Caddo Origins, A Smith County Perspective

Attempting to trace Caddo Origins in Smith County and surrounding counties depends a lot on what we end up defining as Caddo. Separating the Caddo culture from previous cultures in East Texas becomes tedious when trying to fit the available archaeological record to existing models of Woodland cultures. Krieger stated that there was no evidence in East Texas of a Woodland (or Hopewellian) culture, with Mississippian culture beginning as early as 500 B.C. I mention this partly for the sake of argument, but also to point out that in this area there is not such a clearcut difference between the archaeology of Woodland and Early Caddo cultures. If Krieger is correct, it might make better sense to have a Formative phase of Caddo rather than trying to make a Woodland culture fit the transition from Archaic to Caddo. Based on current thinking, Caddo culture developed around A.D. 800-900, based primarily on work conducted at the George C. Davis site. In Smith County there seem to be few sites that fit into what has been termed either the Formative and Early Caddo (A.D. 800-1200) periods or Alto phase sites as defi ned at the Davis site. Rather, in this area, Caddo culture reached a florescence during the following Middle Caddo (A.D. 1200-1400) time period, at least in the number and visibility of sites on the landscape. Then, for reasons that we do not fully understand, this area was apparently abandoned by prehistoric groups.


INTRODUCTION
Attempting to trace Caddo Origins in Smith County and surrounding counties depends a lot on what we end up defining as Caddo. Separating the Caddo culture from previous cultures in East Texas becomes tedious when trying to fit the available archaeological record to existing models of Woodland cultures. Krieger stated (Suhm and Krieger 1954: 158) that there was no evidence in East Texas of a Woodland (or Hopewellian) culture, with Mississippian culture beginning as early as 500 B.C. I mention this partly for the sake of argument, but also to point out that in this area there is not such a clearcut difference between the archaeology of Woodland and Early Caddo cultures. If Krieger is correct, it might make better sense to have a Formative phase of Caddo rather than trying to make a Woodland culture fit the transition from Archaic to Caddo.
Based on current thinking, Caddo culture developed around A.D. 800-900, based primarily on work conducted at the George C. Davis site. In Smith County there seem to be few sites that fit into what has been termed either the Formative and Early Caddo (A.D. 800-1200) periods or Alto phase sites as defined at the Davis site. Rather, in this area, Caddo culture reached a florescence during the following Middle Caddo (A.D. 1200-1400) time period, at least in the number and visibility of sites on the landscape. Then, for reasons that we do not fully understand, this area was apparently abandoned by prehistoric groups.
There is some question whether Caddo culture was introduced to the area either by the actual movement of peoples or an infusion of ideas, or rather developed out of the traditions of existing cultures; more likely, both processes took place. Assuming for the moment that the Caddo culture in this area evolved out of an existing culture (with an infusion of new ideas?~the glue that held it all together), what would that culture be? The Woodland period (1000-500 B.C. to A.D. 800) is characterized by the introduction of several new technologies: introduction of the bow and arrow, more intensive agriculture in some areas, but there is little evidence to support this in East Texas, and the use of pottery. This is in conjunction with a more sedentary lifestyle. Sites of this time period have as diagnostic traits some combination of the following: small contracting stem Gary dart points, stemmed arrow points such as the Friley and Steiner types, and pottery, mostly plain, and never in great amounts. Supposedly the cultures associated with the Woodland period lead to the development of what we call Caddo culture.
One idea put forth by Schambach (1970) is that Caddo culture evolved out of the Fourche Maline culture with its particular traits. Another idea, put forth by Story (1990), is that Caddo culture evolved from not one predecessor group but rather from several different groups with distinct but relatively similar sub-traditions, these being the: (1 ) Arkansas River Valley; (2) Woodland edge; (3) Red River valley; and (4) Piney Woods. However, the Woodland culture in the Smith County area differs in some degree from the Fourche Maline culture on the Red River and the Mossy Grove/sandy paste culture in the southern part of the Piney Woods. On several grounds, Perttula defined this area between the two as being occupied during the Woodland period by the Mill Creek Culture (Perttula and Nelson 2004: 155-170). In this part of East Texas, the Woodland sites differ from Fourche Maline in that they lack the intense middens and large amounts of pottery that otherwise characterize Fourche Maline sites; there are also no chipped hoes or Poole pipes. The sandy paste ceramic tradition of the Mossy Grove culture does not extend this far north into the northern part of Smith County. The Mill Creek sites appear to have been smaller and occupied for shorter time periods than is the case with Fourche Maline sites in the Red River valley, for instance, and are journal of Northeast Texas Archaeology, Volume 31, 2009 lacking the intense middens associated with the latter sites. There also seems to be very little pottery associated with these sites when compared to later Caddo sites.
Differences between Woodland and Caddo sites in this area include:

Browning Site (41SM195A)
The Browning Site is one example of the Mill Creek culture in this area of East Texas (Walters 2004). The main occupation is dated by radiocarbon and Oxidizable Carbon Ratio dating to between A.D. 600-800. This date and associated artifacts indicate it was occupied during the latter part of what has been termed the Woodland period in the region.
Friley and Steiner are the most commonly recognized types of arrow points collected at the Browning site. Other stone tools include flake tools, seven ferruginous sandstone ground stone tools, and 12 dart points (with small Gary points the most common type).
The 40 sherds from the Browning site are discussed in detail by Walters (2009). It is noteworthy that the ceramics at the Browning site closely resemble later Caddo ceramics in thickness, surface treatment, firing, and hardness, and to some extent with respect to vessel decoration, and they would be very hard to separate from the ceramic sherds on any nearby Caddo assemblage. Twenty-eight (70%) of the sherds are grog-tempered. Another 15% have grog/hematite temper. Four (10%) have a combination of grog and bone as tempering agents; none of these sherds have an abundance of bone. Two (5%) sherds have no discernible temper. Twentysix (65%) of the sherds have been fired and cooled in a reduced oxygen atmosphere. Thirteen (32.5%) were fired in a reducing atmosphere, and then allowed to cool in the open air. One sherd (2.5%) was completely oxidized during firing.
Six of the sherds at the Browning site are decorated. Three body sherds have single straight incised lines, while two sherds had two parallel straight incised lines; the distance between the incised lines ranges from 12.2-13.0 mm. The one decorated rim has a single straight horizontal incised line on it. A second rim is from what appears to be a plain carinated bowl.
The amount of ceramics at the Browning site is meager when compared to later Caddo sites in the area that are distinguished by their sheer volume of sherds. At the Browning site, the sherd density is only 1.96 sherds per m 3 . By comparison, in excavations at the 14th century A.D. Leaning Rock (41SM325) Caddo site, the sherd density is 280.4 sherds per m 3 (Walters 2008). It is uncertain why there are so few sherds represented at Mill Creek culture sites compared to what is seen on Fourche Maline or Mossy Grove sites, but evidently ceramics played a minor role in the lives of the people that lived at the Browning and other Mill Creek sites.

Boxed Springs Mound Site (41UR30)
The Boxed Springs site is an Early Caddo (ca. A.D. 900-1200) multiple mound center in the middle reaches of the Sabine River (Perttula and Wilson 2000). The site consists of four mounds a1ranged around a plaza with borrow pits, midden areas, and at least one large Caddo cemete1y where some 150 graves were looted. Sam Whiteside excavated Mound A, a circular burial mound, in the 1960s (Figure 1), uncovering a rectangular pit with evidence of at least three individuals. This burial is quite similar to several shaft burials at the Gahagan mound site on the Red River in n01thwestem Louisiana, in that the latter were large rectangular tombs with multiple inte1ments with grave goods placed along the walls and in the comers of each tomb. Offerings in the Boxed Springs burial tomb included 55 arrow points (Alba, Hayes, and Catahoula-like) in four clusters. There were also two large Gahagan bifaces, five celts, and polishing stones in the tomb. Additional grave goods included seven ceramic vessels: two plain bottles; a Spiro Engraved beaker; an everted lim jar with a pinched body decoration and zoned incised-cane punctates; a plain calinated bowl; and two plain jars. The first sherd is a jar lim with a suspension hole (Figure 2a). The rim is direct and has a rounded lip; its olifice diameter is 20.0 em. The sherd is from a vessel that was fi1ed in a reduced atmosphere and it had grog temper. It is classified as Coles Creek Incised, var. Coles Creek, with horizontal incised lines that a1e slightly overhanging and smoothed. A row of tliangular punctates have been placed below the horizontal lines. The sherd was submitted a few years ago for instrumental neutron activation analysis and the results indicated the vessel was made from local clays. The next sherd (Figure 2b) is from the same vessel. There is another example of Coles Creek, var. Coles Creek from the site that is a body sherd (Figure 2d).
Two sherds are decorated with randomly or freely-placed v-shaped tool punctates. One sherd is a lim, slightly everted, with a rounded lip (see Figure 2e ). It comes from a vessel fired in a reduced atmosphe1e; charred plant matelials we1e the only temper. Decoration on the sherd consists of randomly or freely-placed v-shaped tool punctates. The other v -shaped punctated sherd is a body sherd from a vessel fi1ed in a reduced atmosphere. It has been tempered with grog and small, finely cmshed bone; the sherd also has a sandy paste with visible qua1tz grains. The sherd is also decorated with v-shaped random tool punctates defined by a single straight incised line (see Figure 2f).
The last sherd is a body sherd with grog temper. The sherd came from a vessel that had been fired in a 1educed atmosphere. The decoration on the sherd is similar to Marksville Stamped, var. Troyville with dentate rocker stamping (see Figure 2c). The Henry Chapman site (41SM56) has examples of Early Caddo pottery (see Walters, this volume). Holly Fine Engraved and Hickory Fine Engraved wares are common in the ceramic assemblage, although few examples are as well executed as the examples from the Geotge C. Davis site. Whether this means that the Smith County sites date later in time or were just poor copies of the vessels that were being produced at the Davis site is not known. Other examples of Early Caddo ceramic types present at the Henry Chapman site are Crockett Curvilinear Incised, Davis Incised, and Pennington Punctated-Incised; Weches Fingernail Impressed, found on many Early Caddo sites, is absent.