University of Texas at Austin wordmarkCollege of Liberal Arts wordmark
Texas Beyond History
TBH Home
Tejas Main

Early Caddo, A.D. 800-1200

Artist's deptiction of an early Caddo village scene
Early Caddo village scene as envisioned by artist George C. Nelson. All of the details are based on archeological evidence from the Davis site or on early historical descriptions of Hasinai Caddo groups. Courtesy of the artist and the Institute of Texan Cultures.
villages walking up bluff
Artist's depiction of villagers excavating soil from a barrow pit and lugging it up a steep bluff edge on their way to an Early Caddo mound. This scene is based on archeological evidence from the Davis site in east Texas. Painting by Nola Davis on display at Caddoan Mounds State Historic Park. Courtesy Texas Parks and Wildlife Department.

Click images to enlarge  

barrow pit
The large depression left by barrow pit at the Davis site is still visible today, over 800 years later. TARL archives.
hand cut out of copper
Hand cut out of thin sheet copper, one of a pair found in Burial Pit 2 at the Early Caddo Gahagan Mound on the Red River in northwest Louisiana. Probably related to hand-and-eye symbolism found on Southern Cult objects at Spiro and other Mississippian sites. Webb-Dodd Collection, Louisiana State Exhibit Museum. Courtesy Pictures of Record, Inc.
cleaning the walls of excavation trenches
Archeologists cleaning walls of excavation trenches at Mound C, Davis site, Cherokee County, Texas. This mound was built over a 300-400 year period, basketload by basketload, by the early Caddo community as a special cemetery for their leaders, who were entombed within deep pits in the original soil below the mound. The contrasting layers of fill visible in the walls reveal the complex history of the mound. TARL archives.
aerial photo of Crenshaw site
1960s aerial photograph looking south across the Crenshaw site. The Red River, visible at the top of the picture, once flowed closer to the site. Mound A, the site's largest is near the bottom of the picture, the tree-covered area jutting out into the cleared field. Other mounds are in the central tree-covered area in the ploughed field. Courtesy Frank Schambach.
Lemley crew
Judge Harold J. Lemley hired this crew to dig at the Crenshaw site in the 1930s. Although primarily an artifact collector, Lemley's foremen (two men on left, Glenn Martin and S.D. Dickinson) took photographs and some notes of the mass burials his crews encountered that are today attributed to Fourche Maline culture. Click to enlarge to see details. Courtesy Frank Schambach.
raised causeway
Unique raised causeway that connects Mounds A and E at Crenshaw. This unusual feature dates to the Caddo occupation of the site and may have been built at the end of the site's history when a raft lake on the Red River apparently flooded the site. Photo by Steve Black.
triple-chambered bottle
Triple-chambered Early Caddo bottle from the Crenshaw site. Photo by Frank Schambach.
refuse pile of antlers and skulls of deer
This massive refuse pile containing the antlers and frontal skull sections of over 1000 white-tailed deer gives the "Antler Temple" its name.
outlines of two large houses
The outlines of two large houses uncovered at the Davis site thought to be left by beehive-shaped grass-thatched houses like those described by early Spanish observers. Dubbed Features 139 and 160, they measured about 10.5 and 12.5 meters (34 and 41 feet) in diameter and probably each housed a large extended family or several smaller families. TARL archives.
artifacts made from exotic stones
Artifacts made from exotic stones that were left as offerings in the Mound C tombs at the Davis site. In the center is a chunky stone similar to those used by many Indian groups in the Southeast to play a game. TARL archives.
ceremonial staffs
Probable ceremonial staffs of office (dubbed "spuds" by archeologists, an unfortunate, but memorable label) from tombs of paramount individuals (perhaps caddis?) entombed in Mound C at the Davis site. TARL archives.
Davis community layout
Davis community layout as depicted on a brochure at Caddoan Mounds State Historic Park. Ongoing investigations are developing a much more detailed picture of how the community was organized. Courtesy Texas Parks and Wildlife Department.
Mound A Inner Precinct
Mound A Inner Precinct as exposed by the WPA excavations. Upon and around the mound were a series of large public buildings 12-16 meters (40-53 feet) in diameter that probably served as temples, council or clan houses, and perhaps the residences of community leaders and priests. Graphic by Dee Ann Story.
stone plummet
Polished stone plummet included as a grave offering. Such artifacts are characteristic of Poverty Point culture (ca. 1700-1000 B.C.). Thus, this object is interpreted as a heirloom piece perhaps representing a symbolic connection with the ancient past. From Mound C, Davis site. TARL archives.

About 1200 years ago the Caddo cultural tradition begins to emerge on the western frontier of the Southeastern culture area as a unique and early expression of the larger Mississippian cultural tradition (see Mississippian World section for background). Parallel developments happened in many other areas of the Eastern Woodlands, but not necessarily at the same time and in the same ways. The Mississippian world was a mosaic of cultural traditions sharing, to varying degrees, what we might call common "themes." These overlapping themes include maize agriculture, settled life, elaborate ritual life, ranked societies, mound-building, ancestor veneration, ritual/political centers, long-distance exchange, warfare, competition, effective methods of food gathering, hunting, processing, and storage, and sophisticated craft production. Such themes were shared mainly through the exchange of ideas and things from group to group rather than through migration. In this section we look at how these themes were expressed in the Caddo world.

Here, we use the term "Emerging Caddo" to refer to the 200-year period between A.D. 800-1000 when the Caddo cultural tradition emerges from its Late Woodland origins; the term Formative Caddo is often used for this period. We use "Early Caddo" to refer to the succeeding 200 years (A.D.1000-1200), when the Caddo archeological tradition is in evidence across the Caddo world. We will also use the lower-case "early" Caddo to refer to the combined 400-year-long era (A.D. 800-1200). Keep in mind that these archeological concepts and periods are standardized approximations of developments that, collectively, unfolded over at least 16 generations. Much of what happened during the 400-year development of early Caddo society is not very well understood for two interrelated reasons: not enough investigation and too few radiocarbon dates from key investigated sites.

Only a few Emerging and Early Caddo sites have been excavated in East Texas and, of these, only the George C. Davis site has seen extensive scientific investigation. Most of the excavators of early Caddo sites have concentrated on burial mounds, yielding fascinating, but one-dimensional information. While we now know quite a bit about how elite members of early Caddo society were treated in death, we have little concrete understanding of their roles in life. And we know even less about the lives of ordinary people. There are good reasons to suspect that early Caddo society was quite different in some respects than the Caddo society known from early historic accounts. Few societies remain unchanged for eight centuries.

The second, companion problem is that we have comparatively few radiocarbon dates from secure contexts and even fewer sequences of dated occupations at early Caddo sites. As a result, the timing of the emergence and development of early Caddo culture is unclear. Based on our reading of the data, Emerging Caddo culture dates at least as early as the mid 9th century (ca. A.D. 850) and perhaps a half-century earlier. At the time, a Caddo group related to peoples living along the Red River in extreme northwestern Louisiana established a new community near the Neches River on what became the southwestern frontier of the Caddo Homeland in the Piney Woods of east Texas.

This community, known to archeologists as the George C. Davis site, consisted of a village and a ritual center marked by three earthen mounds. Since this is the best known of the early Caddo sites in the main Caddo Homeland, we will discuss it in some detail, but in doing so we do not want to give the impression that all early Caddo sites were like the Davis site or that it is the earliest identifiably Caddo site. Neither impression rings true.

The Davis site looks like it was established anew, as if its founders set up a new community based on ideas and experience gained elsewhere. For this reason we think that the Davis site cannot have been the first recognizably Caddo community and that somewhere to the northeast, perhaps along the Sabine or, more likely, on the Red River are even earlier Caddo communities that represent the true emergence of the diverse Caddo tradition. It is possible, we suppose, that the founders of the Davis site were one of the earliest groups to develop what we see today as early Caddo culture. In such a scenario, splinter groups (colonies) bent on breaking from tradition and following new ideas (reflecting differing religious and political/organizational wonts) may have moved to the western frontier—the Piney Woods of east Texas. We could settle this question (and quite a few others) with a much larger set of well-dated early Caddo sites.

Crenshaw Site

The Crenshaw site on the north side of the Red River in Miller County, Arkansas (and, more generally, the Great Bend area) is a prime candidate as one of the key places where the Caddo cultural tradition developed. This mound center on an old channel of the Red River has yielded a wealth of late Fourche Maline and early Caddo artifacts, graves, and architectural features. Many excavators have worked at the site beginning with C.B. Moore in 1912. Unfortunately, most of the digging since then was done by looters or by minimally trained amateur archeologists. Trained professional archeologists, most notably Frank Schambach of the Arkansas Archeological Survey in the late 1960s and early 1970s, have investigated small parts of the site.

Crenshaw is a very complex site with a long history, but most of the dating is educated guesswork because so much of what is known comes from looters and early digging. Besides that, the Crenshaw site is mighty … peculiar, for lack of a better way to put it. Beginning in Late Fourche Maline times, roughly A.D. 600-900, the site was used as a burial ground and ceremonial center, a place where mass burials and rituals took place, some being unknown elsewhere in the Caddo Homelands. Sometime after A.D. 900 people with a recognizable Caddo culture (possibly the direct descendants of the Fourche Maline people who used the site earlier) began using the Crenshaw site as a ceremonial center, adding some bizarre twists of their own. While part of the reason that the Crenshaw site looks peculiar is because we have not yet examined very many contemporary sites, the site may well have a uniquely important place in Caddo history. Here are a few examples of the seemingly unusual ritual practices we are talking about.

In late Fourche Maline times, isolated human skulls and isolated mandibles (jaws) were placed on low earthen platforms and then covered over, forming a burial mound. Some skulls and jaws were accompanied by a few grave goods such as pots and beads, but most were not. Schambach thinks the bodies were buried elsewhere, perhaps at outlying villages, and that the heads and jaws were kept in charnel houses or on scaffolds until time came to bury them at the ritual center of Crenshaw. Hundreds of years later, between about A.D. 1250-1400, Caddo people placed numerous small groups of skulls and mandibles in small pits within a large open area at Crenshaw that is known as the Plaza of the Skulls. While a few pits containing isolated skulls reportedly have been found at at least one other Caddo site, the Crenshaw site may have been the focus of some sort of cult that involved the handling and burial of skulls, sometimes weathered and defleshed, sometimes freshly severed heads. Nothing like this is known from historic accounts in the Caddo area.

The Crenshaw site also has a number of mass graves dating to the late Fourche Maline and possibly the Emerging Caddo periods. Three of the mass graves have rows of intact skeletons (10, 27, and 43 individuals) that were laid neatly, one against the next, on a mound surface that was then covered with a new layer of earth. Still other mass graves were rectangular pits with rows of bodies neatly laid side-by-side. Another mass grave is a circular pit about 50-feet across into which over 60 individuals were placed. Some were intact and completely articulated, suggesting that the deceased were placed in the pit soon after death. Other skeletons were weathered and jumbled, suggesting that these had been processed in a charnel house and brought to the pit in bundles. The pit was then covered with a thin layer of soil, atop which a large fire was built and a feast was held, judging from the large number of broken animal bones. These remains were then covered quickly by a large earthen mound. Once again, nothing like these behaviors is known from historic accounts of Caddo life.

A final example of fascinating but unique early Caddo behavior at the Crenshaw site is a set of features dubbed the "Antler Temple" and an associated refuse pile containing the antlers and frontal skull sections of over 1000 white-tailed deer. Schambach thinks the inferred temple dates to early Caddo times and might have been the residence of a shaman or priest, similar to the Hasinai fire temples documented in early Spanish accounts. (While no postholes or other direct structural evidence was detected, the debris patterns suggest that a structure had been present.) The finding of human teeth on the apparent temple floor suggests that human skulls and mandibles may have been stored in the temple prior to burial in the nearby Plaza of Skulls. (New radiocarbon dates published in 2016 indicate the adjoining ash bed structure or temple dates from about A.D. 1160-1250.) Schambach notes that Spanish accounts also describe successful Hasinai hunters bringing their deer to the caddi (Caddo leader) for blessing and ritual skinning. Deer heads were apparently part of the rituals. While the Antler Temple seems to represent a relatively early example of ritual practices still present among historic Caddo groups some 700 years later, it is still unique among known Caddo groups.

Crenshaw may seem so peculiar simply because, fortuitously, the evidence there, particularly bone, is well preserved by thick layers of slackwater mud deposited when timber log jams on the nearby Red River formed and flooded low valley areas. In contrast, very little bone is preserved at the Davis site, which remained high and dry throughout its history. Nevertheless, Crenshaw seems to be a very atypical site with a special place in Caddo history. Schambach thinks that during early Caddo times that the site was a vacant ceremonial center where only a few priests and attendants lived. In this scenario, the ritual center would have served outlying communities presumably living in nearby villages and hamlets. Given the recent recognition, discussed below, that numerous hard-to-recognize house patterns occur around and between the mounds at the Davis site, the vacant ceremonial center notion is open to question. The same mud deposits that provide good preservation conditions at Crenshaw may yet conceal an early Caddo village. For now Crenshaw must be regarded as an unusual but highly important early Caddo site that needs further work and a solid radiocarbon chronology before its role in the development of early Caddo society can be fully understood.

Davis Site

The George C. Davis site in the Neches River valley in east Texas has long played a central role in our understanding of early Caddo history. Much of this site in Cherokee County, Texas is now protected by Caddoan Mounds State Historic Site and can be visited by the public. The Davis site's importance is an accident of the history of archeological investigation more than it is a reflection of the site's actual role in Caddo history. There are other early Caddo mound centers that are larger than Davis and located in more central parts of the Caddo Homeland, such as along the Red River and the Sabine River. But no other site has had as great an impact on what we know about the ancient Caddo. Because of its importance, the George C. Davis site deserved to be the subject of a future major exhibit on Texas Beyond History. In the meantime, we provide a brief summary of some of the most important aspects of the site, beginning with a short history of its investigation.

History of Investigations

In 1939 the Davis site was chosen for major excavation by the University of Texas at Austin for a WPA project, in part because it was the nearest major Caddo site to Austin and because there were many unemployed men in Cherokee County who needed work. Over the next 18 months a sizable WPA work crew excavated about half of the site's largest mound (Mound A) and its surrounds under the leadership of Perry Newell, an excellent field archeologist and meticulous recorder. Newell and his crew uncovered the often-overlapping outlines of an amazing series of large houses and special buildings around, beneath, and within the mound.

During and just after World War II an extraordinary scholar, Alex D. Krieger, analyzed the artifacts from Davis and compared them to materials from dozens of Caddo sites in Texas and adjacent sites. He worked out the long sequence of Caddo prehistory and demonstrated that the Davis site dated long before the historic period. The resulting 1949 report by Newell and Krieger provided the first truly detailed description of a major early Caddo center coupled with a bold interpretive framework that provided the first synthesis of early Caddo history. It is still regarded as a classic archeological site report.

In 1968, Dee Ann Story, professor of anthropology at UT Austin, took up the Davis site where Newell and Krieger left off. Over a three-year period Story's team excavated major parts of the remaining two mounds at Davis and investigated a number of other features at the site. Mound C proved to be, as Krieger suspected, a burial mound that contained a series of large tombs placed in deep pits. Mound B, the site's smallest and lowest mound, was found to have been a capping mound build over the foundations of at least one large, dismantled, special-purpose building and a platform upon which additional special buildings may have been built. In succeeding years during the late 1970s and early 1980s, Story returned to the Davis site with undergraduate and graduate students and carried out off-mound investigations in many areas of the site. One critical aspect of her work is that she obtained over 125 radiocarbon assays, making the Davis site the best-dated of all Caddo sites. The many scholarly articles, theses, and dissertations that resulted from Story's investigation made the Davis site that much more famous and important.

The third era of investigation at the Davis site began in the early 2000s has begun under a new generation of researchers at UT Austin. Darrell G. Creel, then director of TARL, and Samuel Wilson, then professor of anthropology, directed new work at the site using state-of-the-art geophysical survey equipment. A device that senses subtle changes in the earth's magnetic field was used to create a new map of the site. This magnetic map revealed the existence of dozens of previously unseen structural patterns representing houses, special buildings, and various other features. This work gives us a much more complete look at the Caddo community that called the Davis site home for centuries.

Community Layout

The Davis site occupies a sizable area of a flat-topped river terrace remnant created long ago by the Neches River, which flows south to the Texas coast. For thousands of years prior to the arrival of the first Caddo group, hunting and gathering peoples (including those of the Mossy Grove culture) had camped in the vicinity from time to time, leaving behind stone tools and meager traces. Sometime in the mid to late 9th century, Caddo peoples established a community that was to last for some 400-450 years. Through time the community grew to perhaps 250-300 people and covered an area of about 60 acres (roughly 600 x 900 meters or 2000 x 2900 feet), within which three large earthen mounds were constructed.

Today the three mounds are the only visible reminders of the community of ancestral Caddo people whose houses, granaries, temples, dance grounds, and fields once covered much of the Davis site. While we would love to be able to hop into a time machine and visit the Caddo community in its prime some 800-900 years ago (and at various other points in its history), we cannot. The artists' depictions that accompany this exhibit give us good ideas of what the community may have looked like. Artists Nola Davis and George Nelson based their paintings on archeological details, early historic accounts of Caddo life, and educated guesswork. Using a similar approach, we can try to paint a word picture of the ancient Caddo community at its prime.

Around A.D. 1100, only two of the mounds existed. Mound A was the focal point of what Dee Ann Story has called an "inner precinct." Upon and around the mound were a series of large public buildings 12-16 meters (40-53 feet) in diameter that probably served as temples, council or clan houses, and perhaps the residences of the community's leader(s). Also in this inner precinct were several buildings with unusual floor plans that differed from the ordinary circular pattern as well as several very small structures that may represent granaries. Open areas within the inner precinct are probably small plazas perhaps used during ritual gatherings and dances. Just outside the inner precinct was a village area where smaller, ordinary houses (9-12 meters in diameter) had been built and rebuilt in the same area for generations.

Five hundred meters north of Mound A was a special cemetery marked by Mound C. Within and beneath the mound were several dozen tombs containing the bodies of community leaders, some of whom appear to have been accompanied by retainers or individuals sacrificed for the occasion. It is likely that only select individuals living in the inner precincts were interred in the special cemetery. As far as is known, there were no houses built near Mound C.

Later in the site's history, a second, smaller precinct began forming 300 yards north of Mound A with the building of the site's largest known building, a circular structure some 18 meters (59 feet) across, and probably several other buildings. This building was dismantled at the end of its life, burned, and then capped by Mound B and additional structures may have been built atop the mound. Not enough excavation has been done in the area to be able to determine much else about this second inner precinct.

The houses of many ordinary villagers were probably scattered across the areas between the mounds and outside them, presumably surrounded by garden areas where corn and other crops were grown. Outside work areas would have also been associated with this "outer village."

The community sketch just given is based mainly on what is known about the 10% of the Davis site that has been excavated, leaving us to fill the other 90% largely by conjecture and the results of geophysical surveys. Community reconstruction is also greatly hampered by the difficulty of precisely dating the various known structures, many of which were excavated by the WPA before radiocarbon dating was invented.

These same difficulties are faced by archeologists trying to interpret any ancestral Caddo site; for most, far less is known than is the case for the Davis site. Happily, modern technology has recently been applied to the Davis site with very encouraging results. The maps below show the magnetic patterns detected in one area of the Davis site within an open area that had seen little previous investigation. Clear structural patterns are visible as well as a number of suggestive but as-yet unclear patterns. Test excavations in 2003 confirmed that the structural pattern outlined in red on the map on the right is indeed that of a large building with a central hearth and four central posts that served as roof supports.

communal work party
Artist's depiction of a communal work party dumping basketloads of soil to form a new layer of an Early Caddo mound. This scene is based on archeological evidence from the Davis site in east Texas. Painting by Nola Davis on display at Caddoan Mounds State Historic Park. Courtesy Texas Parks and Wildlife Department.
eagle effigy pipe
Eagle effigy pipe from a deep, Early Caddo shaft grave (Pit Q) in Mound C at the Crenshaw site, Miller County, Arkansas. Such striking artifacts are typically found only in graves thought to be those of important leaders. The bowl of this stone pipe is visible on bird's back. A reed or bone pipe stem would have been inserted into an intersecting hole that exits through the bird's back. Courtesy Pictures of Record, Inc.
Crenshaw site
Mound F at the Crenshaw site as it appeared in the late 1960s. This large mound may be a platform mound upon which Early Caddo temples were built. C.B. Moore dug into it in several places in 1912 searching for burials, which he did not find. Courtesy Frank Schambach.
map of Crenshaw site
Map of Crenshaw site, Miller County, Arkansas. Courtesy Frank Schambach.
Crenshaw site today
Crenshaw site today looking across the main site area toward one of the mounds. The landowners of this portion of the site protect it by, among other means, maintaining it as an pecan orchard. Photo by Steve Black.
drawing of mass burial
Schematic sketch of late Fourche Maline mass burial of 14 individuals neatly laid out in a rectangular pit beneath Mound B at Crenshaw. This 1935 drawing was done by Glen Martin of Texarkana who was employed by Judge Lemley to run the dig. The two intruding pits are later, early Caddo shaft tombs that were dug through the mass grave. Courtesy Frank Schambach.
Crenshaw site map
Crenshaw site map showing the location of mounds, middens, cemeteries, and the Plaza of the Skulls, so named because of the finding of numerous buried clusters of human skulls. Courtesy Frank Schambach.
Caddo Antler Temple
Artist's reconstruction of the early Caddo Antler Temple at the Crenshaw site. Courtesy Frank Schambach.
Davis site map
The Davis site occupies a flat terrace that overlooks the floodplain of the Neches River. When Anglo settlers moved into this area of east Texas in the mid-1800s, the site had been abandoned for at least 500 years. They named the open, flat area "Mound Prairie." Graphic by Ken Brown.
Mound A
Mound A under excavation in 1940 by WPA crews under the direction of Perry Newell. The outline of one of the large special structures has been left in place while the excavations continued deeper around it. TARL archives.
Mound C
Mound C under excavation in 1969 by archeologists from the University of Texas at Austin under the direction of Dr. Dee Ann Story.
geophysical survey
Geophysical survey in progress, 2003 by archeologists from the University of Texas at Austin under the direction of Drs. Darrell Creel and Samuel Wilson. Here the archeologist is dragging a sensing device back and forth across the site at regular intervals. The sensors mounted on the lightweight wheeled framework detect subtle differences in the earth's near-surface magnetism. Magnetic disturbances created by pits and postholes dug by the ancient Caddo some 800 years ago can be detected with this equipment. Photo by Steve Black.
daily life at the Davis site
Daily life at the Davis site as depicted by artist Nola Davis. Residential houses would have had nearby outdoor work areas, shade ramadas, raised granaries, and cooking pits, surrounded by cornfields and gardens. Courtesy Texas Parks and Wildlife Department.
Mound B
Mound B covered the site's largest building, a circular structure some 18 meters (59 feet) across, and probably several other buildings. The large building was dismantled and burned before being capped by Mound B. Additional structures may have been built atop the mound. Graphic by Dee Ann Story.
Davis radiocarbon dates
The 130 radiocarbon dates from the Davis site show that the main occupation period was between the mid-9th century through the early 14th century, a span of over 450 years. Graphic by Dee Ann Story.
geophysical map
Section of geophysical map showing a small area between Mounds A and B in which at least four structural patterns are readily discernable. The orange square highlights a circular pattern of a building about 11 meters (36 feet) across with four internal roof supports and a hearth in the middle. The angled line running through the image is one of the signature of an interpretive footpath at Caddoan Mounds State Historic Park. (The right-angled glitches are minor data/processing gaps.) Courtesy Darrel Creel and Dale Hudler, TARL.
colorized printout of geophysical data
Colorized printout of geophysical data centered on the same building (F237) with four internal roof supports and a hearth in the middle. This was used by the field crew in May, 2003 to guide excavations. The red-outline shows the excavation area, while the labeled internal features (internal support posts and hearth) are designated F237.1, F237.2, etc. Photo by Steve Black.
uncovering the hearth
The building pattern proves to be just what was expected. Here Darrell Creel takes small soil core samples from the dark stain of Building F237's central hearth (F237.4). The samples will guide the excavation of the hearth. Photo by Steve Black.
stone celts
Polished stone celts such as these were included as grave offerings in Mound C and were also found within postholes. They were used as axes, but some of those found as offerings are so small they do not look like functional objects. Davis site. TARL archives.
"Ceremonial Maze"
The "Ceremonial Maze"(Building F35) as outlined by WPA excavators. View looking west across the "wings" thought possibly to represent those of a turkey, a special bird to the Caddo. The orientation of the narrow extended entranceway (or "neck") on the right (north) appears to have shifted a few degrees during remodeling, making it look as if the entrance was blocked. The raised ridge of baked clay on the opposite "tail" end is the result of the intentional burning of the structure when it was dismantled hundreds of years ago. TARL archives.
building F9
Building F9 during the WPA excavation. Over 4,600 pottery sherds, a piece of copper attached to a cord, fragments of three clay pipes, a ceramic ear spool, a ceramic effigy, and various stone tools and flint flakes littered this unique semi-rounded building at the Davis site. Some of the artifacts are visible in this photo. TARL archives.
Bujilding 125
Building 125 during excavation. Thought to be a fire temple, the intensively burned deep central hearth had yet to be excavated when this 1970 photo was taken, yet its location is obvious. The two large holes nearby are where two of four large posts stood around the hearth (the postholes have been cleared by the excavators). TARL archives.
plan map of building 125
Plan map of Building 125 and cross-section of its deep, multi-layered central hearth. Graphic by Dee Ann Story.
excavation walls
The excavation walls in the central area of Mound C revealed a complicated sequence of cut and fill layers. The intentional use of contrasting soil colors by the ancient Caddo probably had symbolic meaning. It also allowed the archeologists to unravel the sequence of events. TARL archives.
Plan map of "log tomb"
Plan map of the "log tomb" (Feature 134) so named because of the two rows of massive log posts that ran down the walls of the tomb and presumably supported a low roof only about three feet above the tomb flour. The construction of the log tomb is inferred from stains and traces as no wood survived. On the floor of the log tomb were eight individuals in extended positions (heads to the north) and neatly arranged into four pairs. Graphic by Dee Ann Story.
incised stone earspool
Front and back of a set of incised stone earspools. Mound C, Davis site. TARL archives.
small tomb
This small tomb is typical of many of those under the outer part of Mound C. It measured about 7 by 5.5 feet and contained a single engraved pottery bottle as a grave offering. No bones were preserved, probably because the burial pit was relatively shallow (8 feet deep) and on the edge of the mound. Graphic by Dee Ann Story.
floor of feature 119
Floor of the cane tomb. (Click to enlarge.) As can be seen, the four individuals were arranged in extended positions in two pairs. The pair on the right was obviously paramount (most important) as they were widely spaced with arms outspread (the legs were not exposed by excavation), heads placed facing up, and accompanied by high-status grave offerings. In contrast, the other pair of individuals was closely spaced, with arms to the sides, had few grave offerings, and had been arranged so that the heads faced east, toward the paramount pair. The lesser pair (perhaps slaves) were probably sacrificed in honor of the paramount individuals (of whom the one on the far right was accompanied by the greatest amount of offerings and personal ornaments.) Graphic by Dee Ann Story.
layer of river cane
Drawing showing the layer (or coppice) of river cane that may represent the symbolic roof of the tomb. Graphic by Dee Ann Story.
cluster of grave offerings
This cluster of grave offerings on the north side of the log tomb included a celt, a marine shell "dipper," and an upside-down ceramic bowl. TARL archives.
gahagan bifaces
Gahagan bifaces—extremely thin knives made of high quality flint—left as offerings in the cane tomb. TARL archives.
huge tomb
This large tomb in Mound C contained a single young adult resting on his or her (sex not determined) back with arms and feet spread out. Around the waist was heavy belt made of tubular marine-shell beads (item 9). At the right knee was a greenstone scepter. On either side of the head were copper-covered stone earspools apparently held in place by a pearl-beaded band that went over the top of the head. Frank Schambach thinks this individual is dressed and positioned as the Bird-Man, a suggestion we find compelling. Graphic by Dee Ann Story.
early and middle Caddo sites
Shown here are some of the better-known and important Early and Middle Caddo sites dating to roughly A.D. 1000-1400. Map by Dee Ann Story.

Special Buildings and their Earthen Mounds

The three earthen mounds at the Davis site are visible reminders of the Caddo community and of the generations of villagers for whom the site must have been the sacred center of their world. The mounds, which were built at different times during the site's history, served various functions, all of them important to the life of the community. Mounds A and B were begun as low "capping" mounds that purposefully covered (capped) the floors of dismantled temples and communal buildings (collectively, "special buildings"). Then the top of Mound A (and possibly that of Mound B) served as a platform upon which new special buildings were built. Platform or temple mounds are one of the hallmarks of Mississippian life. Burial mounds like Mound C had been around since early Woodland days if not before.

Each earthen mound at Davis was built in stages, probably during short periods of time when the community pitched in for the common good. The villagers dug up the soil from areas below the low bluff that borders the wide, flat terrace upon which the site sits. A large depression left by one of the barrow pits where soil was removed can still be seen today after 800 years of gradual infilling. Basketload after basketload of soil was carried up the bluff, brought to the mound site, and dumped. This was not a casual process, but a very deliberate one done with care. Often the villagers selected layers of soil having differing colors or textures for particular mound construction layers, particularly in Mound C, the burial mound. There, it is clear that certain colors must have be chosen for their symbolic value. For example, a special green clay obtained elsewhere in the Neches Valley was used to cover certain of the shaft-like tombs that held the bodies of very special members of the community. While we cannnot decipher the specific meanings of the soil colors chosen a thousand years ago, the patterning is clearly purposeful and must have had deep symbolic value.

Mounds A and B mark the inner precincts within the Davis community, a pattern than began before the mounds were begun and one that continued for generations. Within these precincts stood tall and often spacious beehive-shaped wooden buildings with grass-thatched roofs that were larger than ordinary houses. Some were probably the houses of community leaders and their close kin. Others were temples where priests or shamans lived and kept sacred paraphernalia and conducted sacred rites. Still others were probably something like council houses where men gathered on solemn occasion to plot revenge or otherwise deal with crisis, as described by the Spanish many centuries later. Unfortunately, the Caddo rarely left many telltale clues as to what each building was used for. They kept the buildings (and probably much of the village) swept clean. And when an important building had served its purpose, it was very carefully dismantled, its timbers yanked, and everything above the floor was either burned on the spot or taken elsewhere for disposal (or reuse). This is why we use the purposefully vague term "special buildings" to characterize these imposing and obviously important buildings.

At Davis several special buildings stand apart from all others because they are so unusual and so complicated.

Named the Ceremonial Maze by Perry Newell and, less prosaically, Feature 35, Building F35 was uncovered some 11 meters (36 feet) south of Mound A. Its outline is most curious, consisting basically of three concentric, discontinuous wall trenches, one of which was partially offset from the others. An apparent extended entranceway faced north and there was a hearth in the center of the building. Strangely, the innermost trenches and two of the exterior ones had no evidence of posts. All the other trenches had post hole patterns or the actual remains of burned posts.

The two intersecting entranceway patterns on the north side are enigmatic. As described by Newell, they "gave the impression that they might form an entranceway but the narrowness of the passageway between the posts makes this unlikely, although possible. The idea that they might have formed two entranceways for two successive structures seems unlikely as … the uniformity of the whole patterns makes it appear a single unit." (From Newell and Krieger, 1949.)

We think it more likely that two building stages are represented and that the entranceway and perhaps other parts of the building were remodeled (and slightly reoriented) during its life. Regardless, Building F35 is probably some sort of religious temple, perhaps built in the shape of a bird. Bird-shaped buildings and mounds are known elsewhere in the Eastern Woodlands. Use your imagination and look at the plan map of this building. Think about a turkey, a favorite bird of the historic and the living Caddo peoples. The entranceway at the north is the bird's neck, its wings are spread east and west and its tail fans out to the south. Keep in mind that you are looking at only the foundation pattern. If this really was a bird-effigy temple, the roof, walls, and perhaps carved wooden sculptures would have fleshed out the design, so to speak. You might think this notion far-fetched, but Building F35 is most definitely a strange bird.

Newell's "Ceremonial Maze" may not be too far off the mark. We speculate that the narrow passageways and dividers within the room may have been used for ritual performances with costumed dancers entering from the sides unexpectedly at key moments. The trenches lacking posts may have been covered with wooden planks to form foot drums. If so, lines of performers keeping beat with their feet would have created quite an otherworldly din within the smoky and dimly lit confines of the interior of the bird temple. Or the drama may have unfolded on the Mound A platform overlooking the building, an otherworldly place from which religious practitioners/performers emerged. Whatever the actual purpose of this building, it is unlike any other known at the site and, for that matter, unlike any other known in the Caddo world. It was, indeed, a "special building."

On the opposite side of Mound A was another intriguing, if not quite as unusual, structure. Building F9 measured about 11 meters (36 feet) across and stands out for a number of reasons: (1) instead of being circular, it is squared off with rounded corners (about 11 meters across; (2) it was built with a shallow depression 14 inches deep (other structures were built on the ground surface); (3) the exterior wall posts were set in a trench (the posts of ordinary houses at Davis were always set in individual holes, the wall trench is a Mississippian technique that is uncommon in the main Caddo Homeland, but often used in the Arkansas Basin); (4) there were five interior posts set at regular intervals around the circumference of the building (these may have supported benches); (5) the floor was formed by the addition of a thin layer of clay (prepared floors were rare at Davis); (6) the extended entranceway was higher than the floor level; (7) the structure was burned, yet not completely cleared away, as was the usual practice at the site; and (8) a great many artifacts were found within the building, most of which appeared to have been placed there after the building was burned.

Within Building F9 were over 4,600 pottery sherds, a piece of copper attached to a cord, fragments of three clay pipes, a ceramic ear spool, a ceramic effigy, and various stone tools and flint flakes. Only one other excavated Caddo building has been found to have such an array of artifacts near it and that is the apparent fire temple at the A.C. Saunders site, a Late Caddo center some 57 km (35 miles) upstream on the Neches River. While the small hearth in Building F9 does not seem to fit the notion of a fire temple, the building was obviously something special. The purposeful depositing of debris, including broken ritual items, suggests that is was a conscious and presumably meaningful disposal of artifacts in the building site, perhaps following a ritual feast.

There is a better candidate for a fire temple at the Davis site. Building 125 was uncovered about half way between Mounds B and C. It had a simple, circular outline about 7.4 meters (24 feet) in diameter. In the center of the building was a deep and intensively used hearth that had been refurbished at least twice with new clay linings. This hearth is far more complex and intensively burned than any other known from the site. Another unusual feature is that four large interior posts were set at regular intervals around the hearth, presumably to support the roof. Interestingly, the posts were not placed on cardinal directions. This seems to fit the early historic descriptions of the Hasinai fire temple rather well:

The perpetual fire, the distinguishing characteristic of the fire temple, was fed by four large, heavy, logs arranged in cruciform pattern so that the arms coincided with the cardinal points of the compass. In the center of the floor, where the four logs abutted, the flame was kept alive by feeding in small firewood from piles that were maintained outside the structure. The ashes from the fire were removed and were allowed to accumulate in a heap outside. (Griffith 1954)

Special Cemetery

While is accurate to call Mound C a "burial mound," it functioned as a special cemetery for certain privileged ("elite") members of the Davis community and others who were singled out to accompany the elite in the afterlife. We are not sure where the ordinary villagers were buried, but it was not here. This special cemetery was established early in the site's history and periodically refurbished and reused for the next 300-400 years. The mound itself was built over the first pre-mound tomb as a flat-topped mound resembling those built to serve as platforms for temples. Once built, the mound was periodically enlarged and resurfaced, never standing any taller than about 20 feet (6 meters). Although the flat-topped appearance was maintained throughout its long period of use, no trace of any crowning structure was detected and it is likely that none was ever present.

Only about a third of Mound C was excavated, and within this area 11 burial pits were investigated in 1968-1970 by University of Texas crews led by Dee Ann Story. It is estimated that 15-20 burial pits remain in the unexcavated part of the mound. While most of the pits contained no more than one individual (or, more often,had no preserved bones), three large tombs had multiple interments (one with eight individuals and two with four). In contrast to the pattern at the Crenshaw site, none of the later tombs in Mound C at the Davis site significantly intruded into earlier tombs. (This suggests a continuity of community, a continuity that was perhaps broken at times during the history of the Crenshaw site.)

In the center of Mound C were relatively large and complex tombs with elaborate grave offerings. Around the outside of the mound and surrounding the pre-mound tomb, most of the tombs were smaller and contained far fewer grave goods. This pattern suggests that the most important individuals were buried in the central area with less important people buried in the outer mound. These "less important" individuals were still buried in a special place in relatively large tombs and were probably not ordinary members of the community.

We will not discuss each of the excavated tombs at Davis. The accompanying pictures provide glimpses of some of the variation. Instead we will briefly describe a few contrasting examples and then summarize some of the important patterns and unusual finds. The special cemetery at Davis began with the construction of a single, massive log-roofed tomb containing eight individuals and an array of fancy grave goods. This "log tomb" (designated Feature 134) measured about 23 by 18 feet (7 by 5.5 meters) and was dug 11 feet (3.5 meters) into the original earth. Two parallel rows of three large postholes ran along the north and south walls (long axis) of the tomb and presumably supported a low roof only about three feet above the tomb flour. The construction of the log tomb is inferred from stains and traces as no wood survived.

On the floor of the log tomb were eight individuals in extended positions (heads to the north) and neatly arranged into four pairs. The heads of the pairs at either end were turned to face inward toward the center of the tomb. One of the central pairs faced toward the other central pair; the latter includes the paramount (most important) individual to judge from the arrangement of grave goods. Because of the very poor condition of the bone, the age and sex of the many of the individuals is not known. Of those that could be discerned, most were adults 20-30 years of age and most were males, but at least one was a female and another a child of about 6 years.

Most of the offerings were placed on the north side of the log tomb, beyond the heads of the deceased, in clusters or piles. Among the numerous grave offerings were conch shell drinking cups, stone ear spools, stone celts (axe heads), wooden objects covered in thin sheets of copper, various patches of colored mineral pigments, effigy pipes, and arrow points. Many of these items were of exotic materials obtained hundreds of miles away. Interestingly, there were no pottery vessels in this first tomb. Although there are finely engraved vessels in some of the later tombs, these are found in small numbers that pale by comparison to the large numbers of pottery vessels (sometimes dozens) found in elite burials at other Caddo sites, especially those along the Red River.

The apparent paramount individual in the log tomb at Davis had a beaded waist belt made of marine shell (probably whelk obtained from the Gulf coast) and an exceptionally large, beautifully made chipped-stone knife (sword, really). This spectacular artifact is almost certainly a prized symbol of authority and rank. It is almost 19 inches (480 mm) long, yet just over a half-inch thick (15 mm) at its thickest point, making it so delicate that it was broken by the weight of the soil when the tomb roof collapsed. It is made of an almost white exotic chert (flint) possibly obtained in the Midwest from the Mill Creek area of southern Illinois. Other somewhat similar, but much smaller knives were found in the clusters of offerings on the north side of the log tomb at Davis.

When the log tomb was sealed, the extra earth from the pit was neatly piled around it, forming a low ring-shaped berm. The berm was constructed in a deliberate sequence of colored earth that was the reverse of the natural stratigraphy. In other words, the red surface soil was placed at the bottom, then a layer of orange soil, and finally a cap of yellow soil, the site's deepest. The roof of the tomb was covered with a thin layer of an unusual greenish (glauconitic?) soil, then the soil removed from the burial pit. Enough time elapsed for the tomb's roof to decay and collapse, spilling earth into the tomb and doubtlessly crushing some of the skeletons and delicate offerings. The collapsed tomb was then covered over (capped) by several layers of fill, creating the initial flat-topped mound.

At least three burial pits were dug into and extended below the mound at this stage. Subsequently, the first mound was enlarged and capped by more fill layers, renewing the flat-topped mound. More burial pits were dug from the top and sides of this mound and then it too was renewed by a new cap. All burial pits extended into the premound soil, suggesting that reaching the original ground surface was ritually important. This pattern of burial use and mound capping/addition was repeated at total of five times over a 300-400-year period.

In contrast to the log tomb, some burial pits, particularly those dug into the outer edges of the mound were much smaller and did not contain many offerings. For instance, one burial pit under the south edge of Mound C (Feature 156) measured only about 7 by 5.5 feet and contained only a single engraved pottery bottle. No bones were preserved, probably because the burial pit was relatively shallow (8 feet deep) and on the edge of the mound.

The most complicated tomb (Feature 119) was located under the central mound. It measured about 22 by 16 feet and its floor was 16 feet beneath the mound's surface. We can call this the "cane tomb" because there was a layer of cane (or coppice) above the four skeletons. (Woven cane mats or bundles of cane were probably placed on the floors of all of the tombs.) The cane layer was the roof or ceiling of the tomb, perhaps symbolically equivalent to the thatched roof of a house or temple.

The four individuals buried in the cane tomb had been arranged in extended positions in two pairs. One pair was obviously paramount; these individuals were widely spaced with arms outspread (the legs were not exposed by excavation), heads placed facing up, and accompanied by high-status grave offerings. In contrast, the other pair of individuals was closely spaced, with arms to the sides, had few grave offerings, and had been arranged so that the heads faced east, toward the paramount pair.

Like the log tomb, the cane tomb had concentrations of grave offerings on the north side of the tomb, above the heads. There were hundreds of individual items ranging from stream-worn pebbles (from a rattle?) and lumps of pigment to finely engraved bottles and human effigy pipes. The artifacts had been purposefully placed in two layers, one above and one below the cane layer capped by a thin layer of green clay. The paramount individuals had copper-covered wooden or stone earspools, bone pins, a large ceremonial knife, pearl and shell beads, traces of woven mats beneath some offerings (and perhaps the bodies), and copper stains, one of which was once a copper hair ornament, among other items. (Only half of the tomb was excavated; many more offerings were probably present.)

But how do we explain the multiple skeletons? In each tomb, the bodies were clearly interred during the same event while the bodies were still intact, leaving us to explain the coincidental deaths. Because of the lack of obvious signs of warfare or disease and because the tombs often contain individuals of different ages and different sexes, including many young adults, we suspect that most did not die a natural death. We suspect that the larger tombs contain a paramount individual who died a natural death and was accompanied by retainers or family members who were immolated (put to death) to honor the deceased and accompany him (or her) in the afterlife.

Such behavior is well documented in the Southeast by archeological finds and, to a lesser degree, by early historic accounts. In 1725, the Frenchman Le Page du Pratz witnessed the ritual strangling of the wives, servants, and others to accompany the deceased brother of the Great Sun, the chief of the Natchez. Although it goes against our religious and civil sensibilities today, other forms of human sacrifice are well documented in prehistoric and contact period North America as well as in Mesoamerica, most infamously among the Aztecs (not to mention in Africa, Asia, and medieval Europe.) In North America, slaves taken from other tribes were chosen for sacrifice and it is possible that some or even all the retainers in the Davis graves were slaves.

We think it more likely, however, that most of the individuals buried in Mound C were members of the Davis community. The repeated pattern of episodes of multiple interments can be speculatively linked to episodes of renewal and rebuilding at Mound A. There, evidence was found of the periodic dismantling of special structures, followed by capping of the dismantled buildings and the building of new structures, often of about the same size and in the same position as the just-buried building. Such cycles of renewal may be part of periodic ceremonies tied to some sort of ritual or astronomical cycles or perhaps to the deaths of paramount leaders. Unfortunately, the excavated skeletons at Davis were poorly preserved, rendering it impossible to make the detailed observations and comparisons that might confirm or deny our suggestions.

Several intriguing patterns hint at the nature and meaning of the ritual behaviors represented by the special cemetery at Davis. The exact meanings are lost to the past, but the repeated patterns known from the Davis site and others are telling. One pattern is the arrangement of the skeletons within the tombs with multiple individuals. In most cases where the skulls were preserved, the heads of the presumed retainers/slaves faced toward the paramount individual. The paramount individuals had exotic artifacts obviously symbolizing authority and high rank such as the large flint knives, stone earspools sometimes covered in copper, shell beaded belts, and, in two cases, polished stone staffs made of exotic greenstone (these are often called "spuds," an unfortunate, but memorable choice of terms). In the earlier graves most of the offerings are piled on the north side of the tomb, above (north of) the heads of the deceased.

Another fascinating pattern is the purposeful selection and placement of layers of earth having certain colors and textures. Mound C is composed of dozens of cut and fill layers representing the digging and filling of pits and the addition of layers capping and even repairing the mound. Archeological trenches cut through many of these layers and showed both how complicated the sequence of events was and how purposeful the selection of fill was. With so many disturbances and intrusions, one would expect the mound fill would be jumbled and homogenized. That sometimes occurs, but often specific layers or fills were composed only of a certain soil with a markedly distinct color. One especially significant earth at Davis is a green glauconitic soil that may have been obtained from natural exposures elsewhere in the Neches River valley (such exposures are not known to occur in the site vicinity). A thin layer of this greenish gray soil was placed above the high status burials and almost nowhere else. (Newell reported finding an "altar" made of green clay on one of the Mound A platforms, but it was poorly documented.) The purposeful use of certain soils for certain layers suggests that the colors and perhaps the contrasts between colors had symbolic importance.

The very deliberate and carefully arranged patterns represented by a single large tomb can be thought of as a sacred scene designed to symbolize, justify, and affirm the legitimacy, importance, and authority of the leader whose death occasioned the tomb. Such sacred scenes may well have embodied mythological stories such as origin myths and the central characters in those stories. These scenes must have reflected important Caddo icons in much the same way that Christians use the cross and the fish, as well as biblical characters and saints. For example, the pattern of pairs of bodies seen in two of the tombs may relate to Caddo stories recorded many centuries later that feature a pair of mythical twins. According to anthropologist Robert Hall, twin-related myths are known from many different groups in both North and South America and may have been part of the oral tradition of the earliest people in the New World.

Another character common to much of the Mississippian world is often called the Bird-Man, a central icon in the Southern Cult or Southeastern Ceremonial Complex. The Bird-Man is usually a human figure (male) dressed in a feathered-wing costume with a bird mask clearly representing a raptor, probably the peregrine falcon. Variations on the Bird-Man theme are known from many engraved shell gorgets and drinking cups, on copper repoussé (raised relief) plates, and occasionally on pottery or stone artifacts from Etowah, Spiro, Moundville, and sites in eastern Tennessee, among others. In some depictions, the human and bird elements are combined such that the figure is neither human nor bird. Bird-Man or Man-dressed-as-Bird, the figure usually has arms and legs akimbo, as if dancing. Bird-Man usually looks fierce and warlike, sometimes brandishing weapons or grasping severed human heads. Another common element is a wide, beaded belt around the Bird-Man's waist. In sum, the Bird-Man is probably a mythical being associated with conflict and power, but one that was frequently personified in ritual dance at Mississippian centers.

The Bird-Man personification may also be represented by several of the paramount individuals in the major tombs at the Davis site. The clearest examples come from a massive burial pit (Feature 118) measuring (at the top) 29 by 24 feet (9 by 7.5 meters) and dug down about 23 feet (7 meters) from the top of Mound C. This huge tomb contained only a single young adult (sex not determined) resting on his or her back with arms and feet spread out. Around the waist was a heavy belt made of tubular conch shell beads. At the right knee was a greenstone scepter. On either side of the head were copper-covered stone earspools apparently fastened to one another by a pearl-beaded band that went over the top of the head. Four clusters of 24-40 arrow points surely represent quivers of arrows placed around the body. The arrow points were arranged into clusters of different types of flint (chert) including some from central Texas and others from eastern Oklahoma. Other offerings include bits of copper, green pigment, shell, and possible bark cloth around the skull that may represent some sort of headdress, a pearl necklace, and a small flint knife. Frank Schambach thinks this individual is dressed and positioned as the Bird-Man, a suggestion we find compelling. Individuals in several other tombs at Davis may also represent the Bird-Man character.

While we tend to assume that the tombs and graves in Mound C represent burial rites following the natural (or battle) deaths of leaders, this is hard to demonstrate. Given all the ritual patterning and the episodic nature of destruction and renewal cycles in Mound A, it is possible that we are wrong and that the triggering event was not the death of some important person. An alternative possibility that we do not rule out is that all of the individuals interred in the special cemetery at Davis may have been sacrificed during special ceremonies tied to either major events in the community's history, crises (such as severe droughts), or perhaps to celestial events (such as the alignment of Venus and Mars or a major eclipse). In other words, the ritual/astronomical event itself triggered death (sacrifice), not the other way around.

Demise of the Davis Community

We do not know exactly when the Caddo community who lived at the Davis site decided to move on. The community continued into the Middle Caddo period, but was in decline by the late13th century. The latest radiocarbon dates that make sense suggest that some activities continued into the 14th century (early to mid-1300s), well over 450 years after the community was established. It is clear that the abandonment of the site was planned and orderly.

The best evidence of this is that Mounds A and C (and probably B, although the evidence there was destroyed by plowing) were each capped with a final layer of clay that sealed the mounds. At Mound A, the final buildings were first dismantled, just like earlier buildings, and then a substantial layer of homogenous clay was added that physically capped the mound and symbolically sealed its contents. This same pattern has been seen at many other Caddo mound centers. When a Caddo community abandoned a place, ordinarily they did so with forethought and purpose.

We do not know where the community moved. One candidate is the Washington Square site about 30 miles due east within what is now downtown Nacogdoches, Texas. The heyday of that site began in the mid-13th century, not long before construction and activity at the Davis site began to wane. Perhaps prestige and power of one ritual/political center was eclipsed by another, as must have happened many times in Caddo history.

Other Early and Middle Caddo Sites

Hundreds of known Caddo sites have yielded pottery of Early and Middle Caddo styles, which are often very similar. Some of the better-known sites are shown on the accompanying map. Many of the same Early Caddo patterns seen at the Davis site and the Crenshaw site are present or suspected at many of these sites. Ritual centers, mounds capping special buildings, buildings built on mound platforms, special cemeteries, shaft tombs, evidence of social ranking, use of ritual items from distant sources, and so on. Most of the Early Caddo sites that have seen investigation are the larger sites with mounds. We know much less about small, non-mound sites. In the next section we take a closer look at a small Middle Caddo village and the continued development of Caddo society.

stone earspools
Stone earspools once covered by thin sheets of copper. These were worn covering the ears, held in position by cords, traces of which were found in one instance. They are only found in tombs and other ritual contexts and were presumably worn as status symbols.
drawing of building F35
Drawing of Building F35, dubbed the "Ceremonial Maze" by Perry Newell. The complicated outline of this highly unusual and enigmatic structure resembles that of an outstretched bird, perhaps a turkey (some imagination required). We think it is a specialized religious building within which religious performances took place that may have featured dancers/characters entering and exiting the scene through the narrow entranceway and internal passages. The trenches lacking postholes may have been covered by wooden planks to create floor drums. Graphic by Dee Ann Story.
Building F9
Building F9 was an unusual building in many ways including its squared-off shape with rounded corners, regularly spaced interior posts that may have supported benches, and the mass of debris, including broken ritual items, found on its floor. It looks like the building was "ritually trashed" after it was dismantled and burned. (Also visible is Building F29, an earlier structure built and dismantled before Building F9 existed.) Graphic by Dee Ann Story.
central hearth
Partially excavated central hearth in Building 125 (cross-sectioned posthole in background). The intense burning that took place in this feature is readily apparent. This hearth may have held a "perpetual fire" like that in the Hasinai Fire Temples seen by the Spanish in the late 1600s. TARL archives.
mound c
Mound C was a special cemetery for certain privileged ("elite") members of the Davis community. It was established early in the site's history and periodically refurbished and reused for the next 300-400 years. The deceased were placed into burial pits (tombs) that were dug deeply beneath the then-existing surface into the original ground. Once built, the mound was periodically enlarged and resurfaced, never standing any taller than about 20 feet (6 meters). Graphic by Dee Ann Story.
Elton Prewitt
Archeologists Elton Prewitt and Dee Ann Story clean the floor and wall of the initial excavation pit into Mound C in 1968. Sharp trowels were used frequently to cut fresh exposures and trace out the complicated layers of the earth mound. Early morning and late afternoon lighting often revealed subtle differences that could not be seen in the glaring mid-day sun. TARL archives.
finely made arrow points
Finely made arrow points were found in several of the Mound C tombs at the Davis site. Often they were in tight clusters suggesting that they were originally placed in quivers made of perishable material. TARL archives.
This spectacular artifact from the log tomb at Davis is almost certainly a prized symbol of authority and rank—a sword, really. It is almost 19" (480 mm) long, yet just over a half-inch thick (15 mm) at its thickest point, making it so delicate that it was broken by the weight of the soil when the tomb roof collapsed. It is made of an almost white exotic chert (flint) possibly obtained in the Midwest from the Mill Creek area of southern Illinois. TARL archives. Click to see both sides
plan map of feature 119
Map showing location of Feature 119, the cane tomb, and schematic cross-section showing the relationship of this tomb to others nearby. Notice that none of the tombs significantly intrudes into the others. Graphic by Dee Ann Story.
feature 119
Cane tomb (Feature 119) during excavation, 1969. The excavation and recording of this complicated tomb took several weeks of tedious work as each item was carefully mapped and photographed in place before removal. TARL archives.
clusters of grave offerings
Clusters of grave offerings along the north wall of the "cane tomb" (Feature 119). TARL archives.
cane impressions
Cane impressions left by layer (or coppice) of river cane that formed the roof of the cane tomb. TARL archives.
Drawing of "bird-man" depicted on a repoussé copper plate from the Etowah site, Georgia. In this example the bird-man seems to be a human dancer wearing a falcon costume. A human head appears to be dangling by its scalp from the bird-man's left hand. The bird-man also wears a tubular shell bead waist band ( highlighted in red), very similar to that found in tomb F18 at the Davis site.
adding a final layer of clay
Artist's depiction of the addition of a final layer of clay to Mound A at the end of the life of the Caddo community at the Davis site, sometime around A.D. 1300. The clay layers physically and, ritually sealed the mounds and the important temples and buildings that once stood here. Painting by Nola Davis, courtesy Texas Parks and Wildlife Department.
Davis vs. Chupik pottery designs
Striking parallels in pottery designs between the Davis site (left) and Chupik site (right), which is located some 130 miles to the west on the Brazos River west of Waco, Texas. TARL archives.

While most Early Caddo sites occur within the Caddo Homeland, several outlier sites are known that appear to be seasonal farmsteads or temporary camps of Caddo groups far from home. One good example is the Chupik site on the Brazos River just west of Waco, Texas. We do not know why the Caddos camped there, but close parallels in artifact styles between items found at Chupik and at the Davis site suggest that they did. That coupled with the relatively large quantity of refuse, leads us to suspect that Chupik was seasonal farmstead or perhaps a short-lived attempt by a group of Davis-site Caddos to colonize a distant area.

Follow Caddo History

Davis vs. Chupik stone tools
Davis site stone tools on the left are very similar to those on the right from the Chupik site near Waco, Texas. TARL archives.