposed to knapped and ground) beads are another part of the lapidary industry at the John Forrest site that resembles artifacts at Loosa Yokena (McGahey 2005) and the Keenan Cache (Connaway 1981). Ground beads are generally made from greenstone, a softer material than the chert used for chipped stone beads. Employing the Mohs Hardness Scale, greenstone is softer at a five or six as compared with chert, which has a hardness of seven. For this reason, a different manufacturing strategy is represented, so ground beads were analyzed separately from the chipped stone beads. All four of the ground beads at John Forrest are made from greenstone pebbles, which may be trachyte. In her work with Archaic effigy beads, Jessica Crawford (2003:12) identified greenstone as trachyte from the Ouachita Mountains, but it is also likely found in the gravel bars of the Mississippi River and its tributaries.
The presumed first stage for greenstone bead production was grinding and cutting the pebbles into the desired shape. Ninety greenstone pebbles from John Forrest were ground flat on an edge or had small, incised grooves. These pebbles may represent the first stage of manufacture, but are not clearly recognizable as bead blanks. McGahey (2005:13) described a total of 121 similarly faceted pebbles at Loosa Yokena, which he believed represented the first step in testing the workability and color of the pebbles for bead manufacture. Drilling was the second and final stage of the ground bead manufacture. Three of the greenstone beads at John Forrest are ground into cylinders and one has been completely drilled. A fourth bead is teardrop-shaped with an incompletely drilled, conical-shaped hole at the apex. While greenstone is still a hard material, we assume that the use of chert microdrills would enable relatively quick work. Greenstone bead technology is represented at many sites throughout Mississippi, including Loosa Yokena (McGahey 2005), Denton (Connaway 1977; Crawford 2003), and the Keenan Cache (Connaway 1981).
In sum, although the John Forrest site lacks intact deposits, the assemblage is useful in discussing stone bead manufacture during the Middle Archaic period. The site contains a Middle Archaic component based on the diagnostic hafted bifaces. While in need of additional supporting evidence, McGahey (2005:18) has suggested that the exact same occupants were at the John Forrest and Loosa Yokena sites. Loosa Yokena has a radiocarbon date of 5730 ± 110 B.P. (cal. two-sigma) from a feature containing a chert bead preform (McGahey 2005:6–7). This date is significant because, as McGahey (2005:7) notes, it is the first radiocarbon date from a feature containing a chert bead from this region. The feature and artifacts within the lower buried A horizon at Loosa Yokena date comfortably within the Middle Archaic period, bolstering this temporal assignment for the John Forrest bead assemblage.
Stone beads are one of the more interesting classes of Archaic material culture found in the Southeast (see Crawford, this volume; Sassaman 2010). Beyond the intriguing diversity in raw materials, sizes, and forms, consideration of the manufacturing process of chert beads elicits an appreciation for the efforts of prehistoric people rising to a feeling of awe. That chert microdrills were used to perforate Stage 2 chert beads is well demonstrated by the Keenan Cache, which included a broken microdrill in a partially drilled bead (Connaway 1981:63–67). We liken this process to drilling a hole through a pine tree with a pine 2 x 4 board, but on a scale that only adds to the difficulty.
While chert beads have been illustrated in the American archaeological literature for decades (see Crawford, this volume, and references therein; Sassaman 2010:Figure 4.5), certain factors have worked to limit broad recognition of the presence of this artifact class, such that colleagues are often surprised to hear of Archaic stone beads. One of these factors is the limited range of distribution: Louisiana, Mississippi, and adjacent parts of Alabama and Arkansas (Crawford, this volume). Additionally, stone beads are only one of many interesting items in the material culture inventory of famous sites such as Poverty Point (e.g., Webb 1968). Another issue is that complete beads are relatively rare, and bead preforms and associated stone tools were not always recognized as such in the past. Bead preforms may have been identified in other collections as large drills, drill preforms, or bifacial tools. Due to the small size of microdrills, some form of fine screening is necessary for their recovery, but any form of fine screening is rare on lithic sites in the Southeast (Price 2012). With increased interest in Middle Archaic-period bead manufacture, we hope that more sites with associated evidence are identified via appropriate recovery and analytical techniques.
Although the Archaic period was once thought to be a time when prehistoric people simply became better at food extraction and produced the anomalous Poverty Point site, archaeologists in the Southeast have recently recognized the high degree of cultural complexity that existed at the time (Gibson and Carr 2004). Craft specialization has not always received the same level of discussion as other aspects of cultural complexity, such as mound building, sedentariness, and trade (e.g., Kidder and Sassaman 2009), but ethnographic data have long been used to demonstrate a relationship between cultural complexity and craft specialization (Clark and Parry 1990).
If Middle Archaic lapidary assemblages were an elephant, then researchers, while not blind, have hold of different parts of the animal and are wearing variously tinted glasses when examining their particular part. There are at least three different camps: those who believe there were no specialists; those who believe in specialists; and those who suggest that there were what might be termed “semi-specialists.” Evaluation of the evidence presented by each of these camps is difficult because definitions of craft specialization are rarely provided.
Based on the Watson Brake assemblage, Johnson (2000) provides an excellent description of the stages of chert bead manufacture. He asserts that bifacial reduction of bead preforms and production of microblades is “relatively unsophisticated” and “would require more patience than skill” (Johnson 2000:103). Interestingly, this is despite Johnson’s observation that Stage 2 beads from Watson Brake have encircling striations perpendicular to their long axes, reminiscent of turning on a lathe, and that “bow drills must certainly have been used to drill the beads” (Johnson 2000:100). This is significantly earlier than the initial presence of the bow and arrow in the LMV (Blitz 1988). Yet Johnson (2000:103) states, “There is no indication of craft specialization at Watson Brake, either in the technology of production or in the end products themselves.”
The no-specialization camp is also occupied by Joe Saunders (2004:154), who echoes Johnson’s argument that “the blade/drill technology at Watson Brake did not require a level of specialization” and who is concerned that craft specialization is indicative of “economic support for a specialist by his/her society.” He does describe one phase of occupation at Watson Brake Mound D as a “bead workshop” (J. Saunders 2004), but the intensity of production is viewed as being at such a low level that it could not have provided economic support for the makers. Of course, this assumes that craft specialization means full-time specialists. Joe Saunders (2004:153–154) goes on to reconcile the large count of microdrills with the low count of beads and lack of finished beads by suggesting that microdrills were used in the production of bone or shell beads, or, alternatively, that finished chert beads were being removed from the site. In considering the Frenchman’s Bend Mounds (16OU259) and Plum Creek Archaic sites along with Watson Brake, Saunders (2004:157) sees economic redundancy but variation in the lapidary industry, attributable in part to sample size and recovery methods. He not only concludes that there was an absence of craft specialization, but also that these Middle Archaic people were “affluent hunter-gatherers” with egalitarian social relations, who built mounds (J. Saunders 2004:160)
In contrast, Connaway (1981) and Brookes (2004) argue for lapidary specialists, especially if the Keenan Cache and zoomorphic beads are considered as ideotechnic artifacts. In particular, Connaway (1981:70) views the Keenan Cache as the work of “an individual artisan whose job it was to manufacture them for some specific reason.” He speculates that zoomorphic beads had a socioreligious purpose, perhaps as fetishes (Connaway 1977, 1981). The more recent argument by Brookes (2004) for lapidary specialists is based on a variety of observations and inferences, including ones concerning technology. Using the Keenan Cache as evidence, Brookes (2004:106) suggests:
the cache demonstrates aspects of Middle Archaic technology never before demonstrated. The bead blanks, after being flaked to desired shape, were hafted and turned on a lathe. It is probable that this was a simple lathe, but the technology was there. Also the beads were drilled, often from one end, with a small chert bit from a microblade. Some beads are in excess of 9 cm in length. This suggests that drilling was not done by twirling a bit that was loosely held in the hand. A drill press of sorts, well braced and solid, would be needed for such an operation. It is possible that the bow was used in this process, though it could be that a pump drill, perhaps used with a spindle whorl, was employed in the manufacturing process. At any rate, the manufacture of stone beads in the Middle Archaic was far more complex than previously suspected.
Brookes goes on to suggest that the size of the cache, the time involved in manufacturing and equipment maintenance, and how the technology differs from that used at the Denton site are all indicators that some people were specialists. Additionally, the 80-plus “strikingly similar” zoomorphic beads are taken as evidence that only a few individuals were involved in their manufacture (Brookes 2004:107). This is counter to observations by Webb (1971b:113) specifically regarding “locust beads,” for which he sees variability in size, raw material, and appearance not as representing manufacture at a single site, but as representing the “spread of a basic concept.” Similarly, due to their rarity Blitz (1993:33) does not view locust beads as fetishes, arguing instead that “locust beads may have been emblematic of a corporate group and its descent-sanctioned rights, perhaps functioning as a totemic symbol.” Despite some disagreement over the specifics of manufacture, all four of these researchers would agree that zoomorphic beads fit the description of ideotechnic artifacts. Brookes (2004:108) goes on to say, “Specialists are necessary not only for their skill in flaking or carving but also for their power or magic, which becomes part of the artifact.”
Following from consideration of Loosa Yokena, Denton, and the Keenan Cache, McGahey (2005) argues for some form of specialization, with some individuals having been better at, and more engaged in, the production of beads. Interestingly, he suggests that perforated discoidals found at Loosa Yokena were “flywheels for pump drills” used in chert bead manufacture and that “personal experience indicates that the pump drill is more effective than a bow drill” (McGahey 2005:16). He notes that sites with significant bead-manufacturing evidence are rare, and suggests that zoomorphic beads at Denton were not made on-site based on a lack of manufacturing debris, which further suggests “a certain degree of specialization.” While this is a vague statement, to his credit McGahey (2005:18) provides a definition and more specific characterization of specialization:
Specialist, in considering lapidary work of the Late and Middle Archaic groups of the area under consideration, has been implicitly understood to mean one who works at least almost exclusively at one task and is predominately supported by others while he goes about his work. . . . the individual responsible for the Keenan Cache was relatively specialized in that he had spent much more time in learning and practicing his craft than most of the members of his group. His lapidary work was probably somewhere on the continuum of development from an arrangement where everyone produced his own beads to that of full-time specialist subsidized by the rest of the group.
While this characterization reflects some other scholars’ views regarding economic support of specialists (e.g., J. Saunders 2004), the substantial literature on craft specialization shows that researchers are moving beyond simple attempts to document specialization. Efforts to unpack the complexity often subsumed under the term “specialist,” and to investigate gender roles in relation to specialization rather than assuming the specialists were men, are underway (e.g., Cobb 2000). Therefore, this compromise scenario “between no specialists and specialists” needs as much critical examination as the others.
An intriguing definition of craft specialization was offered by Clark and Parry (1990) that has labor, rather than time, as its focus, leading to consideration of who benefits from that labor. In their view, if one’s production of durable goods is for oneself or one’s dependents, it is not craft specialization, no matter the scale of production, consistency of product, knowledge required, and so on. They succinctly define craft specialization as “production of alienable, durable goods for nondependent consumption,” allowing for more than one type of craft specialization (Clark and Parry 1990:297). Distinctions are made as to whether craft specialization is independent or attached, and production scales are characterized as ad hoc, part-time, or full-time. They use these distinctions in examining relationships between specialization and cultural complexity, through statistical testing of data derived from ethnographic cases (Clark and Parry 1990:Tables 1 and 2). Their definition has the advantage for archaeologists of identifying craft specialization through the distribution of end products relative to the location of manufacture, as Brookes (2004) and McGahey (2005) did in the case of zoomorphic beads at Denton. Qualifying the economic focus of Saunders (2004) and McGahey (2005), Clark and Parry (1990:298) state, “craft specialization can be voluntary or compulsory and does not necessarily entail compensation in subsistence products or their equivalent.”
Some may question the utility of this definition because of a lack of focus on technology. Also, craft specialization defined in this manner is likely present in all societies, in the least complex form of independent craft specialization at an ad hoc scale of production (e.g., Costin 1991; see reaction by Clark 1995). In terms of an organization of technology model, one could argue that craft specialization is a technological strategy because it involves the keepers of knowledge and tools that form a technical system. This point is made by Brookes (2004) with regard to the knowledge and equipment used in drilling long, tubular chert beads. However, craft specialization as defined by Clark and Parry is best considered as an aspect of social and economic strategies, with implications for technological strategies.
With regard to the ubiquity of craft specialization, McGahey (2005:18) notes that the definition of craft specialization traditionally used in studies of Archaic-period lapidary finds implies that specialists would occur “only at or above the chiefdom level of sociopolitical organization.” However, Clark and Parry’s definition of craft specialization, with its focus on labor, allows for variation in craft specialization in different types of social structures to be explored. This is pointed out by Cobb (2000:36), who states:
Because specialization is a form of production, it must be examined within the wider arena of social relations that constitute the labor process . . . one must closely examine the underlying relations that structure production as it may be manifested by specialization. These deeper relations involve a complex articulation of production, distribution, exchange, and consumption at an economic level, as well as mediation of these processes through social and symbolic means.
Ultimately, variability in the production of chert beads can be explored by applying the organization of technology model. As currently developed, application of the model allows us to consider systematically: (1) data regarding lapidary assemblages, and where gaps in the data corpus exist; (2) the model itself and how it can be expanded to include ideotechnic artifacts; and (3) whether craft specialists were responsible for Middle Archaic-period lapidary assemblages.
The chert beads and related stone tools at the John Forrest site provide a way to explore the organization of this lithic technology and the question of craft specialization. While we follow others regarding the use of additional equipment, such as a pump or bow drill, in the manufacture of stone beads, specific discussions of such equipment in the organization of technology model are beyond the scope of this paper. We include data from related sites, including Loosa Yokena and Watson Brake, to explore this bead production technology. Chert beads do not seem to be part of what Hayden (1998:2) dubs a “practical technology . . . to solve practical problems of survival and basic comfort.” Hence it is necessary to broaden the scope of the model, not simply to allow its application to “prestige technologies” (Hayden 1998:11) or ideotechnic artifacts such as beads (Brookes 2004:101), but also to consider the greater cultural milieu. Based on the work of others (Johnson 2000; McGahey 2005), blade cores, blades, and microdrills are key components in the production of beads and are considered here. The organization of technology model is best applied by starting with the top two levels for context, then moving to the bottom and working up to social and economic strategies.
Social and economic strategies shape and are implemented by technological strategies, and these strategies are conducted in and are shaped by the environment and demographics. Because our focus is on chert beads and related lithic assemblages, the overall environmental description for the Middle Archaic is broad while the raw material environment receives attention at a local level. Due to data constraints, discussion of demographics is quite general. Certainly, examining both the environment and demographics in more detail at multiple scales would enhance our understanding of social and economic strategies during the Middle Archaic period.
The Middle Archaic (8900–5700 cal. B.P.) corresponds to the Mid-Holocene Altithermal, Hypsithermal, or Climatic Optimum (Anderson 2001; Brookes and Twaroski, this volume). Early characterizations of Mid-Holocene paleoenvironments focused on the climatic stability of this period, noting that in the Southeast the climate was generally warmer and dryer (Schuldenrein 1996). More recent paleoclimatological reconstructions taking account of the synergistic effect of multiple factors, including the atmosphere, ocean, and vegetation (Ganopolski et al. 1998) demonstrate that this period was marked by global climatic variability relative to earlier and subsequent periods (Sandweiss et al. 1999). Climate during the Mid-Holocene was different from today, but these differences were highly variable across space and time. For example, the Southeast was likely wetter, with more extreme seasons (Anderson 2001:158; Watts et al. 1996:36). While this variable environment did not determine the strategies used by Middle Archaic peoples, environmental conditions would have had an impact on strategy choice and the success of a particular strategy (see Brookes and Twaroski, this volume). A final point to consider here is that our current paleoclimatic reconstructions are incomplete (Peacock 2008), and with future work these reconstructions may very well change.
Raw material availability is an important environmental factor to be considered in the organization of Archaic-period stone bead technology. The overwhelming focus at the John Forrest site was on locally available materials. The majority of the chipped stone assemblage consists of Citronelle Gravels, with minor amounts of Fort Payne chert from the Tennessee River Valley (e.g., Futato 1999) and Tallahatta Sandstone (a.k.a. Tallahatta Quartzite, found in east-central Mississippi and west-central Alabama—Curren 1982; Lloyd et al. 1983; Maudsley 1998). The Citronelle Formation is a “band of secondarily deposited gravels” that extends along the entire eastern rim of the Mississippi Alluvial Valley (Stallings 1989:38). This material is dominated by tan-colored cobbles, but colors range from reds to grays. A water-rolled cortex was commonly observed on early stage beads at the site. These gravels are abundant in James Creek, which is located approximately 250 m to the north of the site. Interestingly, the raw materials predominantly used for bifaces, cores, and beads at Watson Brake (Johnson 2000:95) and Loosa Yokena (McGahey 2005:7) were also tan gravels. In fact, Johnson (2000:102) describes the Watson Brake assemblage as resembling a “generalized Archaic source-area technology” to which drill and bead production was added. The raw material used for the ground beads was likely available in the gravels (Crawford 2003; Hadley 2003).
You cannot eat rocks, so settling next to a lithic source in an area devoid of subsistence resources is not a strategy well suited to hunter-gatherers. Given that the occupants of John Forrest, Loosa Yokena, and Watson Brake had choices about where to settle, how much of a consideration was it to be near a raw material source? That is, were there sacrifices made in access to subsistence resources or social networks in order to be near lithic material because of the importance of stone in implementing social and economic strategies? If so, then this observation provides interesting insight into the energetics and workings of Middle Archaic peoples in the region.
Brookes (2004:107) points out the lack of exotic raw materials used in bead production. Although Fort Payne chert was being moved long distances and appears in the John Forrest assemblage in a very minor amount in the bifaces (n = 60, 4.6 percent) and hafted bifaces (n = 17, 4.6 percent), it is noteworthy that people in the LMV did not acquire large amounts of Fort Payne chert nor use this material in the production of beads, either due to choice or lack of access.
With regard to demographics, specific discussions are rare for the Middle Archaic period in the Southeast, and data are even rarer. In a recent overview, Kidder and Sassaman (2009:677) state for the Middle Archaic, “Human groups, their mobility already restricted by rising population pressure resulting from natural demographic increase, would have encountered greater competition for restricted or scarce resources.”
This is as specific as the available information is for this period. Certainly more detailed demographic data would provide additional insights. For example, McGahey (2005) has pointed out that the scale of production is much greater at Loosa Yokena and the John Forrest sites compared to others in Mississippi. Is this because more people at these sites engaged in bead production? Or, were many beads being produced by a few people? Answering such questions will lead to a better understanding of the complex world of Middle Archaic peoples.
The lowest levels in the organization of technology model are artifact form and artifact distribution. Artifact form is the aspect of Middle Archaic bead technology for which we have the greatest amount of information. Descriptions of the various forms of stone beads date to the original analysis of the Keenan Cache by Charles Rau back in 1877 (Connaway 1981; Crawford, this volume). In terms of artifact form, the chert beads at John Forrest are restricted in size and color because they are made from the local gravels. Bead blanks vary in length from 20 to 77 mm, with the one complete bead measuring 21.2 mm. Other tubular beads from sites discussed here measure from 10 to 24 mm, with some variability outside of this range, and zoomorphic beads are on average longer at 37.2 mm (Crawford 2003:9). Chipped stone beads tend to be tubular, while ground stone beads are more variable, including tubes, discoidals, and teardrop shaped. The majority of the chert beads at John Forrest were heat-treated, resulting in a waxy sheen and reddening of the material. Chert beads at Watson Brake and Loosa Yokena are also commonly red, although they display more variation in raw materials. Also, for the zoomorphic beads the most common raw material is red jasper (Connaway 1981; Crawford 2003:10). McGahey (2005:11) notes that color was an important factor in choosing materials for beads, but does not elaborate. The selection of red for both the chipped and zoomorphic beads suggests a pattern in its selection. However, heat-treatment is also used to improve the quality of the material before knapping, and the red color may be an unintentional result of the manufacturing process. Additional research is needed to determine what significance red beads had during the Middle Archaic.
Chipped beads have fewer shapes and colors compared to ground beads. However, while we characterize these types of beads differently based on material and manufacturing process, there may or may not have been a difference to Middle Archaic peoples. The questions remain: did the chert beads and ground beads function differently? Were the chert beads harder to make and thus represent a more valuable object, or identify a person of a different status or clan? One way to explore these questions would be to look at the distribution of beads at sites and look at their association with other artifacts and features.
The details of artifact distributions can be observed at two scales: regional and site-level. The regional distribution of stone beads still needs refinement within the boundaries, but the general limits have been established (see Crawford, this volume). As Brookes (2004:107) states, “Zoomorphic beads are found in five states, Louisiana, Arkansas, Tennessee, Alabama, and Mississippi, [and] the beads were being made and put into the Middle Archaic exchange network. Since so many have been found in Mississippi, and the remainder are all from adjacent states, it is suggested that zoomorphic beads were made in Mississippi.” There is a very limited amount of data available on the site-level distribution of chert beads and manufacturing material, due to limited excavations and much locational data being limited to site-level provenience. However, the identification of a workshop at Watson Brake Mound D and the fact that two units at Loosa Yokena produced the majority of beads in various stages points to a need for more intrasite data.
At this point in Middle Archaic bead research, the only lithic artifacts that are confidently linked to bead manufacture are blade cores, blades, and microdrills. Blade cores are characterized by their size and shape, which are likely influenced by the exclusive use of Citronelle Gravels. The blade cores at Watson Brake, Loosa Yokena, and the John Forrest site share similar shapes, lengths, and number of platforms (Johnson 2000:98; McGahey 2005:10). Blades were not studied in depth at the John Forrest site because, as noted earlier, the landowner had indicated a lack of interest in collecting debitage and it was assumed that his collection was biased. The debitage from the John Forrest site surface collection is composed of larger flakes, which probably do not represent those that were used to make the microdrills. However, the blades from excavated contexts at Loosa Yokena and Watson Brake were small enough to be products of the blade cores. The microdrills at Loosa Yokena and Watson Brake both demonstrate retouch from the ventral side of the blades, as well as narrow or tapered ends that were hafted (Johnson 2000:99; McGahey 2005:11).
Blade cores, blades, and microdrills are viewed as a continuum of tool manufacture. Blade cores were used to make blades, which were then worked into microdrills. Many of the details about this manufacturing process are not well understood. Johnson (2000) demonstrates that many of the microdrills were not removed from blade cores, and the variability of this production step in other assemblages has never been explored. Was it easier to make microdrills from blade or flake cores? Blade cores generally require more skill to produce, but were there advantages to blades over other flakes in making microdrills? Blade cores allow for a more standardized flake, which may have been important in the production of microdrills. As Joe Saunders (2004:153–154) demonstrates in his discussion of microdrills, there is still uncertainty as to how microdrills were used, and on what material. The type of drill used, hafting element, and material drilled could be identifiable with careful experimentation and micro-wear analysis.
In terms of the distribution of blade cores, blades, and microdrills at a regional level there is limited information. However, it is significant that these three sites with chert beads also have numerous blades, blade cores, and microdrills. It is also interesting that these tools were not found in the Keenan Cache, which has unfinished beads. Blades, blade cores, and microdrills appear to be evenly distributed in the unit excavations at Loosa Yokena, with large numbers of blades throughout (McGahey 2005). Conversely, blades and blade cores are highly concentrated in Mound D at Watson Brake (Johnson 2000; J. Saunders 2004). These artifacts are best identified together to determine if beads were manufactured at a site. Microdrills pose a problem because they can easily be missed without fine screening, and blade cores and blades have other uses beyond bead manufacture. Thus it would be easy to overlook a bead-manufacturing site if the beads, bead preforms, and microdrills were not found or identified properly. To understand the regional distribution of blades, blade cores, and microdrills, more excavation that specifically takes note of this technology is needed. Additionally, excavated collections could be reanalyzed specifically to identify these artifacts.
Moving through the organization of technology model, design and activities are the next considerations. Traditionally, tool design is considered as reliable or maintainable. Activities can be thought of as everything that people do, but for practical purposes with regard to technology, activities minimally include raw material procurement, artifact manufacture, use/reuse, and discard.
Not serving the same practical purpose as projectile points and other stone tools, beads were not designed with either reliability or maintainability as primary considerations. Interestingly, reliable designs involve high-quality materials and overdesigned parts to ensure that a tool will work when needed. Chert beads could be said to be overdesigned in that other materials, especially bone, would serve to make a tubular bead and, with chert, small rounded chert pebbles occurring in abundance in James Creek could simply be drilled through the center with less chance of failure than that of the thin-walled tubular beads. Clearly, design choices were dictated by something other than attempts to minimize time and energy expenditure. Expanding the concept of design beyond considerations of reliability and maintainability is necessary to investigate adequately the organization of chert bead technology and that of other prestige artifacts, or what Clark and Parry (1990) call “hypertrophic artifacts.”
Hypertrophic goods are the polar opposite of the products made by craft specialists in a factory, which are “cheap, common, utilitarian goods made for an impersonal market” (Clark and Parry 1990:293). Hypertrophic goods are special products designed to communicate social messages, and Clark and Parry (1990:296) argue that “exaggerated energy costs endow an item with more information.” Interestingly, they make the point that because hypertrophic goods are designed with a different purpose than simple utility, their production is not inefficient. This is because the increased effort has the added benefit of conveying information. From a Darwinian evolutionary perspective, such apparent wasteful activities could be explained as costly signaling (see Peacock and Rafferty 2013 and references therein). Beads are manufactured based on design choices, and the choice of using chert as a raw material, given the level of technology, provides important information about social and economic strategies however it may be viewed theoretically. We consider the tubular chert beads manufactured at the John Forrest site as hypertrophic goods and will return to this point when discussing subsequent levels of the model.
Activities lie at the same level in the model as design. For our purposes, the activities minimally necessary for discussion are those involving the use history of tubular chert beads. This history played out in the context of many other activities and included everything from subsistence practices to leisure time. Greater consideration of activities in the broadest sense would add clarity and accuracy to inferences regarding social and economic strategies, but is beyond the scope of this work. Here, we discuss tubular chert bead raw material procurement, manufacture, use, and discard.
Raw material procurement was simplified for Watson Brake, Loosa Yokena, and John Forrest because all three sites are situated near sources of gravel chert. The Keenan Cache is located in Jefferson Davis County, Mississippi, which has sources of Citronelle Gravel chert. This suggests that site location near a raw material source for chert bead manufacture and other chipped-stone tool-using activities was a consideration in the context of the other activities that were necessary for survival.
Based on current evidence, similar manufacturing techniques were employed at all three sites. McGahey (2005) essentially describes the same manufacturing process as outlined by Johnson (2000) for Watson Brake. The John Forrest bead materials fit comfortably in the Watson Brake stages, while McGahey (2005) notes the similarity of Loosa Yokena with John Forrest and goes so far as to suggest that these assemblages were produced by the same people. All stages of manufacture are present in the John Forrest assemblage. The manufacture of a chert bead during the Middle Archaic period apparently required new technologies and tools, such as a means to secure the bead during the drilling process and perhaps a pump drill. The number of people who had access to these technologies and tools and provided labor for the manufacturing process is considered at higher levels in the model. We assume that chert bead manufacture was a labor-intensive process. Experimentation is necessary to test this assumption and reveal specifics of the manufacturing process.
The use of beads, especially tubular beads, has not been given much consideration. Presumably this is because beads were obviously used for personal adornment. Zoomorphic beads have been called fetishes (Connaway 1981; Sassaman 2010) and totemic symbols (Blitz 1993b), and Brookes (2004) sees them as having been made by specialists who imbued them with magic (see also Crawford, this volume). Long tubular beads could have served as shaman “sucking tubes,” used to suck and blow away diseases, accounts of which can be found in both ethnographic and ethnohistorical accounts (Frison and Norman 1993; Mandelbaum 1940; Venegas 1759). Stone tubes that possibly functioned in this way are found at archaeological sites dating to Paleoindian times (Sellet et al. 2010; Wheat 1979). As prestige items or hypertrophic goods, chert beads served as “texts” that “become the objectification or instantiation of meaning or ideology” (Clark and Parry 1990:296). It is also possible that beads were simply valued as objects of adornment and had no other function or meaning. Unfortunately, the context of the archaeological beads has not been helpful in resolving their exact use.
Finally, there is discard. A consideration of discard is hampered by the lack of undisturbed contexts at John Forrest. The landowner did suggest that beads were more concentrated in one area of the site, but preliminary analysis of shovel-test data does not indicate any particular concentration of beads and/or microdrills. At Watson Brake a bead workshop was identified, but microdrills were widely distributed in the excavation units. For Loosa Yokena, spatial analysis was not conducted, but data tables indicate that chert bead preforms are evenly distributed for test units, while three of the 20-meter square surface collection units had concentrations of 10 to 13 bead preforms (McGahey 2005). The Keenan Cache provides an interesting case and generates many questions. Were these materials buried with the intention of retrieval or were they part of a ritual (Brookes 2004:107)? Was it the result of the work of a single individual, as some have assumed (Connaway 1981; McGahey 2005; Rau 1878), or did multiple people contribute to the cache? Due to a lack of preservation and disturbance, it is unknown whether the complete bead from the John Forrest site was in a burial or represented a small cache that had been dispersed.
One of the more intriguing aspects of the Keenan Cache, as well as the assemblages recovered from John Forrest and Loosa Yokena, is the number of “discarded” artifacts in various stages of production that do not appear to be failures. Yet identifying these artifacts as viable in the chert bead-manufacturing process could be a result of our lack of understanding of the constraints of bead manufacture. However, the fact that certain pieces appear to be taken well into the production process, such as a Stage 3 bead that is ground but not yet drilled, raises questions concerning the production process. For example, did this artifact remain at this stage of production because the maker lacked access to equipment to drill the bead? Certainly, other potential explanations are plausible. Greater consideration of artifact form and distribution, coupled with experimentation, will allow us to sort through these explanations and choose the one that best fits the evidence.
Together, design and activities provide information concerning the technological strategy. Traditionally, technological strategies are characterized simply as expedient or curated (Nelson 1991). Clearly, several aspects of curation apply to stone beads in that they were produced in advance of use and transported from location to location (Odell 1996), as evidenced by some sites emphasizing production and others with only finished products. Our conception of technological strategy is that it includes the knowledge, tools, and skills to accomplish a task, as well as a plan to expend time and energy efficiently within social and economic situations. Given the potential nuances of technological strategy, it is difficult to imagine where any technology fits into the expedient-to-curated continuum, and this is perhaps especially true for prestige or hypertrophic items such as chert beads. One aspect of the technological strategy evident for the three major bead sites under consideration here is their location at a source of raw material. Affording knappers easy access to raw material by site location is part of the technological strategy. A relatively sophisticated knowledge base and set of tools for the manufacture of chert beads was also necessary, which ranges from the production of microdrills to a pump drill, and includes a vise or means to secure beads during the drilling process. Whether everyone in the group had access to these tools, possessed the knowledge and skills for their operation, and had the time and energy to expend in the production of chert beads is under debate.
Craft specialization as defined by Clark and Parry (1990) concerns labor and who ultimately possesses the fruits of that labor. Clearly, this is a social/economic strategy that has implications for the technological strategy. Typical studies investigating the organization of technology restrict considerations of social and economic strategies to mobility and the forager-collector dichotomy. If chert beads were made by people with high rates of residential mobility, then we would expect the distribution of activities and artifacts to be greatly different than what would be expected in a sedentary settlement pattern, which is proposed by Peacock and Rafferty (2013) for the Middle Archaic period in the LMV. While we do not explore these implications further here, it also clear that social and economic strategies must go beyond a singular focus on mobility.
On the one hand, one could hypothesize a social/economic strategy that includes egalitarianism, in which the occupants of the John Forrest site had equal access to prestige items because everyone possessed the knowledge, tools, and skills to manufacture these items. The fact that a local raw material was chosen for the manufacture of the chert beads supports such a hypothesis of equal access. On the other hand, one could hypothesize that not everyone in the group was of equal skill with regard to chert bead technology. These skills would include knapping the preform, manufacturing and maintaining a pump drill, and so on. An efficient expenditure of time and energy would be to allow those with the highest skill levels to specialize in the production of chert beads at the cost of being supported to some extent by others. This support would come from the makers’ ability to trade these items or by being attached to elites who controlled the items.
Interestingly, Bar-Yosef and Kuhn (1999) see the production of composite tools during the Upper Paleolithic of the Old World as evidence of changes in social and economic strategies related to the support of skilled individuals. Bar-Yosef and Kuhn (1999:332) argue:
The creation of elaborate technological aids to foraging or other work carries with it a significant amount of “frontloading,” expenditure of time and energy well in advance of any possible return. On one hand, this requires a certain degree of foresight on the part of toolmakers. Perhaps more importantly, it requires a significant amount of cooperation and coordination of activities among members of a social group. The investment of significant amounts of time and labor in the production of elaborate technology, the potential benefits of which might not be realized for days, weeks, or even years, means that individuals were free to divert this time and labor from more immediately pressing concerns such as getting food or shelter. If an individual is able to devote many hours or even days to the manufacture of an artifact, someone else must be carrying at least part of the load with respect to gathering and processing other resources. This “division of labor” might have been transitory and minor compared to the types of occupational differentiation seen in later and larger-scale societies, and we are not arguing for rigid permanent occupational specialization during the Upper Paleolithic. Nonetheless, the ability to shift the burden of daily subsistence labor to another individual at least temporarily would be vital to the evolution of some complex technologies that began to appear in the Upper Paleolithic. Conversely, the absence of such options for cooperation would inhibit the amount of time and energy any single individual could afford to put into tool manufacture, regardless of the potential payoffs of having more elaborate implements.
Presumably, this temporary shift of labor resulted in the manufacture of composite tools that were shared within the group, thus conforming to the definition of craft specialization followed here. This highlights the fact that craft specialization, broadly defined, is potentially part of a common human adaptation and not restricted to ranked and state societies.
In our opinion, Brookes (2004) has made a strong argument for craft specialization for Middle Archaic-period lapidary makers in Mississippi, using a wide variety of sites and range of artifacts, including zoomorphic beads (see discussion by Crawford, this volume). No zoomorphic beads or preforms were recovered from Watson Brake, Loosa Yokena, or the John Forrest site. We infer from the John Forrest assemblage that craft specialists manufactured the tubular beads there, based on the lack of finished products in comparison to unfinished items in various stages of production. Alternatively, individuals could have made these items for themselves and their dependents and, upon leaving the site for some reason, took the completed beads with them. Of equal importance to examining craft specialization is investigating chert bead manufacture as wasteful behavior and exploring alternate hypotheses (see Peacock and Rafferty 2013).
That alternate explanations exist does not indicate that this intractable problem must remain unresolved. Experimentation to replicate the bead-manufacturing process would provide important insights. The type of drill used, how it was used, and the material upon which it was used can be determined through use wear and experimentation. More extensive excavations are needed to produce adequate samples for intersite and intrasite comparisons. Additionally, analyses specifically focused on artifact standardization would be useful. Logically, with craft specialization there are fewer producers, resulting in relatively little variability within the artifact category (Costin 1991). Measuring the amount of artifact standardization is the most common approach to looking for craft specialization in the archaeological record.
A final important point is that ideology should be added as having an overarching impact on the entire organization of technology model. While new to the model, archaeologists have had a long and recurring interest in ideology. Additionally, Brookes (2004) labeled chert beads as ideotechnic items and suggested that the key element in the manufacture of chert beads was not the knapping or drilling, but rather a supernatural component. While some may say it is impossible to get into the minds of prehistoric peoples, ideology ultimately has an impact on artifact form and distribution. Failure to recognize this point and include ideology in the model would mean that some variation in artifact form and distribution would always remain unexplained or misunderstood.
Following Brookes (2004), we have employed data from a number of sites that have produced chert beads to investigate the question of Middle Archaic-period craft specialization in the Southeast. We focused on a relatively new piece of the “elephant” by incorporating the John Forrest site assemblage, and we also differed from previous researchers by using an organization of technology model. It is clear that there is variation in the final form of chert beads and in their distribution within and between sites. That being said, the organization of technology model was not particularly well suited to considering chert beads. These prestige/hypertrophic artifacts caused a rethinking of certain aspects of the model, especially the levels of technological strategy and social and economic strategy. The addition of ideology as an overarching consideration is significant. Certainly, some variation in artifact form and artifact distribution is explicable only with a consideration of ideology. Thus, the model is incomplete without the addition of ideology.
Our employment of an accepted and broad definition of craft specialization could be viewed as a means to support the case made by Brookes (2004:106) that the makers of Middle Archaic-period chert beads at the John Forrest and other sites were craft specialists. In actuality, this definition allows for a shift in focus, from the simple question of the presence or absence of craft specialization in a situation where craft specialization is undefined to questions concerning whether specialists were full- or part-time and attached or unattached. New studies are required to answer these questions, such as experimental replication of chert beads; use-wear analysis of microdrills; more specific study of artifact form to address standardization; and, more fieldwork to better understand intersite and intrasite distribution. We would recommend that a single individual conduct the standardization analysis to minimize bias, and note that such a study would likely have the fewest obstacles to completion.
There is a growing interest in Archaic-period research in the Southeast, especially the Middle Archaic. Due to the work of Sam Brookes and others, the dominant mode of thinking about hunter-gatherers, dating back to the nineteenth century and reified in the 1950s, is no longer tenable. Such thinking is captured in the following quote describing Archaic-period cultures in the Eastern Woodlands:
Individuals in this society among the men would be outstanding or important according to their ability as individual hunters or according to their talent as shamans or medicine men. We have little or no indication of any degree of specialization in particular crafts, or of the development of specific role in the community [Griffin 1952:354].
While we would dispute the lack of craft specialists based on current data, the role of “shamans or medicine men” is something that has largely been ignored in Archaic-period archaeology in the Southeast outside of discussions of chert beads. The work of Sam Brookes and others in this regard will continue to inspire new research into this fascinating artifact class, place, and time.
We are grateful for Samuel Brookes’s career in Mississippi archaeology and for his excellent insight into the Middle Archaic. Without his enthusiasm and assertion that important cultural innovations and movements begin in Mississippi, we might never have explored the social implications of this collection. We would also like to thank Evan Peacock for inviting us to participate in the SEAC symposium, which developed into this volume. Thanks are due to Andy Hilburn for the nice map and Sarah Mattics for the artifact photos. We appreciate the many conversations that we have had with colleagues since we first started working with the John Forrest site and collection. We especially thank Andrew Bradbury for his comments on a draft of this paper. Finally, we would like to thank the landowner, John Forrest, for donating the surface collection and allowing us to excavate on his property, which inspired us to look into Middle Archaic cultural complexity.
