Caddo Ancestors

The Middle Caddo period (A.D. 1200-1400) was a time of changing settlement patterns, economic changes, and exuberant experimentation with engraved and utilitarian pottery styles. In fact, archeologists sometimes have difficulty recognizing Middle Caddo sites as such because there are few well-defined pottery styles. Compared to the Early Caddo pottery, which looks very similar from place to place, the fine wares made by Middle Caddo potters express far greater individuality and creativity.

This upsurge in ceramic creativity during the 13th and 14th centuries coincides with the expansion of Caddo settlements in many places across the homeland. A shift seems to have taken place from a predominance of larger mound centers like the Crenshaw and Davis sites to smaller communities including small villages and hamlets as well as agricultural farmsteads. This shift may be mainly a reflection of an economic shift from a more diverse economy toward increasing reliance on corn agriculture. Corn farming requires lots of space and the growing crop must be closely watched, weeded, and, as it matures, guarded to keep birds and mammals (especially raccoons) from ravaging the corn. Many Caddo farmers moved to the country, so to speak.

In recent decades archeologists have investigated various large and small sites dating primarily to the Middle Caddo period. They have found everything from small seasonal campsites or single-family farmsteads that seem to lack substantial houses and refuse middens to ritual (mound) centers that continue many of the patterns seen at the Davis site (which continued well into the Middle Caddo period). The larger sites were important civic-ceremonial centers containing multiple mounds and associated villages. The multiple mound centers are rather evenly spaced along the Red River, the Sabine River, and Big Cypress Bayou, and those that are contemporaneous may represent an integrated regional network of interaction and redistribution. For example, the Jamestown (eight mounds and village), Boxed Springs (four mounds, village, and large cemetery), and Hudnall-Pirtle (eight mounds and 60-acre village) sites appear to represent the apexes (central places) of three Early and Middle Caddo networks in the Sabine River basin.

Among the premier mound centers in the Neches-Angelina river basin, the Washington Square site in the middle of what is today Nacogdoches, Texas, dates mainly to the Middle Caddo period. The site once consisted of at least three mounds separated by a long plaza area and with an associated village. Most of the site has been destroyed as Nacogdoches has grown. Two mound remnants have been partially excavated by Jim Corbin and his students at Stephen F. Austin State University and the 1985 field school of the Texas Archeological Society. Extensive excavations documented a circular structure under Mound 1/2, an assortment of pits and postholes in non-mound contexts, and several large burial pits in a mortuary mound (the Reavely-House Mound). Although little evidence of the village apparently survives, large sherd-filled pits (representing many vessels) were encountered in an area between the two mounds. These are thought to be deposits from public feasting activities led by the Caddo elite that used the Washington Square site as a ceremonial center in the 13th and 14th centuries.

The Haley site is another ceremonial center dating to the Middle Caddo period and one of the most famous of all Caddo sites. It is located on a levee of the Red River in southwestern Arkansas about 20 miles north of the Louisiana state line. The Haley site had two prominent mounds and probably once had other smaller mounds. C.B. Moore excavated most of one mounds in 1912, finding nine shaft tombs now known to date to the Middle Caddo period. He tested the other mound and, finding no burials, concluded (probably correctly) it was a "temple mound." Since the 1960s, amateur archeologists and pothunters have found two sizable cemeteries at the site and have dug up over 100 burials.

Although some archeologists have suggested that Haley was strictly a "ceremonial and burial center," we suspect that associated village areas were close by. The Haley site is best known for its unusually ornate pottery, which exemplifies Middle Caddo pottery creativity. Tellingly, many of the pottery styles found at Haley are known only from a short stretch of the Red River Valley extending southward into northernmost Louisiana. A few obvious trade pieces do show up elsewhere, but the limited distribution area must demark the territory of the local group (polity).

Some 100 miles to the southwest from Haley, a fascinating Middle Caddo village was discovered during archeological investigations in advance of a large lignite (coal) surface mine. The Oak Hill Village site in Rusk County, Texas, occupied a small ridge overlooking Mill Creek, a tributary of the Sabine River. More is now known about this Oak Hill Village site in Rusk County, Texas than perhaps any other site dating to the period. On behalf of TXU Mining Company LP, archeologists from the private consulting firm PBS&J used heavy machinery and hand excavation to open up broad areas, and thus were able to expose the "footprint" of virtually an entire small village. Postholes and pits dug through the sandy soil into the underlying clay were visible once large areas of the site were scraped off with machines and then cleaned by hand. The ridge upon which the village was built had been plowed and contoured by modern farming, thus disturbing the upper deposits of the site.

Oak Hill Village had at least 42 circular and rectangular structures representing at least three successive villages. The structures that were part of the last and largest village were arranged over a 3.5-acre area in a circular pattern around a central plaza area. Sorting out which of the structures were contemporaneous was very difficult because some of the structures had been rebuilt over and over and some overlapped earlier structures, particularly at the northwestern end of the ridge and plaza. Working out how the village changed during its 300-year span was also hard because most of the structures were only recognized after the heavy machines had scraped away what would have been the floor level (thus, removing any evidence of the actual house floors and artifacts that had survived modern plowing). Nonetheless, analysts were able to work out the basic sequence of the village phases, called here the early, middle, and late villages.

The Early Village at Oak Hill began as early as A.D. 1150 (or perhaps a few decades later). It consisted of four large rectangular structures, each measuring about 7 x 11 meters (23 x 36 feet). These are thought to be houses that would have been suitable for 10-20 people. A total of 40-70 people probably lived at the site during this phase. Rectangular houses are not common in the region, but are known at other Middle Caddo sites, including the Ferguson site in Arkansas, the Belcher site in Louisiana, and the Hines site in Wood County, Texas. The early houses at Oak Hill were not arranged around a plaza.

By A.D. 1250 the Middle Village had been reorganized and the house style had changed from rectangular to circular buildings like those typical of most Caddo sites in northeast Texas. The middle village phase was the most intensive of the three and lasted a full 100 years. Most of the superimposed building patterns (most of them surely houses) date to this phase. The exact number of houses in use at any one time is unknown because of the overlapping outlines. The structures making up the middle village at Oak Hill were arranged around an open plaza covering at least 15 x 25 meters (about 50 x 80 feet). The total village population is estimated at between 75 and 100 people.

One of the clusters of three overlapping structural patterns at the north end of the Middle Village was apparently covered by a low mound, upon which a new building may have been placed (the evidence was destroyed by plowing and machine scraping). This mound probably covered a special building of some sort, perhaps the house of the village headman or a small temple.

During the Late Village phase, dating from about A.D. 1350 to as late as 1450, the village expanded to the south and new types of structures were built along with more of the circular houses. Total village population is estimated to have reached an all-time high of at least 100, perhaps a bit more. Two of the new structures at the south end have long extended entranceways pointing to the north (and the main village area). These are clearly "special buildings" although it is not known whether they were temples or the houses of the village headman or priest. Also new during the Late Village is the addition of interior partitions to several buildings and the construction of four small circular structures thought to be granaries.

Although the Oak Hill Village deserves a full TBH site exhibit, a few tantalizing details are presented below and in the accompanying photographs, based on the work of the archeologists who analyzed and reported the Oak Hill Village, Robert Rogers and Timothy K. Perttula, and their colleagues.

The main construction timber used throughout the life of the Oak Hill villages was oak. There is no evidence that any of the buildings were burned, suggesting that they decayed naturally after use life of no more than 15 years. One technique used to prolong the life of the upright timbers was to char the ends of the posts that were set into the ground. Some of the overlapping and intersecting house patterns suggest that some of the houses were rebuilt using borrowed timbers (perhaps those that had snapped off at the ground level).

Many charred corn cobs were found in several pits within houses that were either cooking fires or what archeologists call "smudge pits," fires intended to produce lots of smoke either for smoking meat or keeping mosquitoes at bay. The fragmentary cobs were studied by plant experts (paleobotanists) who suspect the diverse sample represents as many as three or four different varieties of corn. Eight, 10, 12, and 14-row forms were found, although the number of rows is not clearly linked to variety (ears with different numbers of rows occur within a single variety). While it was not possible to identify these as known varieties, it does seem likely that the Oak Hill villagers were growing several different types of corn. This fits with early historic accounts that describe the Hasinai groups growing an early maturing flint corn and a later maturing flour corn. You can see the advantage of having varieties that could be grown at different times of the year and those that would tolerate drought better.

The 300-year span of the successive Oak Hill villages illustrates nicely the increasing importance of corn. Corn was found in 32% of the Early Village soil samples from hearths and pits, 50% of the Middle Village samples, and 97% of the Late Village samples. The Late Village granaries reinforce these data. The village economy, though, depended on a mix of crops, wild plants, and wild animals throughout the site's history. Hickory nuts are present in almost every sample. Other nuts include acorns and walnuts. Fruits that were eaten include persimmons, dewberries, and grapes. The finding of charred maygrass and goosefoot/pigweed indicates that these starchy seeded weeds were probably grown (or encouraged) as well. Deer was the hunter's favorite target and the main protein source. Other hunted or trapped animals include rabbits, buffalo (a single animal), and squirrel. Fish, birds, and reptiles round out the list and that is only what was preserved. Dozens of other plant and animal species would have been part of the diet.