Hank's House 1

The prospect of investigating Hank's house was exciting because its well-preserved remains offered an opportunity to learn a great deal about the architecture of a Plains Village pithouse. We wanted to "salvage" the information before the next big flood completed the erosional process and washed away the remaining half of the house. Such floods are highly unpredictable—it might be 20 years or 24 hours before a real Panhandle gully washer stalled over the upper end of the West Pasture and remodeled the landscape, sending Hank's house tumbling down the streambed. So we decided to conduct a salvage excavation as soon as possible. But first we had to find a crew and solve some logistical problems, the heaviest of which (literally) was removing the thick mantle of up to two meters (seven feet) of sand covering the pithouse.

Finding a willing crew proved easy enough once Doug Wilkens and Doug Boyd spread the word of what we had found. The Panhandle has a small, but active group of experienced avocational archeologists, most of whom are members of the Panhandle Archeological Society and the Texas Archeological Society. We also persuaded several professional archeologists active in Southern Plains archeology to take time "off" and join us. Meanwhile, Erickson took care of the logistics including establishing a field camp at the site where we could store equipment and preparing an excavation platform along the creek cutbank.

We returned to the M-Cross Ranch in late November, 2000, with a sizable contingent of volunteers. Before most arrived, landowner John Erickson used his small bulldozer and Bobcat to remove the overburden (overlying sand), stopping a foot and a half (40-50 centimeters) or so above the pithouse layer. Over the next week, the volunteer crew put in 406 hours of labor as we exposed most of the remaining half of the pithouse and several of the associated features. We had hoped to finish the dig in one session, but couldn't quite do it. In mid-May, 2001, we returned with a smaller crew for a three-day dig to wrap things up. The rest of this section explains what we found.

Sliced in Half

Hank's site is formally known as 41RB109, being the 109th officially recorded site in Roberts County, Texas. The unnamed creek through the West Pasture had, in essence, given us the kind of view that archeologists often like to start with—a long cross-section or slice through Hank's site. The map on the left shows the area we excavated and the archeological features we identified along this slice. Before discussing the prehistoric house proper, let's look at some of the other features we found outside the house. Some of these were very close to Hank's house and very likely "functionally associated," meaning they were probably created and used by the occupants of the house. But others are farther away and were probably part of another household—without additional excavation we can't really tell.

Hank's house faces east, like many Plains Village houses, and has an extended entranceway (see house diagram on right). We excavated around the front of the house and found several noteworthy features. In front of Hank's house was a trash midden, a thin sheet-like layer of household trash —scattered bits of artifacts, such as flint flakes and pottery sherds, mixed with dark, charcoal-stained soil. Within the midden were three distinct groups of objects (archeologists call such groupings cultural features), all of which indicate that the area in front of the house—a sort of "front porch"—was used not only for discarding trash, but also served as a storage area and probably a work area.

Caliche Cobble Cluster: The first feature was a cluster of unmodified caliche rocks that were probably collected and piled with the intent of being used at a later time.

Mussel Shell Cluster: The second was a tightly packed cluster of six complete mussel shell valves that had been stacked, one inside the other. These, too, appeared to represent materials that were stockpiled for later use.

Grinding Stone Tool Cache: The third feature within the trash midden was a tool cache of eight rocks, seven of which were grinding tools. A "cache" is a group of objects that were intentionally stored, perhaps even hidden. A metate, a pestle, and five different sizes and types of manos were found in a tight cluster and appeared to have been stored in a small pit. These grinding implements made up a diversified tool kit useful for a variety of different functions. The eighth rock was a large chunk of very grainy and poorly consolidated conglomerate. The large rounded sand grains that comprise this rock look very similar to the grains seen in many of the potsherds found on the site. Pieces of this rock may have been crushed up and added as a tempering agent to the clay used for making pottery.

Storage Pits: On the edge (as we perceived it) of the sheet midden was the bell-shaped pit (Pit 3) found just outside the entrance to Hank's house. We found two other bell-shaped pits 12 to 15 meters east of Hank's house. Like Hank's house, all three of these pit features had been sliced in half by erosion. The bell shape of the pits is typical of underground pits used by many different prehistoric and historic Indian groups across much of the Great Plains, and they were typically used by farming peoples for storing crops that were harvested in the fall for use later in the winter. The relatively narrow entrance at the top of the pit was necessary so that the opening could be easily covered by a large flat rock, a flat piece of wood, or perhaps a smoked buffalo hide cover held down by rocks and covered by dirt. The covered pit would have been hidden and protected from most varmints (at least until some lucky rat dug his way into it).

Because Pit 3 was located just past the entryway to Hank's house, it likely served as an underground storage bin for the people who lived in the house. The pit could have been used for many years but probably was abandoned when some food rotted inside it or burrowing rodents dug into it. Once the pit became useless for food storage, the people living in the house probably filled it with household trash, bits and pieces of which we found.

Because Pits 1 and 2 were located some distance from Hank's house, they were probably not associated with Hank's house and may have been used by people who lived in another residential structure nearby. Like Pit 3, Pit 2 was filled with cultural debris (trash) and had been converted from a storage pit to a trash pit. In contrast, the fill inside Pit 1 contained no cultural debris, the sediment fill inside was very clean (there was no charcoal or ashy sediment), and the pit outline was well defined. This may indicate that the last function of the pit was for storage of perishable foods. An attempt was made to identify any organic remains that might have been stored in this pit, but none were preserved.

Architecture of Hank's House

In this section, we take a closer look at what we found in the pithouse and what we think it means. Excavation is a process of discovering evidence, but detailed and accurate records must be kept so that you can study and interpret that evidence after it is dug up. Each layer of fill in Hank's house was carefully excavated and all artifacts and features were plotted on excavation maps. Descriptions of the excavations and features were written, and hundreds of forms and maps were created. Thousands of photographs (black and white, color slides, and digital images) were taken to document every detail the archeologists saw. All of the artifacts were collected in bags marked with precise location information (provenience), and many charcoal and soil samples were taken. All of these steps are time consuming but important because what remained of Hank's house was destroyed in the act of excavating it. A site cannot be reconstructed to reveal the stories of life long ago unless such detailed records are kept and studied. The process is similar to how forensic scientists investigate a crime scene to learn exactly what happened.

During the excavation, we separated the sediment inside the house into three stratigraphic layers called the upper fill zone, middle fill zone, and floor zone. The upper fill zone represents sediment and artifacts that were deposited inside the depression created after the house was abandoned and collapsed. The sediment was mainly eolian sands that were blown in by the wind, and the artifacts were probably thrown into the shallow depression by prehistoric people. We suspect that the depression filled in quickly, probably within a few years of when the house fell down.

The middle fill zone was marked by the appearance of charcoal staining and small fragments of burned clay daub, and these burned materials became larger and denser throughout this zone. Large intact pieces of charred wood and burned daub found in this zone clearly indicated the house had burned. These materials were especially abundant in the western portion of the house, where the burning was quite intensive, but sparse in the east side and southeastern corner. It soon became clear that the debris represented the materials from the roof of the house that had burned and collapsed. Some of the burned wood represented long segments of tree branches, while many of the burned daub fragments had parallel stick impressions indicating that wet clay had been pressed onto the inside of the roof. The burned daub was most abundant in the center of the house, indicating that lots of clay was used to line the smoke hole above the central fire pit. The people who lived in Hank's house must have been concerned about the possibility of sparks from the fire pit rising and igniting the roof.

The floor zone represents materials found in the last 4 to 6 inches above the floor of the house. This zone contained abundant charred wood and burned daub, mainly in the western area, and some artifacts. Where the burning was intensive, the floor itself was a thin layer of baked clay sediment, and the excavators could easily follow the floor. In this area, small hollow tubes of charred grass found lying directly on the floor had once been part of the roof. In the eastern part of the house where little or no evidence of burning was observed, the floor was difficult to follow. The unburned clay on the floor had gradually melted away (dissolved) as water percolated down through the sandy sediments over many hundreds of years.

When the excavations ended, we had exposed exactly one half of the original pithouse and the house layout was obvious and familiar. Hank's house was similar in many ways to other pithouses found and excavated in the Texas Panhandle, but it also was unique in many respects. The burning of the structure preserved many important architectural details. Some are typical of prehistoric pithouses of the Antelope Creek and Buried City cultures, others are not. Each of the key architectural features of Hank's house is described below.

Clay-plastered Entryway: Hank's house had a nine-foot-long entryway that was plastered with clay. It began about three feet inside the east wall of the house and extended eastward six feet beyond the east wall. It was essentially a paved ramp that sloped from the original ground surface at its east end down into the house, and the walls of the entrance extended into the house at least two feet. The ramp bottomed out just inside the house, but rose sharply on its west end forming a hump or step that was several inches higher than the floor or the lower edge of the entry ramp. This was obviously a high-traffic area, and multiple layers of clay indicated that the step had been replastered several times. It also formed a depression in the entryway where the bottom of the ramp met the step. Fine, laminated layers of sediment in this depression suggest that it was a water trap that prevented water from running into the house during heavy rains.

West Wall: The west wall of the house was the best-preserved architectural feature because it had been intensively burned. Along the west wall, eight wooden posts had been completely charred and clay wall plaster between them was burned and still in place. Below the floor level there were six post holes, round to oval in shape, found beneath charred posts. The two burned posts that did not have corresponding postholes were resting directly on the floor and had not been set into the ground. The posts ranged in size from 3 to 4 inches (7 to 11 cm) in diameter, and from 8 to 16 inches (20 to 40 cm) in depth below the floor. In cross-section, some charred posts were round and had been complete logs or branches, while others were arc-shaped (half branches) or thin slats that represent split branches. It is quite possible that wood was scarce enough that the builders split some branches to conserve their lumber. The west wall plaster and charred posts were preserved to a height of about 2 feet above the house floor. Most of the posts were burned down only to the floor level, where the fire probably stopped due to lack of oxygen, but a few posts were charred down below the floor level and deep into the postholes. Such burning is not common, but modern ranchers have seen wooden fence posts burn down deep into the ground in some cases.

South wall: Eight post holes were found along the south wall, but only one remnant of a charred post. The post closest to the southwest corner of the house had been burned in place, but only a small portion of it was found above the floor level. It appears that the intensity of the fire was greatest in this corner, but the burned debris found on the floor diminished toward the east. Three large flat chunks of burned clay were found along the south wall near the southwest corner of the house. These appear to have been remnants of clay plaster from the wall, and two of the pieces appeared to be in place along the south wall. The post holes along the south wall ranged from 7.5 to 17 inches (9 cm to 43 cm) deep below the floor level. Based on the diameters of these postholes, the posts were probably about the same sizes as those on the west wall—about 3 to 4 inches (7 to 11 cm) in diameter.

East wall: The east wall was marked by five post holes but no charred posts were found. Remnants of wall plaster were found adjacent to and near the entryway step. The absence of burned posts and scarcity of burned wall plaster indicates that the fire was not very intense toward the front of the house. The east wall post holes ranged from 7 to 12 inches (18 to 31 cm) deep below the floor level. These posts were probably about the same sizes as those along the west wall.

SW Central Post: This was a very large post, set deeply in the ground. This post was very obvious to the excavators because a large chunk of a burned log, measuring nearly 6 inches (15 cm) in diameter, was preserved and stuck up several inches above the floor level. This large post had burned in place, but the burning stopped just below the floor level. When a deep test was excavated through the floor to reveal the profile, the post hole was 10 inches (25 cm) wide at the top and tapered to a point at a depth of about 24 inches (60 cm) below the floor level. The tip of the post had been charred before it was set into the ground.

SE Central Post: We almost didn't find this posthole—a fact that serves as a cautionary note to archeologists digging pithouses like Hank's. As we excavated Hank's house, it seemed to be following a pattern of interior features that was similar to that of other pithouses in the Antelope Creek area. Once the southwest central post was found, it seemed logical that there must be a corresponding southeast central post. But when we reached the floor level at this location, the expected burned post or a posthole wasn't there. Instead, it seemed to be a continuous floor surface. Some burned debris was across the floor in this area, and there were lots of rodent burrows, but there was still no hint of this post hole as we troweled through the floor. We joked that it was obvious the Indians must not have read an archeological report on how their house should have been built.

Puzzled, we kept on digging below the floor with our trowels because we were convinced that there had to be a southeast central post to support the roof. It was not until we had dug a full 4 inches below what appeared to be an intact floor surface that the first hint of a post hole was seen. At this point, we found a circular stain that measured about 9 inches (23 cm) in diameter. We then dug a small pit down below the floor to create a profile and found that the posthole extended down 20 inches (50 cm) below the floor level. Like the southwest central post, the tip of the southeast central post had been charred before it was set into the posthole.

Central Firepit: In the very center of Hank's house was the firepit or central hearth. The hearth was created by digging a large, bowl-shaped pit and lining it with clay plaster. When viewed from above, the hearth appeared as a circular pit ringed by a ridge of clay that was about 5 cm higher than the surrounding floor. The outside diameter of the clay ridge was 20 inches (50 cm), and its inside diameter, the mouth of the pit, was 18 inches (45 cm). When viewed in profile, the hearth had a deep bowl shape and tapered to a rounded bottom that measured about 11 inches (28 cm) in diameter. The pit was 11.5 inches (29 cm deep).

Several layers of fill were evident in the firepit. The layer of burned debris across the floor of Hank's house extended down about 4 inches (10 cm) into the hearth pit. This represented roof fall material from when the house burned. Below that was a layer of brown clayey sand, 4 to 5.5 inches (10 to 14 cm) thick, that looked like wind-blown sand. On the very bottom was a 4- to 5-inch (10 to 13 cm) thick layer of gray ash, presumably from the last time the hearth was used. At the top of this ash layer, a single sherd of cordmarked pottery was found.

The clay layers that lined the hearth were interesting. There was some hint of lamination, suggesting that several layers of clay might have been added at different times. There also was an outer layer of clay that was definitely added at a later time. Several inches of clay had been mounded up to reform the rim of the hearth, and this added layer extended about half way down inside the hearth before abruptly ending.

Central Channel: Hank's house had a distinctive depressed floor area in its center. If the house were complete, it would have formed a rectangle with a central roof support post at each corner. Often called a central channel, the rectangular area around the hearth was about 4 inches (10 cm) lower than the surrounding floor. Such depressed floor areas are common features in many Antelope Creek houses, but its precise function is not known. Archeologists have speculated that it may have helped with airflow around the central hearth, and it might have served as a water trap to keep most of the house dry if water seeped (or poured) inside during heavy rains. The central channel also served as the main kitchen and living room, around which people would do many different activities. The area around the channel would have been used for storage of belongings and sleeping areas.

Post Notch: An arc-shaped notch was found along the southern channel lip, about half way between the southwest and southeast central posts. This notch appears to have been added sometime after the house was initially built, presumably to accommodate the bottom of a vertical post. It may mean that the house was used for several years, at least long enough for the roof to begin sagging and require an additional support post.

Artifacts

The artifacts recovered from Hank's site can be classified into six groups based on how they were associated with the house:

Post-Occupation Fill: artifacts found in the upper levels of sediment inside the pithouse. These materials probably represent trash from occupations of other houses nearby that was thrown into the pithouse depression after Hank's house was abandoned.

House Refuse: artifacts found in the sheet midden in front of the entrance to the house. These materials were probably used and discarded by people who lived in Hank's house.

House Storage: artifacts associated with the three storage/cache features found in the midden area.

House: artifacts found on or near the floor of Hank's house. These materials represent items used by the inhabitants of the house.

Pit 3: artifacts found inside Pit 3. These materials were probably used and discarded by people who lived in Hank's House.

Pit 2: artifacts found inside Pit 2. These materials were probably used and discarded by people who lived in a separate house nearby.

Several of us involved in the dig along with some outside specialists are still studying the artifacts and samples from Hank's site. Proper analysis takes time and isn't considered complete until a final technical report is prepared. Since this is a volunteer effort, we work on it as we can and keep chipping away at it, bit by bit. Some of the artifacts in the house assemblage are important for understanding the people who lived in Hank's house and are mentioned here.

The house assemblage is really very small, and it was obvious which artifacts had been in the house when it burned. A few flakes of Alibates flint, some turtle bones, and some fragments of cordmarked pottery were found directly on the house floor. A Washita arrow point was found in a mass of burned debris along the west wall. It is also made of Alibates flint but is permanently discolored due to the intensive heating. Because its tip was broken off, it seems likely that someone had brought an arrow home to take off the broken point and lash on a new one. The broken point was probably discarded or lost in the house.

One feature found on the floor of Hank's house is a cluster of artifacts found along the south wall. This group of artifacts consists of two bone tools, a mussel shell scraper, a rounded caliche pebble, a rounded silicate pebble, and a flake with a worn edge. All of these tools were found in a small pile, as if they had been intentionally laid there or perhaps inside a rawhide bag or small basket (which subsequently rotted). Although we can't say for sure what these artifacts represent, my working hypothesis is that they represent a "potter's tool kit." Such tool kits were groups of implements that would have been used to make pottery.

The pottery we've found at Hank's site doesn't match the thin Borger cordmarked pottery found in Antelope Creek sites. Most of it is relatively thick-walled and is tempered with chunks of scoria, a locally available volcanic rock. Scoria tempering has not been reported from most Plains Village sites, and was once regarded as a diagnostic trait of Woodland-period pottery. We're finding scoria-tempered pottery all over the West Pasture, and its presence in this area makes perfect sense because large chunks of scoria are found in the upland gravels along much of the Canadian River valley in the eastern Texas Panhandle.

The local scoria is a dark red to black, porous-looking igneous rock full of little air pockets that form tiny holes along its edges. It is, in fact, a volcanic slag ejected out of a volcano during an eruption. The most important characteristics of the local scoria are that the weathered nodules are easy to crush up, and the tiny bits of rock have sharp angular edges and make excellent temper for binding clay together in pottery making.

In the West Pasture sites, scoria-tempered cordmarked pottery is found in contexts that are definitely Plains Village, such as on the floor of Hank's house and in the trash-filled storage pit nearby. Jack Hughes reported finding it at a number of Plains Villager sites in the eastern Panhandle. David Hughes also noted scoria tempering in some of the ceramics at Buried City. Thus it appears to be a common pattern in the eastern Texas Panhandle and a continuation of a more widespread Woodland pattern.

Special Samples

Besides the artifacts recovered from Hank's site, many samples of charred materials and sediment (from soil or fill layers) were collected. Sediment samples were processed using the flotation method to recover charred plant remains. All of these samples were then examined by Dr. Phil Dering, then with the Archeobotany Laboratory at Texas A&M University (Dering now has his own independent lab in Comstock, Texas). Many of the charred samples were of wood or grass that was part of the house and are discussed below. Other charred samples from the floor of the house, the storage pits, or the trash midden in front of the house represent plant foods eaten by the people who lived there. Several charred fragments were identified as corn (Zea may) and indicate that the inhabitants were farmers. Corn has been found in the village sites of the Antelope Creek and Buried City peoples, but this is the first time corn has been found in a village in the eastern half of the Canadian River valley. Corn remains were found on the floor of Hank's house and in the trash-filled storage features (Pits 2 and 3).

Besides the corn, lots of fragments of charred sunflowers seeds were found on the house floor and in the central hearth. Because the seeds were all very small, we know that these were from wild sunflowers rather than the larger domesticated varieties. Plum pits also were found, particularly in the trash midden area in front of Hank's house. It was difficult to determine whether they were unburned or burned, but a few appear to have been roasted lightly, with the pits having been broken open to get at the tiny seed inside.

Another interesting find in Hank's house was a burned clay mud-dauber's nest. This nest was built by flying wasps, probably the black and yellow mud-dauber (Sceliphron caementarium). Using mud from the creeks, they create organ-pipe nests composed of multiple tube-shaped cells. The wasps lay an egg inside each cell, and they add a collection of tiny spiders, still living but paralyzed by venom. When the egg hatches, the wasp larvae feeds on the spiders until it matures and breaks out of the cell. Mud-daubers commonly build their nests in protected areas, such as underneath the overhanging eaves of houses or barns. Because mud-daubers have two generations of offspring a year—one in the winter and one in the summer—the nests found in prehistoric sites are not good indicators of the seasons when occupations occurred. The mud-dauber nest found at Hank's house has one sealed tube and one open tube, but it is not uncommon for larvae to never mature and emerge, and mud-dauber nests may survive intact for several years unless people remove them. These mud-daubers are not particularly aggressive, but people at Hank's site would have viewed them as pests because their stings are rather painful. Mud-dauber nests are very common in abandoned houses, and the burned nest found inside Hank's house would have come from the inside. Although it is not conclusive, this could be evidence that the house had been abandoned for at least a few months in the winter or summer before it burned.

Dating: When was Hank's House Built?

Three radiocarbon dates provide the best evidence of when Hank's house was built. The first date was obtained from a large chunk of charred wood taken during the initial site investigation. This sample represented wood from the roof of the house that burned and collapsed onto the floor, but we believe the date on this sample is too old because of the old wood problem. To illustrate the old wood problem, bear in mind that the radiocarbon dating method measures the time elapsed since an organism, a tree in this case, died rather than the time elapsed since the burning episode. This is because living organisms continue to "cycle" atmospheric carbon and accumulate tiny fractions of radioactive carbon-14 (C14). Once they die, the accumulated C14 slowly decays, or breaks down, from its radioactive form into nitrogen-N14. By measuring the ratio of C14 to stable carbon isotopes C12 and C13, we can estimate the elapsed time since the organism died.

Since the first sample from Hank's was only a section of a burned branch, it may have been from the core of the branch rather than the outer rings. Since the interior core wood of a tree branch is dead while the outer rings are living, a date on core wood could be much older than the actual time the tree died. Unfortunately, there is often no way to evaluate this. It also is possible that the sample represents old wood that was used long after the tree had died. If a branch that had been dead for many years was used in the construction of the roof of the house, a radiocarbon date from it would be much older than the actual time when the house was built and used. Keep in mind that in circumstances where wood is scarce, like the Texas Panhandle, prehistoric peoples often reused the timbers of old houses to build new houses. Such recycled timbers protected inside a house could be tens or even hundreds of years old. It is not uncommon to encounter situations where a radiocarbon date representing the death of a tree is many years older than—even a century or two before—the actual event that an archeologist is trying to date.

When we got back the first date in the fall of 2000, we had no reason to doubt it, but we knew from experience that it is risky to rely on a single radiocarbon date. Radiocarbon dating is a statistical estimation technique and any statistician knows that the larger your sample size, the more confident of the estimate you can be. In our case, "large sample size" meant getting more than one date, and that is exactly what we did. So two more dates were run; this time we were careful to minimize the risk of the old wood problem.

The other dates are better indicators of the true age of Hank's house. The second date was from the outer rings of the large juniper post found in the house. While the 8-inch-diameter post is composed of many tree rings, only the outermost ring was living at the time the tree was cut. By selecting a sample of only the outer two or three rings, the radiocarbon date will represent the last two or three years of the tree's life. The third date was on charred corn cupule fragments recovered from the floor of Hank's house. Since corn plants only live for one year, the radiocarbon date on this sample represents the year the corn cob was harvested. It is likely that the corn cupule was burned when the discarded corn cob was used as fuel and trampled into the floor the same year that the plant was harvested. The corn fragment was trampled into the floor where it lay for about 700 years until archeologists dug it up.

When these two radiocarbon ages (numbers 2 and 3 in above chart) are calibrated (a method of increasing accuracy by comparing individual dates with a well-dated chronology of individual tree rings), the resulting dates overlap. The cutting of the tree marks the construction of the house and the harvesting of the corn marks an activity that occurred while people were living in the house, probably no more than 15 years after the house was built. These two dates are essentially contemporaneous (statistically speaking, they overlap at the 1-sigma level) and may be averaged using a special statistical technique. This technique showed that there is a 95% probability that the construction and use of Hank's house occurred between A.D. 1276 to 1391. This places the occupation of Hank's house squarely within the Plains Village period in the Texas Panhandle, which is thought to date from A.D. 1200 to 1450. Unfortunately, a more precise estimate is not possible due to the accuracy limitations of radiocarbon dating. All things considered, we think it is likely that the house was used during the early 1300s. (We could improve the estimate with additional dates, but this is quite expensive—radiocarbon dates on small samples cost about $600 each.)