Introduction. Texas prehistory extends back at least 13,500 years and is marked by a variety of Native American archaeological sites and cultural remains. The "historic" era began in 1528 with the shipwreck of Pánfilo de Narváez's expedition and the subsequent account written by Álvar Núñez Cabeza de Vaca.
The prehistory of Texas has been studied by both professional and avocational archaeologists for many decades. The first excavations were apparently at the "Old Buried City" (the Handly ruins) in Ochiltree County, directed in the early 1900s by T. L. Eyerly of the Canadian Academy. The most notable pioneer, from an academic perspective, of early Texas archaeology was James E. Pearce, of the University of Texas, who began excavations in Central Texas around World War I. His techniques were crude and his analyses limited, but his work provided the first insights into the prehistoric past of the state. The beginnings of avocational archaeology in the state can be traced largely to Cyrus N. Ray of Abilene, who in 1929 was instrumental in founding the Texas Archeological Society, which continues today to play a major role in fieldwork, training, and publication in the state.
Broader views of the ancient past came from the excavations at many sites in different parts of the state, many of them supervised by A. T. Jackson, on behalf of Pearce, for the Work Projects Administration (WPA) during the depression years of the 1930s. Again, however, techniques were poor in most cases and little interpretation was done. Early museum research was conducted by the Witte Museum at the Shumla Caves in the lower Pecos region in the early 1930s; in addition, the Panhandle-Plains Historical Museum was established in 1932 and carried out research in the Panhandle of Texas. The research of Alex D. Krieger at the University of Texas from 1939 to 1956 served to integrate some of the WPA work of the 1930s. His publications provided thorough site reports and syntheses of broad aspects of the archaeological record. His research brought national attention to the prehistory of Texas. Krieger's influence culminated in the first major synthesis of Texas prehistory, published in 1954. Also of national interest in the 1940s and early 1950s was the work on "early man" in the New World (the Paleoindian period) by E. H. Sellards and Glen Evans of the Texas Memorial Museum. Great strides in learning the cultural history of ancient Texas came in the 1960s, when archaeological teams carried out excavations in proposed reservoir basins along many Texas rivers. Notable among these was the work at Amistad Reservoir on the Rio Grande and the research at Lake Sam Rayburn on the Angelina River in East Texas.
From the late 1960s through the early twenty-first century, archaeological work and research has expanded greatly in the state. Numerous universities established departments of anthropology, and archaeologists began new research programs in various regions. In addition, federal environmental legislation from the 1960s and beyond led to the study of "cultural resources" that were to be affected by proposed development projects; this legislation and implementation of regulations spawned a large number of archaeological studies, most now carried out by a wide variety of cultural resource management firms in the state. The post of state archeologist, held initially by Curtis D. Tunnell, was inaugurated in 1965. The Texas Antiquities Code, passed by the legislature in 1969, required cultural resources investigations on state-owned lands. The explosion of archaeology that first accelerated in the mid-1970s has produced a vast literature on Texas prehistory, and a wealth of information on the chronology and cultures of the ancient Native American peoples of the state.
It is important, however, to understand the scientific approaches used by archaeologists. These make possible the advances in knowledge that have taken place and distinguish archaeology from the collecting of Native American relics or the myths advanced by the rediscovery of Native American culture in the late twentieth century. Archaeological methods include a variety of techniques to ensure the systematic collection of data, including GPS, drones, LiDAR, remote sensing machines, and hand and machine excavations with established horizontal and vertical provenience data. The use of grid systems to record horizontal information and the excavation of site deposits using arbitrary or natural levels to record vertical data are at the heart of archaeological systematics. The critical factor for the scientific analysis of archaeological data is its context: we have to know how, where, and under what conditions an object was found; what other objects were associated with it; and what patterns they may be a part of. A cigar box of "arrowheads" may be of interest to the archaeologist and, if collected from a specific area, have some interpretative value, but these relics lack context and are thus of little value in exploring Texas prehistory. Similarly, the random digging of a prehistoric site does not contribute to a knowledge of archaeology, but prevents it, like ripping pages from a book destroys it. Texas archaeology has long been plagued, and especially since the 1990s, by commercial digging of sites. Such looting destroys the prehistoric record and greatly diminishes the prospects of learning detailed information about the ancient peoples of the state.
The Texas archaeological record comprises sites: any place on the landscape that has been modified by human beings. Common among these are campsites, where daily life took place; quarries or lithic processing areas, the locales of stone-chipping; temporary campsites, representing brief hunting or gathering forays; kill sites, where bison or other mammals were slaughtered and butchered; rock art sites, overhangs, caves, or shelters with pictographs or petroglyphs, such as Seminole Canyon State Park and Historic Site and Hueco Tanks State Park and Historic Site; caves and rock shelters, protected overhangs in canyon walls, which some Native American groups, particularly in West Texas, used for daily occupation (these provide extremely well-preserved organic remains reflecting everyday life); mound sites, purposeful accumulations of earth found in ancestral Caddo sites in East Texas, used as platforms for dwellings or for burials; burned-rock middens, incidental accumulations of fire-cracked rock, often in mounds, used for food-processing, and found associated with campsites in Central and West Texas; shell middens, accumulations of marine shells from shellfish collected as food, principally oyster shells on the central coast and Rangia brackish-water clams on the upper coast; and cemetery sites, areas set aside for the disposal of the dead, found in the Late Archaic and Late Prehistoric eras in Central, Coastal, and East Texas.
This abbreviated list of site types suggests the diverse ways in which ancient Native Americans utilized the terrain and took advantage of its plant and animal resources. At many sites, the only surviving cultural remains are stone tools. If we are to understand what was done at such sites, it is vital that we know the context of the stone tools. For instance, at kill sites, proper excavation will usually discover projectile points and cutting or butchering tools in association with animal bones. The most common type of kill site in Texas is the bison kill of Paleoindian times (before 10,000 years ago). For example, at Bonfire Shelter, near Langtry in Val Verde County, excavations revealed a mass of bison remains associated with Folsom and Plainview points, accompanied by flakes and bifaces used for processing the slaughtered animals. At quarries or lithic processing areas, controlled surface collection will often yield great numbers of large, crudely chipped bifaces, rocks in the early stage of tool-making known as quarry blanks. Rarely are projectile points or other finished tools found, since this is a locality where the basic levels of stone-working took place to secure good chipping materials using a hammerstone to remove the rough exterior from the cobbles, and roughly shaping the blanks for further reduction elsewhere. Although they have long been ignored, lithic processing areas are important sites for archaeological study, as they shed a great deal of light on a fundamental activity of prehistoric cultures. One quarry site that is open to the public is the famed Alibates Flint Quarries, on the Canadian River in the Panhandle.
Campsites are found across the state along streams or other water sources; most are "open occupation" sites, although caves and rock shelters were also often used for habitation. Many represent the villages of hunters and gatherers, whose foraging was the main way of life throughout Texas until later times, when farming was introduced in East Texas and in parts of the Panhandle and far West Texas. Campsites, the locales of daily life, were perhaps occupied for a few weeks or months before the group moved on to exploit the plant and animal foods of another area. These are the most common sites and contain great quantities of stone tools, flakes, and other debris. Context is particularly important in these sites. To be able to understand the behavior of the ancient inhabitants and the activities they carried out at these places, a large block or open-area excavation is necessary. In it, the projectile points, scrapers, choppers, flakes, animal bones, snail shells, and other items can be plotted in place, allowing the study of their spatial patterning. The distributions of plotted artifacts often show where tool-making took place, where animals were skinned and butchered, where bone tools were made, or wooden spear shafts fashioned. The relationships of the tools to the areas of the site and to other stone tools provide, then, contextual information critical to archaeological interpretation.
Much of what is found in Texas prehistoric sites is artifacts of chipped stone (such as projectile points), pottery, antler, bone, and shell. If excavated with systematic methods, the context of these artifacts—taken as a whole—provides a picture of ancient life at certain periods of time. Critical to our interpretations is the dating of these materials and contexts. The most widely used method for absolute dates is radiocarbon analysis, in which associated organic remains (such as wood charcoal from a hearth) can be assayed to yield a date for cultural remains at the same level of the hearth. Excavations can also provide relative dates by determining which styles of artifacts are earlier or later than others. Once a chronology is established at several regional sites, types of known dates can then be "cross-dated" by distinctive artifacts to other sites. Gradually, a framework of prehistoric cultures can be built up in a sequential fashion. In Texas, research has shown that in most regions, distinctive changes occurred in the shapes of projectile points through time. These artifacts occur in two forms: as dart points—large, heavy points used on the tips of spears thrown with the spear thrower or atlatl, common in the Paleoindian, Archaic, and much of the Woodland periods—and as arrow points—tiny, thin points that tipped arrow shafts, often made of cane, when the bow and arrow was adopted by ancient Texas cultures around A.D. 700. This weapon appears to have wholly replaced the spear thrower, as it was more accurate and more effective at longer distances (tiny arrow tips could penetrate a bison, a man, or a smaller creature). Many of the dart and arrow points can be sorted into "types" of distinctive shapes that are restricted in distribution in both time and space. This makes the points time-sensitive and valuable chronological aids for archaeological research.
Archaeological work has continued in parts of Texas for more than 100 years. Some areas, such as Central Texas and East Texas, have been intensively studied, and detailed archaeological sequences have been established there. In other regions, such as South Texas and North Texas, research intensified only in the 1970s, and much remains to be learned. Cultural change proceeded at different rates over the vast area of Texas; in some regions, hunting and gathering cultures persisted throughout prehistory; in others, cultures with farming and settled village life appeared. Research has divided the Texas prehistoric archeological record into five general periods: pre-Clovis (ca. 18,000–13,400 years ago), Paleoindian (13,400–10,000 years ago), Archaic (10,000 years ago to around the beginning of the Christian era or later, depending on the region), Woodland (ca. 2500–1150 years ago in only a few regions), and the Late Prehistoric (ca. 1250–1150–420 years ago).
Pre-Clovis. The term Pre-Clovis is associated with archaeological sites that have been accepted as reliably dated to beyond the age of Clovis technology which appeared in the archaeological record around 13,400 years ago. For decades, archaeologists across Texas and the Americas thought that Clovis represented the oldest stone tool culture. This view has been challenged by the growing evidence for much older sites. Debate surrounding the presence of Pre-Clovis people in the Americas intensified largely because these early sites are often difficult to recognize. However, some areas where Clovis-age deposits have been unquestionably identified have the potential and/or have helped archaeologists detect early sites around Texas because cultural material lies stratigraphically below the Clovis material.
Evidence indicates that the earliest known modern human inhabitants of Texas date back to around 20,000 years ago at the Gault site in Central Texas. Gault has evidence for Pre-Clovis human groups occupying the Buttermilk Creek valley area in Bell County. The tools that were left behind include small projectile points, large bifaces, macro-blades, and basic flake tools, and demonstrate that people were living as hunter-gatherers in Texas between 16,000-20,000 years ago. The majority of the stone tools were made from chert, a local material that can be identified as seams in limestone outcrops that form part of the Edwards Plateau. These stone tools show that people were engaged in generalized hunting and foraging activities, including processing meats, animal skins, and harvesting plant materials, as well as manufacturing and maintaining tools made of stone, bone, antler, and wood.
Further downstream along Buttermilk Creek lies the Debra L. Friedkin site, where evidence of cultural materials show that people were there between 13,000–15,000 years ago. The tools from Friedkin include bifaces, blades, and flake tools used for processing hide, cutting meat, grooving and incising a variety of materials including wood and bone, very similar to the types of activities seen at the Gault site. Both of these sites are found in Central Texas along the ecotonal edge of the Edwards Plateau and Blackland Prairie. The areas surrounding the Gault and Debra L. Friedkin sites are rich in natural resources and have attracted human inhabitants to these regions in Texas almost continually for millennia.
Other Pre-Clovis sites have been reported across Texas, but the evidence is limited. At the Wilson-Leonard site in Williamson County, a small number of chipped stone flakes below the known Clovis occupation level indicates the presence of people before Clovis times. In western Travis County, a sparse lithic assemblage in Zone 1 of the Levi site may also predate the Clovis occupation level. The Bonfire Shelter site near Langtry (Val Verde County) is the oldest mass bison-kill site in the Americas, and dates taken from the oldest bone bed suggest that butchering was taking place at the site around 14,000 years ago. The Cueva Quebrada site, a small cave high on a cliff overlooking the Rio Grande, also in Val Verde County, has evidence for burned bone with butchering marks, and may also be 14,000–16,000 years old. In Nueces County, research of the Petronila Creek site recovered tools made of stone and bone, including bone tools with butchery marks. Dates from mammoth bone recovered from the Petronila Creek site revealed that these cultural materials are around 18,000 years old. All of these sites indicate that there were humans occupying specific regions of Texas long before Clovis technology appears on record.
Given the evidence that humans were present in Texas by at least 18,000 years ago, human groups of hunter-gatherers likely began to move out of Eurasia sometime between 20,000–30,000 years ago. Migrations along the coastal margins were likely facilitated by early watercraft, and highly successful groups were able to continue expanding their traditional hunting territories by exploring new regions; eventually they ventured upriver and expanded hunting ranges to the interior landscapes of the American continents. The stone tool technologies present in Texas likely represent only part of the early evidence, and it is likely that multiple groups arrived at different times in the Americas during this early period. These groups would have spread out across the landscape, and some eventually settled in Texas. It is from these groups that both Clovis and a number of contemporary stone tool cultures developed around 13,500 years ago.
The search for the earliest humans to arrive in Texas, and indeed the United States, is one of the most hotly contests debates in archaeology. Scientists and archaeologists continue to look for answers as to where, when, why, and who the first humans were to set foot in the Americas. Regardless, Texas is one of a handful of states that has a deep human chronology, and as archaeological evidence from the Gault and Debra L. Friedkin sites demonstrates, Texas has been home to humans for millennia, and they were drawn to the region for its abundant natural resources that continue to sustain Texas life through the modern era.
Panhandle. Native Americans ranged over and eventually settled in what is now the Texas Panhandle more than 13,400 years ago. A mammoth kill site, Miami, of early Paleoindian age with Clovis points, is found in Roberts County in the Panhandle. The Clovis points, found with the remains of five mammoths, had been resharpened and used as butchering tools (see MAMMOTH HUNTING). At Lubbock Lake, several species of megafauna were processed and butchered in Clovis times. Later Paleoindian bison kills in the region are marked by Folsom and Plainview points, bone butchering tools, and flake tools, including sites such as Lipscomb, Lake Theo, Lubbock Lake, and Plainview.
Hunting and butchering activities continued in the Panhandle during the Archaic periods, along with plant food processing in camp sites in earth ovens, with the excavation of wells around 6,500 years ago at the height of droughty periods. Sites of the Early to Middle Archaic periods are found in playa basins and stream valleys and spring-fed canyons along the Canadian River breaks and Caprock Canyon escarpment. Modern environmental conditions on the Panhandle were established by 4,500 years ago, and sites known include camp sites with rock hearths, barbed dart points, lithic tools, and lithic debris from tool manufacturing activities, occupied rock shelters, and lithic procurement sites. With a returned abundance in bison in the region, bison kills are known in the Panhandle and South Plains in Late Archaic times, after ca. 2,500 years ago; small game was also hunted. These sites have bison bone beds where hunters used arroyos as natural traps to capture and contain bison so that they could be killed with spears; very few artifacts other than the bones and the dart points are present at these kills.
The Ceramic period (ca. 2,000–1,500 years ago) and the Late Prehistoric I period (ca. 1,500–800/900 years ago) on the Texas Panhandle and Southern Plains is marked by the manufacture and use of ceramic vessels and arrow points. Some of the earliest evidence of the use of these material culture remains come from Deadman’s Shelter at Mackenzie Reservoir, at between 1,900–1,700 years ago. The pottery in the earlier part of the Ceramic period, or the Plains Woodland period, is a coarsely-tempered and thick-walled cordmarked jars. In early Late Prehistoric Lake Creek and Palo Duro culture sites the material culture assemblage includes corner-notched Scallorn and basally-notched Deadman’s arrow points as well as other chipped stone tools (unifaces and edge-modified flakes, bifaces, and cobble tools), and ground stone manos and metates. A few of the ceramics appear to be non-locally made and plain Jornada Mogollon brown ware. Open campsites and rock shelter sites are known during this period, as well as a few residential bases such as the 1,300–1,000-years-old Kent Creek site in Hall County. The site had rectangular pithouse structures with entranceways, baking pits, and hearths. Goosefoot, purslane, and charred acorns were food stuffs, along with deer, antelope, bison, and a range of small mammals, turtles, and freshwater mussel shells.
After ca. 1,000 years ago, lithic tools found on Panhandle sites include triangular arrow points (such as Fresno, Washita, and Harrell types), scrapers, and large beveled bifaces similar to those seen on contemporaneous Southern Plains village farming societies. In addition to hunting bison and gathering plant foods, other game animals such as pronghorn antelope, mule deer, elk, rabbits, and smaller rodents were also procured by Late Prehistoric groups.
In certain parts of the Panhandle and Llano Estacado, settled villages with masonry architecture developed. The inhabitants engaged in bison hunting and agricultural pursuits, such as the cultivation of corn, beans, and squash, and are found in the Antelope Creek phase of the Plains Village or Middle Ceramic period in the Canadian River valley between ca. 800–500 years ago and in other parts of the Panhandle in Buried City Complex sites in Ochiltree County. The masonry architecture employed in house construction in Antelope Creek phase sites used large stone slabs plastered using adobe mortar for wall construction. Circular or rectangular house structures occurred singly on sites, and there are also small hamlets and multi-room dwellings of semi-subterranean rectangular plastered rooms at sites such as Antelope Creek Ruin 22. Other structures in the region, such as the Hank site in Roberts County, were rectangular earth lodges in pits with a hip beam design, picket post walls, and clay-lines smoke holes in the roof. The technology of the Antelope Creek phase peoples included chipped stone tools such as arrow points, scrapers, knives, and beveled Harahey knives, stone pipes, bone and shell tools, along with grinding basins or metates, manos, and abraders. The most important lithic raw material source in the region is the agatized dolomite from the Alibates Quarries on the Canadian River in the Panhandle, and many of the quarry pits at the site likely date to the Antelope Creek phase. Village sites in the region with links to southeast New Mexico appear around the same time. Ceramics are primarily Borger Cordmarked jars with clear cordmarked or smoothed-over cord marks on the rim and body of vessels, and they have collared or non-collared rims; some cordmarked rims have punctations or incised lines at the lip-rim juncture. Obsidian artifacts occur in these Panhandle sites, most coming from sources in the Jemez Mountains of northern Mexico, and Olivella shell beads from western sources are also present; these were part of Plains-Pueblo trade in Late Prehistoric times. Some of the burials found in Antelope Creek sites have evidence of violence in the form of traumatic injuries, including fractures, puncture wounds, cut marks from scalping or dismemberment, and arrow points embedded in the human remains. These may be a byproduct of territorial defense and competition with other Panhandle groups over resources and trade, including fertile soils, access to bison herds, meat, and hides. Other Antelope Creek phase sites have evidence of defensive features.
Far West Texas: Jornada Mogollon. The Jornada Mogollon region lies at the far western edge of Texas and encompasses El Paso and Hudspeth counties. As originally defined by Donald Lehmer in 1948, the term Jornada Mogollon refers to the Late Prehistoric, or Ceramic era, cultural manifestation that was linked to the greater prehistoric Southwest. The Jornada was defined as a branch of the prehistoric Mogollon cultures of southern New Mexico and Arizona characterized by the use of brown ware pottery and pit house architecture. In modern times, the term Jornada region refers to the prehistoric and historic cultures of far western Texas, south central New Mexico, and north central Chihuahua, Mexico.
Seventy years of professional and avocational archaeological research have identified settlements ranging in age from the Paleoindian period (circa 13,000 years before present) to about 400 years ago. The types of settlements include camp sites, pit house villages, agave baking pits, pueblos, and rock art localities.
The earliest periods of human occupation are known as the Paleoindian and Early Archaic. The Jornada region was sparsely populated during these periods, especially the latter, due to the climate of the time (the Mid-Holocene Warm period or Altithermal period). Coinciding with the beginning of more favorable climatic conditions at the end of the Altithermal, major changes took place throughout the Jornada region at around 4,500 years ago. During this period, known as the Keystone phase, the first village settlements appeared, usually consisting of several small hut structures, and the earliest evidence of corn farming appeared. The first evidence of communal gatherings is found during the Keystone phase in the form of massive agave baking pits used for communal feasts. Specific rock art motifs appeared, such as the zigzag and other abstract designs.
At around 3,400 years ago, a series of technological and subsistence changes marked the transition from the Middle to the Late Archaic period. Three phases have been defined for this period: Fresnal (3,400–2,750 years ago), Arenal (2,750–2,300 years ago), and Hueco (2,300–1,600 years ago). Several distinctive spear points or darts appeared, including side-notched, corner-notched, and basal-notched forms. The Late Archaic Fresnal phase was clearly a period of incipient agriculture. The following Arenal phase is a distinctive anomaly in the Late Archaic sequence. The phase is essentially a 450-year-long hiatus in the radiocarbon record of baking pits and storage pits that represents a decline in—and in some instances the near abandonment of—the uses of certain technologies and subsistence practices. The final Late Archaic interval, the Hueco phase, was a period of increasing incorporation of corn into the diet, an expansion and intensification of settlement throughout both mountain highlands and desert basins, and population growth prior to the introduction of ceramic technology and the bow and arrow. Beyond the boundaries of the village settlements, the social dimensions of Late Archaic settlement are also manifested in rock art and ritual landscapes. The earliest evidence of shrine caves, such as Ceremonial Cave in the Hueco Mountains of El Paso County, appeared during this period.
The end of the Archaic period and beginning of the Formative period (also known as the Ceramic period) is marked by the appearance of ceramics and the bow and arrow. In general, the Formative period is characterized by drastic increases in regional population, a greater emphasis on food production, increasing sedentism and social complexity, and widespread evidence of more visible and public ritual. Four phases have been defined for the Ceramic period: Mesilla (1,000–1,400/1,800 years ago), Early Doña Ana (1,000–850 years ago), Late Doña Ana (850–700 years ago), and El Paso (700–550 years ago).
The Mesilla phase was essentially a continuation of the subsistence and settlement practices of the preceding Hueco phase but with the introduction of ceramic container technology and the bow and arrow hunting technology. The first communal civic-ceremonial structures (kivas or special communal pit houses) appeared during this period. The seventh century A.D. was a major period of population increase and settlement changes throughout the Jornada region and greater Southwest. Some researchers have described this period and its pronounced settlement and population changes as the Neolithic Demographic Transition of the U.S. Southwest. It is also during this period that the first distinctively Southwestern iconography appeared in regional rock art.
The Early and Late Doña Ana phases represent transitional periods from the pit house settlements of the Mesilla phase to the pueblo villages of the El Paso phase. Settlements were occupied more intensively, village layouts became more structured, and house structures were often arranged around common courtyards. The first decorated pottery appeared during this period in the form of red-on-brown vessels during the Early Doña Ana phase and an early variant of red and black-on-brown (El Paso Polychrome) vessels during the Late Doña Ana phase.
The final time interval of the Formative period is the El Paso phase. During this time, Jornada people resided in pueblo settlements, some of which consisted of up to 150 rooms. Each pueblo room block had a large civic-ceremonial room with special floor features and subfloor pits, and these rooms were ritually burned when the pueblo was abandoned. The pueblo dwellers subsisted on cultivated corn, beans, and squash. The El Paso phase may be considered the apex of iconographic representations of religious belief. Symbols of ancestral mountains, emergence caves, rain, ancestors, clan symbols, and deities were painted on rock art panels. Symbols also appeared on El Paso Polychrome vessels, and were carved in stone, bone, and marine shell. El Paso phase people obtained pottery, turquoise, and marine shell through widespread trade networks with people in northern Mexico and western and central New Mexico.
The fifteenth century marked a period of profound change, not only in the Jornada region but across much of Texas, the southern Plains, and the U.S. Southwest and Mexican Northwest. The numbers of radiocarbon dates, a proxy measure of population densities and energy consumption, show a precipitous decline between 550–600 years ago. In fact, the radiocarbon record, and by inference the momentary population of the region, declineed to levels equivalent to the Middle Archaic period, some 4,000 years earlier. The underlying cause or causes of such profound regional changes remain speculative and may have involved drought, pandemics, and population migration. Small populations of horticulturalists and hunter-gatherers remained in the region until after the sixteenth century.
Eastern Trans-Pecos and Big Bend. The eastern Trans-Pecos and Big Bend region of West Texas is the northeast division of the Chihuahuan Desert physiographic province—the largest desert in North America. The total desert covers some 500,000 square kilometers (or more than 310,000 square miles). The subdivision in Texas is differentiated by montane/basin environments that grade in elevation from almost 9,000 feet in the Guadalupe Mountains to just below 2,000 feet at the Rio Grande. Overall this desert region of Texas is characterized by hot dry climates, xerophytic environments, and a volcanic or sedimentary geology—converging factors that preserve a phenomenal archaeological record.
Early diagnostic Paleoindian projectile points (13,450–11,950 years ago) such as Folsom and Clovis are found in the region and at deflated sites like Chispa Creek, but intact occupational sites have yet to be discovered. Several sites containing Late Paleoindian occupations have been excavated in the region, the most notable is the Genevieve Lykes Duncan site that contains some of the earliest evidence of plant baking of desert succulents with earth ovens and ground stone tools used in the region. There is a noted increase of Late Paleoindian (11,000–8,450 years ago) lanceolate projectile points with a broad distribution across the Big Bend landscape, possibly indicating an increased presence during this period. However, archeologists have little direct evidence of hunting from this period but do know these early groups were adapted to a desert environment by 10,500 years ago and were utilizing key succulent species as part of an early broadening diet.
The emergence of new projectile point forms in addition to an increased broadening of the diet with the inclusion of more plant species is thought to mark a transition to an Archaic lifeway. In the Big Bend region, the transition in diagnostic projectile point forms from stemmed and unnotched lanceolate points to corner-, side-, and basal-notched varieties coincided with the changes in subsistence. There are few sites that date to this Early Archaic period (8,450–4,450 years ago), and of these the David Williams site best characterizes what is known about the period. Excavation at this short-term camp produced a limited amount of bone identified to deer or antelope, as well as several small thermal features, and Early Archaic diagnostic projectile points. Among some of the most important findings from the few sites of this age are the paleobotanical results that indicate the longevity of key floral species of the Chihuahuan Desert.
The Middle (4,450–2,950 years ago) and Late Archaic periods (2,950–1,115 years ago) are well represented at sites in the region as signified by a diversity of projectile point styles and related subsistence technologies. Coeval with the emergence of several varieties of contracting stem dart points is an increased repeated use of plant-baking earth ovens. This increased use may have been a result of sustained aridity and population increase, both in the context of broadening technologies. Coeval with the changing technologies are a diversity of settlement patterns, including the montane environment at sites such as Rosillo Peak (41BS762), noted use of rock shelters like Spirit Eye Cave (41PS25) and Wolf Den Cave (41JD191), as well as the earliest dated example of a placed boulder mosaic and associated cache at the Lizard Hill Cache in Big Bend National Park. As a whole, the Middle Archaic period marked a noticeable increase in the diversity of settlement patterns, technologies, and an increased presence of sites and radiocarbon dates, a trend that continued into the Late Archaic.
The proliferation of sites dating to the Late Archaic represents a continuation of a hunting-gathering lifeway, but also seen is the first direct evidence of maize in the region by 2,100 years ago. Because of the increase in the number of projectile points and radiocarbon dates from Late Archaic sites, population size is thought to have increased again in this period. In addition to maize marking a routine interaction with outside southern groups, there was the emergence of technologies associated with Plains bison hunters found in multiple Big Bend sites, being further evidence of the influx of different groups into the region, a trend that persisted into the Late Prehistoric period.
The dynamism of the Late Archaic period continued into the Late Prehistoric (1,115–415 years ago) with the spread of bow-and-arrow technology into the region and the increased use of cultigens along the Big Bend of the Rio Grande. Early adopters of the bow in the region wielded distinct projectile points that are part of the Livermore phase, numbers of which have been recovered from several mountaintop caches. The development at ca. 900 years ago of pit houses and Pueblo-like structures at the confluence of the Rio Grande and Rio Concho known as the La Junta de los Ríos site district is a dramatic change wherein horticulturalists moved into the region and developed a mutual trade relationship with local hunter-gatherer populations. Ceramics from these villages indicate that early occupations were by outside groups, but by the late period, after 400 years ago, local traditions had developed as clusters of manufacturing locales. There is no evidence for outside manufacture of ceramics from villages in New Mexico or Mexico. Near these La Junta villages are sites of the Cielo Complex, the archaeological expression of a local non-ceramic and hunter-gatherer population group who used stone-based wickiup rings.
Lower Pecos Canyonlands. The state’s smallest archaeological region, the Lower Pecos Canyonlands, is defined by the geographic distribution of the 3,000-year-old Pecos River Style polychrome pictographs centered on Val Verde County in Southwest Texas and extending south into the Mexican state of Coahuila. The limestone cliffs towering above the Pecos and Devils rivers and their tributaries as they empty into the Rio Grande contain hundreds of overhangs—rock shelters and caves—frequented by prehistoric foraging societies. These sheltered sites preserve pictographs and ancient occupational debris, including perishable organic materials such as fragments of basketry, matting, and leather. Archaeologists and artifact collectors have dug into Lower Pecos rock shelters in search of perishable artifacts and information since the 1920s.
Humans likely first explored the Lower Pecos region during the Late Pleistocene geologic era some 14,000–16,000 years ago and left behind traces such as fragmented animal bones with possible butchering marks at Cueva Quebrada. More convincing evidence dates to around 12,000 years ago at Bonfire Shelter near Langtry where a massive deposit of burned and butchered Bison antiquus bones known as Bone Bed 2 occurs with Early Paleoindian projectile points and butchering tools. The Late Pleistocene in the Lower Pecos Canyonlands, like most of North America, was considerably cooler and wetter than the present day and supported grasslands that at times drew herds of now-extinct herbivores like the Ice Age bison found at Bonfire Shelter. The small highly mobile groups of hunters and gatherers who hunted the bison may have only visited the region in favorable conditions.
By 10,000 years ago as the climate became warmer and drier, the native plants and animals characteristic of the Lower Pecos region today took hold, as did human populations. Key to year-round occupation was developing predictable food resources from a broad spectrum of native species ranging from top prey such as deer to a wide variety of smaller animals such as rabbits, rodents, lizards, and fish and many kinds of plants from nuts and berries to mesquite beans and grass seeds.
The most telling pattern of human-plant interaction focused on green rosettes such as agave lechuguilla, sotol, and yucca that provided both food and strong fibers used to make baskets, cordages, sandals and more. These arid-adapted plants were available year-round but required specialized labor-intensive processing techniques. The central stems or hearts of lechuguilla and sotol yield sugary carbohydrates but only after being slowly cooked over several days in underground plant-baking facilities known as earth ovens. Earth ovens are layered arrangements of heated rocks, green plants such as prickly pear pads to provide moisture, and the green rosette hearts covered by a thick layer of earth. Archaeological evidence of such facilities, oven pits surrounded by discarded processing debris, can be found in many rock shelters and hundreds of open campsites along the rivers and across the uplands. The most prominent archeological signature of plant baking is the often-massive accumulation of spent cooking stones known as a burned rock midden. Literally tens of thousands of burned rock middens and smaller cooking or hearth features occur in the Lower Pecos and across the adjacent Edwards Plateau and the Big Bend.
What may be the earliest well-documented evidence of earth oven cookery in North America occurs in the Lower Pecos and nearby Big Bend dating between 10,000–10,500 years ago. This marked the beginning of what some have called the “Carbohydrate Revolution,” a broad pattern of consuming carbohydrate-rich plant foods that spread across much of temperate North America during the Holocene or recent geologic era.
Archaeologically, the span of time between roughly 10,000 and 1,000 years ago is known as the Archaic period. In the Lower Pecos Canyonlands, the Archaic is typically broken into Early-Middle-Late chronological periods and sometimes divided into six somewhat shorter local subperiods with evocative names like Eagle Nest, Viejo, and Cibola. Around 1,000 years ago the principal weaponry changed from throwing spears or darts propelled with atlatls (spear throwers) to the bow and arrow, a change considered to mark the end of the Archaic and the start of the Late Prehistoric era. Hunting and gathering lifeways, including plant baking, continued to characterize the Lower Pecos Canyonlands until the arrival of the Spanish in the 1600s.
Throughout most of the twentieth century archeologists portrayed the prehistory of the Lower Pecos as representing a very conservative hunting and gathering tradition that changed little through time. A static picture emerged of small-scale human societies tethered to permanent water and living much of the year in the Canyonland rock shelters. Today most researchers suspect that Lower Pecos prehistory was much more eventful and dynamic with migrations in and out of the region in response to climatic change, continental population increase, and wider social movements.
One obvious example is represented by Bone Bed 3 at Bonfire Shelter, a massive deposit of Bison bison bones created by bison “jumps” in which hundreds of stampeding animals were driven off the cliff above the shelter by well-orchestrated communal hunts by Late Archaic groups who may have been newcomers to the region. These events occurred about 2,800 years ago at the same locality used by Paleoindian bison hunters some 8,000–9,000 years earlier. The Late Archaic bison jumps coincided with a cooler and wetter interval when grassland expansion brought herds south from the Great Plains. Later influxes of new people are thought to have occurred around 1,000 years ago, during the Flecha subperiod, marked by a new pictograph style (Red Monochrome), and around 500 years ago in the Infierno subperiod. The latter influx may have also been triggered by the return to the region of bison and bison hunters, peoples who set up temporary encampments marked by clusters of wikiup rings on upland promontories overlooking the canyons. By the time Spanish expeditions skirted the rugged Lower Pecos Canyonlands in the sixteenth and seventeenth centuries, they briefly encountered and recorded the names of several dozen hunter-gatherer bands.
Lower Pecos Canyonlands Rock Art. For thousands of years, Native Americans wrote their sacred stories and documented their histories on the rough limestone walls of the Lower Pecos Canyonlands. They communicated this information through painted images of humans (anthropomorphs), animals (zoomorphs), and a wide range of enigmatic imagery. These ancient visual texts serve as a window into the past, a window into the lifeways of the people of the Pecos (see INDIAN ROCK ART). Pictographs in the Lower Pecos have been categorized into five styles or classification: Historic, Red Monochrome, Bold Line Geometric, Red Linear, and Pecos River.
Pecos River style is the most abundant and visually impressive rock art in the region. These paintings in red, yellow, black, and white were made with mineral pigments and other natural ingredients, such as animal fat and plant extracts. They were applied to the rock surface using brushes made from animal hair, plant fibers, feathers, or fingers. The precision and clarity of lines found in Pecos River paintings would present a significant challenge for even the most accomplished artist today.
Archeologists know of more than 200 murals north of the Rio Grande. There are likely as many or more south of the border in Mexico. These ancient murals vary in size and complexity. Some are huge, measuring more than 100 feet long and 30 feet tall. Scaffolding and ladders were necessary for their production. Others are small and tucked away in secluded alcoves. Artists in the Lower Pecos first began painting these murals around 4,500 years ago and continued to create them for almost 3,000 years.
It was long believed that each rock art panel represented numerous painting episodes conducted by different artists over hundreds or thousands of years. Through detailed analysis of these panels researchers now know that many, if not most, are well-ordered compositions, planned arrangements of elements in a work of art to communicate an idea. Lower Pecos artists used pigment crayons to sketch out and organize the elements of their compositions. Vestiges of these preliminary sketches still can be seen today.
Pecos River style murals include anthropomorphs, zoomorphs, a wide range of geometric imagery, and enigmatic creatures that are not identifiable as human or animal. Anthropomorphs are by far the most frequently-depicted element. Although they share many common characteristics, such as elongated, rectangular bodies with disproportionally short arms and legs, there is incredible diversity in their representation.
Most anthropomorphs range in size between three to seven feet. Some, however, are monumental, towering more than twenty feet tall. Others are pocket-sized; the smallest recorded is only three inches tall. Headdresses of varying types, clusters of feathers at the waist, and wrist or elbow tassels often adorn these human-like figures. Generally, the artists portrayed anthropomorphs wielding paraphernalia, such as atlatls (spear throwers) and staffs.
Animals also are portrayed. Deer and felines are the most common, birds and canines to a lesser degree, and bears are conspicuously absent. Some imagery resembles insects, such as caterpillars, dragonflies, and moths or butterflies. Other figures resemble snakes, tadpoles, and frogs.
Scholars long feared that the meaning of these magnificent murals was lost with the artists who produced them. Recent research has demonstrated otherwise. Rigorous analyses of the art have brought compelling evidence about its manifold meaning, revealing a stunning depth of sophistication and complexity.
Red Linear is characterized by animated, small, fine-lined figures of animals and humans. Although the style is referred to as “Red Linear,” implying the use of only red in their production, they were also painted in black, yellow, and white. Until only recently, this style was believed produced after the Pecos River paintings. Researchers, however, have identified numerous examples of Pecos River style superimposing Red Linear pictographs. In other words, Red Linear is either older than or contemporaneous with the Pecos River style.
There is tremendous diversity within Red Linear imagery. While some figures are rigid and static, others are fluid, evoking a sense of movement as humans engage in group activities, such as dancing, hunting, and possibly conflict. Most figures are small, only two to four inches tall, but some extend to more than twenty inches. Unlike other regional styles, Red Linear imagery includes themes associated with sexuality and pregnancy. Other pictographic elements include hand-held implements (such as atlatls and fending sticks), headdresses, looped lines, and other geometric forms.
Animal imagery is common and includes finely-executed canine-like figures, rabbits, and deer. In some cases, humans appear to be driving deer into nets and snares, illustrated by elongated grid patterns or looped lines. Numerous scenes such as these seem to represent hunting or trapping activities, although this may be too simplistic an interpretation. Hunting episodes documented in myth and art often have multiple levels of meaning, some mundane and others supernatural.
Bold Line Geometric lacks the artistry of either Pecos River or Red Linear rock art. It includes roughly-executed broad lines of red or yellow paint forming zigzag, herringbone, and other geometric patterns. Human and animal-like figures are far less common. When present, they are unrealistic, with round bodies and sharply bent arms and legs that mirror the zigzag lines characteristic of the style.
Red Monochrome paintings are characterized by life-size images of humans and animals painted in shades of red. Humans are more realistically depicted than in the other regional styles and are portrayed with hands, feet, fingers, and toes. Their bodies face frontally with arms bent at the elbow and forearms raised, sometimes holding a bow and arrow. The portrayal of bows and arrows reinforces the placement of this style within the Late Prehistoric period. The range of animals portrayed in Red Monochrome paintings is more extensive than the other styles and includes depictions of panthers, deer, turkeys, rabbits, and dogs. Aquatic life includes turtles and catfish.
Rock art of the Historic period refers to paintings created after indigenous people’s first contact with European explorers. The rock art of this period provides a fascinating record of the convergence of the two cultures. Native American artists painted images of Spanish officers dressed in uniform, missionaries, churches, and domesticated animals, such as cattle and horses. And, as their ancestors before them, they continued to paint their sacred stories and document their ceremonies.
Southern Texas. Southern Texas prehistory has seen increased research in recent years, but the chronology remains elusive. Southern Texas is a broad coastal plain extending to the south of the Balcones Escarpment and lying between the Rio Grande and Gulf of Mexico and includes South Texas (the interior), the Coastal Bend, and the Rio Grande delta. Since earliest times, southern Texas has been occupied by hunters and gatherers, exploiting the resources available in their areas.
The early Paleoindian period in southern Texas is known almost entirely from surface finds. Clovis and Folsom points are rare but widespread. The greatest number of these early forms, and those of middle and late Paleoindian age, have been found as a result of massive sand quarrying at the Wilson County Sand Pit and the Atascosa Sand Pit. No kill sites are known; indeed, few excavations have been done on sites of this age. The Buckner Ranch site in Goliad County was dug in the 1940s and yielded much Pleistocene fauna and a variety of projectile points. However, in 2018–19 the excavations were reopened, and a vast sandy Holocene slope along Berclair Creek was found to be responsible for the spreading and mixing of point types. The Coastal Bend has yielded a surf-rolled obsidian Clovis point at Port Lavaca, a fluted point along Oso Creek, and additional fluted points in San Patricio County.
The middle part of the Paleoindian period is marked by point types that include Plainview, St. Mary’s Hall, and Wilson. Wilson points are surface finds in southern Texas and the Coastal Bend. St. Mary’s Hall specimens were identified in 1977 excavations at the St. Mary’s Hall site in San Antonio. Artifacts included point fragments, large end scrapers, a bifacial Clear Fork tool, preforms, and lithic debris. No radiocarbon dates were obtained, but 2011 excavations at 41BX1396 revealed a cultural sequence with St. Mary’s Hall at the bottom, radiocarbon-dated around 10,400 years ago. Hundreds of these points have been documented across the area, as far east as Copano Bay, near Rockport, and at 41VT141 in Victoria County.
The latest Paleoindian sites (ending ca. 10,000 years ago) are broadly represented across the area. From the Coastal Bend to the Rio Grande, many specimens of the Golondrina and Angostura types, and occasional Scottsbluff points, have been found. The Golondrina form occurs broadly, as far south as the San Isidro site in Nuevo Leon, and in dated contexts at Baker Cave (conventional ages that range from 8,910–9,180 years ago). Angostura points are even more common. Radiocarbon dating of late Paleoindian times comes from the Richard Beene site south of San Antonio, where Angostura anchors a long sequence. Site 41VT141 in Victoria County has yielded many middle and late Paleoindian tools, as have locales and beaches along the Coastal Bend. The Berger Bluff site in Victoria County has a deep (eight meters) Paleoindian occupation. A distinctive tool form, the bifacial Clear Fork tool, is found in middle and late Paleoindian sites. Detailed wear pattern analysis shows that these are wood-working tools.
The Archaic in southern Texas is usually thought of as an extensive array of triangular points and unifacial Clear Fork tools. However, with the emerging chronology available, this long period can be roughly divided into Early, Middle, Late, and Terminal eras. The Early Archaic is the poorest known, but marked by Central Texas stemmed point forms (Gower, Uvalde, and Martindale). The Early Triangular form (variously called Baird and/or Taylor) is most extensive from the coast into the interior. Even in those areas dominated by triangular points, like Webb County, Early Triangular can be distinguished technologically from the later Tortugas type. Such points date between 4,800–5,800 years ago. In the Coastal Bend, the McKenzie site has excavated examples of this point type. Another Early Archaic type, the Hidalgo point, occurs in the northern part of the Delta and into eastern Nuevo Leon. The highly distinctive Guadalupe tools may begin late in this era.
The Middle Archaic likely has Early Triangular in its earliest times, co-existing with (or followed by) Calf Creek points. Deep excavations (2.5–3 meters) at Gates-Rowell, a site at Choke Canyon, exposed charcoal-filled pits and concentrations of mussel shells and Rabdotus dating to 5,300 years ago.
The Late Archaic has scattered Central Texas point forms (Pedernales and Montell) along with Tortugas, Abasolo, and Refugio points. A distinctive form called “south Texas Shumla,” with 90 percent or more made on heat-treated chert, is likely contemporary. Choke Canyon yielded archeological data from excavated sites. At Possum Hollow, Middle and Late Archaic hearths and pits and abundant fauna (small mammals, reptiles, and fish) was recovered. The Lino and Boiler sites in Webb County indicate that Refugio points predate the Tortugas type, followed by Desmuke points and Olmos bifaces.
At the large Late Archaic cemetery of Loma Sandia, many burials were accompanied by a variety of temporally associated points (Tortugas, Abasolo, Carrizo, and Morhiss}, along with other artifacts such as tubular sandstone pipes, bone awls, and marine shell ornaments. The Aransas phase sites of the shorelines and bays of the Coastal Bend have distinctive shell middens and often used clay dunes for campsites. Among the typical artifacts are shell tools made of whelk (and conch) shell, including adzes, gouges, and hammers. Point types are Kent, Ensor, and Frio points, with Morhiss points occurring early in the phase. The heavy emphasis on coastal resources is seen in the diverse Gulf species such as oyster, scallop, whelk, conch, sunray clams, quahog, etc. The eroded clay dunes on Oso Creek also reflect intensive collecting of terrestrial Rabdotus snails. Several cemeteries were placed in clay dunes along Oso Creek, the best-known being the Oso cemetery, beginning in the Aransas phase. In the delta, Archaic lithic artifacts of Late Archaic age are found in the north part and northern fringes of the area. Artifacts include Tortugas and a few stemmed points, Olmos bifaces, and occasional Clear Fork tools.
The Terminal Archaic is typified by countless campsites with Ensor, Frio, Fairland, and Ellis points. Excavations indicate these are contemporaneous, sometimes with Marcos points. Tool forms include Nueces tools (beveled, bifacial, or unifacial), probably scrapers, and a variety of other bifacial tools and tubular sandstone pipes; painted pebbles from northern Zavala County are known. The Silo site in Karnes County is a cemetery of Ensor age. Distinctive artifacts are large corner tang artifacts, “cache” bifaces, and bone tools, while the Haiduk site burials (also in Karnes County) have Marcos points and corner tang artifacts.
The Late Prehistoric is first recognized in southern Texas with the appearance of Scallorn and Edwards points around 1,300 years ago. At a number of sites, Zavala points also occur. Certainly, the bow and arrow introduced improved hunting techniques. On Chaparrosa Ranch and Tortugas Creek, sites dated 500–300 years ago have abundant subsistence remains, including bison (rare), deer, antelope, rabbits, rodents, snakes, frogs, fish, mussels, and Rabdotus snails. Other material culture items are perforators, large and small end scrapers, and flake knives. Particularly notable across South Texas are shaft straighteners made on limestone cobbles; most have a series of longitudinal engraved lines, and others have a deep transverse groove across the lines. Most are broken from repeated heating. Settlement data for the Late Prehistoric include occupation sites adjacent to large creeks (at riparian/floodplain boundaries), with lithic reduction occurring at locales on the hills of the valley walls that were covered with Uvalde gravels.
Around 700 years ago, the Toyah phase appeared. Excavated sites point to the importance of bison in the Toyah sites, along with distinctive new tools: small end scrapers, perforators made on flakes, beveled knives, and bone-tempered ceramics. At sites like Berclair, Hinojosa, and Possum Hollow, a variety of other fauna occur, along with deer and antelope, and mussels. A bison kill and butchering site was found at Skillet Mountain No. 4 at Choke Canyon. In some areas, some Toyah traits are found, as at Chaparrosa Ranch, but with little to no emphasis on bison. The Tortuga Flats site has bison, deer, antelope, and a variety of other fauna, along with Scallorn, Perdiz, and Cuney arrow points. Unifacial tools, cores, and heavy choppers were present, along with bone-tempered ceramics.
In the Late Prehistoric, the Rockport complex is found in the Coastal Bend, while to the south, in the delta, is the Brownsville complex. Rockport complex peoples were surely the ancestors of the historic Karankawa. Fishing was important in the fall through spring, and in the summer, upriver hunting was important. Shoreline and clay dune campsites were used, and Rockport materials are found in Refugio County sites, in one case as a Rockport phase occupation on one side of a lake, and Toyah phase materials on the other. Rockport artifacts are typified by asphaltum-painted sandy wares, Perdiz, Scallorn, Bulbar Stemmed, Padre, and Fresno arrow points. A core-blade technology is present; other artifacts include marine and mussel shell ornaments and bone awls. Cemeteries are found at this time, such as the Odem, Redfish Bay, and likely, Blue Bayou sites. In some instances, Scallorn points have been found in skeletal remains as the cause of death.
The Rio Grande delta is known for the distinctive Brownsville Complex artifacts, sites, and cemeteries. Coastal and riverine food resources were presumably used; in addition, the beaches provided abundant shells for the manufacture of shell ornaments. These include whelk and conch disc beads, gorgets and pendants, and Oliva beads and tinklers. Numerous cemeteries are linked to the complex by the grave goods. Several cemeteries are in clay dunes, while others are along the Rio Grande. Whatever the cultural beginnings of the Brownsville Complex shell technology, it apparently became an intensive activity. There is clear evidence of shell ornament trade with the Huasteca. Trade goods from that Mexican culture include pottery vessels, jade and serpentine ornaments, and obsidian, the latter sourced to outcrops in Mexico.
Central Texas. Broad chronological intervals for prehistoric cultures in Central Texas include three major divisions; Paleoindian, Archaic, and Late Prehistoric. The Paleoindian period (ca. 13,400–ca. 9,800 years ago) reflects numbers of small and highly mobile groups. Limited data exists concerning their daily lives, as cultural features are scarce and rocks employed in cooking activities in later periods are absent. Structures have not been identified with the possible exception of rock pavements/foundations at the Kincaid site and Gault. These people did make portable art in the form of incised cobbles at sites such as Gault, Kincaid Shelter, and Wilson-Leonard.
The early Paleoindian period has two complexes represented by highly-crafted fluted lanceolate point types, Clovis followed by Folsom. The Clovis techo-complex had previously been viewed as the earliest period in Central Texas (ca. 13,400–11,900 years ago), with only a few occupations at Gault, Levi Rockshelter, Pavo Real, and Wilson-Leonard. Initially, populations were thought to have focused on hunting large, extinct megafauna like camel, horse, mammoth, and bison, but recent excavations demonstrate populations utilized diverse game resources along with many plant resources (grasses and tubers) for broad-spectrum foraging subsistence practices. Kill sites are not recognized here. Clovis artifacts reflect a distinctive knapping technology, including a distinctive core and blade technology. Tool functions include cutting soft tissues and silica-rich plants, hide and bone working, woodworking, and digging.
Immediately following Clovis, Folsom populations employed small, thin unnotched fluted lanceolate points. These were employed primarily to hunt now extinct species of bison (B. antiquus) as most other megafauna had disappeared. Lithic tool assemblages include ultra-thin bifaces, end scrapers, gravers, large ultrathin bifaces, and edge-modified flakes.
The late Paleoindian period is recognized through a variety of unnotched lanceolate forms (e.g., Barber, Golondrina, and St. Mary’s Hall). The stemmed Wilson point (e.g., Wilson-Leonard) is unique within this period.
The subsequent Archaic period lasted from ca. 9,830 to 1,250 years ago. This is subdivided into the Early (ca. 9,830–6,300 years ago), Middle (ca. 6,300–4,500 years ago), and Late (ca. 4,500–1,250 years ago) sub-periods and is distinguished by sites having smaller notched and stemmed dart points employed with the atlatl. With megafauna now gone, the Archaic subsistence pattern was diversified, hunting large and small game and gathering plants with the emergence of a broad-spectrum subsistence pattern at sites such as Richard Beene in the San Antonio area. Sites are more common and recognized through quantities of burned rock employed in diverse cooking activities through assorted features (e.g., earth ovens, boiling pits, discard piles, and large burned rock mounds). Large burned rock mounds reflect a specialized cooking strategy focused on exploitation of bulk roots and tubers like camas, onions, and sotol bulbs. Stone tool assemblages became more diversified with the addition of manos, metates, and abraders. Other stone tool types include gouges/Clear Fork tools, various bifaces, drills, Guadalupe tools, edge-modified flakes, and burins.
The earliest point type in the Early Archaic is the Angostura, a lanceolate form carried forward from earlier times. This was followed by a series of split-stemmed points (Hoxie, Gower and Jetta types), and then stemmed (Uvalde), and corned-notched (Bandy and Martindale) points. The increased number of point types may indicate population growth.
Andice-Bell-Calf Creek dart points dominate the beginning of the Middle Archaic. These were followed by un-notched triangular forms (Taylor and Baird types) and subsequently by stemmed Nolan and Travis dart points.
The Late Archaic is divided into six intervals based on multiple dart point types. In sequential order, they include Bulverde, Pedernales, Lange-Marshall-Williams, and Marcos-Montell-Castroville, followed by Ensor, Frio, Fairland, and ending with Darl. A broad subsistence pattern continued with reliance on multiple meat resources such as deer, pronghorn, bison, rabbits, fish, birds, and turtles. Diverse plants like onions, camas, walnuts, and grass seeds were also exploited. Cooking features represented by burned rock middens, slab-lined hearths, basin hearths, rock ovens, and rock-filled hearths, and refuse dumps remain common. Small cemeteries with individuals killed during conflict, as at the Loeve-Fox site in Williamson County reflect hostile interactions.
The Late Prehistoric period (ca. 1,250–375 years ago) saw the introduction of the bow and arrow and the use of pottery. The Austin interval from ca. 1,250–600 years ago was characterized by Scallorn, Granbury, and Edwards arrow points employed with the bow. Temporal overlap exists between the Austin interval arrow points and earlier Darl dart points, implying that previous technologies did not terminate abruptly and adoption of new technologies was gradual. Hunting and gathering continued during this interval, although hunting appears less frequent, with more reliance on fresh water mussels. Pottery is absent, with instead the continued use of rock cooking features such as basin hearths and ovens. Lithic assemblages are dominated by projectile points and edge-modified flakes, but drills, awls, perforators as well as Friday and Gahagan bifaces. Cemeteries are documented at some sites.
The prehistoric record ends with the Toyah interval (ca. 600–375 years ago). This interval has been extensively investigated and recognized through Perdiz and Cliffton arrow points and plain bone-tempered pottery, including well-made, thin-walled, Leon Plain ollas and jars. Stone tools (e.g., two and four-beveled knives, small bifaces, small prismatic blades, large end scrapers, flake drills, spokeshaves, gravers, choppers, edge-modified flakes, and hammer stones) reflect emphases on hunting and meat processing. Organic preservation is good with multiple bone tools (e.g., awls, antler tines, billets, incised ribs, and spatulas) and ornaments (e.g., beads and shells). Ground stone (e.g., one-handed manos, shallow metates, and abraders) reveals plant processing. Diet breadth is broad and dominated by bison, although deer, antelope and a vast array of smaller game were also procured.
North Texas. North Texas has been inhabited by aboriginal peoples from at least as early as ca. 13,400 years ago and perhaps substantially before that. Extensive Paleoindian occupations have been reported all across the North Texas area. These includes the occurrence of Clovis, Folsom, Plainview, Dalton, San Patrice, Pelican, Scottsbluff, and Angostura projectile points. Of particular note are the Clovis sites at Aubrey in Denton County and Brushy Creek in western Hunt County. Instead of an isolated projectile point find, these two sites represent seasonal occupations that have yielded a number of other Clovis-age tools, including blades, large bifaces, end scrapers, gravers, choppers, and hammerstones in association with Pleistocene mammal, reptile, and invertebrate remains.
The Early Archaic and part of the Middle Archaic period in the region is characterized by diagnostic Carrollton materials (dating ca. 7,000–5,500 years ago), including Split Stem (Gower), Carrollton, Trinity, Calf Creek Horizon (Andice, Bell), and Dallas points plus Clear Fork tools, Waco sinkers, and Carrollton double-bitted axes. The Late Archaic has Yarbrough and Elam points. Clear Fork tools, Waco sinkers, and Carrollton axes are not present by this time. Along the Elm Fork and main stem of the Trinity River, Archaic sites tend to be situated on the first terrace above the river floodplain.
During both the Paleoindian and Early Archaic periods, there appears to have been regular access to high quality cherts from the Edwards Plateau of Central Texas. Projectile points and large bifacial cutting tools are preferentially made from Edwards chert, while more utilitarian tools, including scrapers, choppers, and hammerstones, tend to be made from local Uvalde gravel quartzites. Beginning in the Middle Archaic and extending through the Late Prehistoric, there was an increasing reliance on local lithic raw material sources, and imported chert became much less common in sites across the region.
West of the East Fork of the Trinity River, the Late Archaic persisted until about 1,300 years ago. The Late Archaic was replaced by a Late Prehistoric culture that used straight to expanding-stemmed arrow points (Alba, Catahoula, Scallorn) and grog-tempered ceramics. This Late Prehistoric period persisted until about 750 years ago when it was replaced by another Late Prehistoric period with triangular arrow points (Fresno, Washita, Harrell), Perdiz points, and shell-tempered pottery (Nocona Plain).
East of the East Fork, the Late Archaic was replaced by a Woodland period culture whose artifact assemblage resembled the Fourche Maline culture of eastern Texas and Oklahoma, western Louisiana, and southwestern Arkansas. This culture has flat-bottomed grog-tempered ceramic vessels (Williams Plain) and Gary, Ellis, and Kent projectile points. By around 1,200–1,300 years ago, these groups began to make straight to expanding-stemmed arrow points. Contact with other indigenous groups to the east developed such that after about 1,000 years ago, Caddo ceramics from East Texas sites appeared in small numbers. The Late Prehistoric peoples along Richland-Chambers creeks at Bird Point Island as well as the East Fork of the Trinity River and its major tributaries also constructed oval rim-and-pit structures at their largest occupational sites. These pits primarily contain abundant fire-cracked rock, white-tailed deer, and turtle remains, and probably represent communal feasting areas. High status individuals were also buried in the rims of these structures. Both Woodland and Late Prehistoric sites tend to be situated on upland landforms overlooking stream floodplains as well as on elevated alluvial landforms along the stream floodplains.
After about 750 years ago to about 400 years ago on the upper Brazos and the Red River drainages, Late Prehistoric Plains Village cultures (including sites of the Henrietta phase) had permanent settlements and engaged in horticultural/agricultural activities and bison hunting. They are characterized by triangular (Fresno, Washita, Harrell) and Perdiz arrow points and abundant shell-tempered (Nocona Plain) pottery. Trade to the east with Caddo groups expanded; at the same time, new trade networks were developed to the west with the ancestral Puebloan peoples of North Central New Mexico. Artifacts of obsidian, ancestral Puebloan ceramics, and turquoise have been sourced to deposits from New Mexico, and shell beads have been linked to the Gulf of California. The local inhabitants of the region likely exploited the native stands of bois d’arc wood for bows to trade for these exotic items.
Northeast and East Texas. Northeast and East Texas were occupied from at least 13,000 years ago to about 200 years ago by Native Americans. The first Native Americans to use these lands were mobile hunter-gatherers (during the Paleoindian period, before 10,000 years ago), and these foragers continued to use the area for millennia during the Early, Middle, and Late Archaic periods (ca. 10,000–2,500 years ago). Their tool kit was based on a variety of stone projectile points of diverse forms and raw materials, along with ground stone tools (pitted stones, manos, and metates), scrapers, and other flake tools made from both local and non-local raw materials. About 2,500 years ago to 1,150 years ago, during the Woodland period, the prehistoric Native Americans living in the region began to settle down in small hamlets and camps dispersed across recognizable territories, and a few of the sites, such as Coral Snake and Jonas Short were used as burial mounds; these sites also have pit and oven features and concentrations of burned rocks used in hot rock cooking, but little evidence of structures is known. These Native Americans of the Fourche Maline, Mossy Grove, and Mill Creek cultures were ancestral Caddo peoples that made several kinds of plain and decorated pottery vessels, and they used Gary and Kent dart points and other chipped stone tools for hunting and other tasks. About 1,300 years ago, these groups began to make and use small-stemmed arrow points for hunting purposes.
The principal occupation of Northeast and East Texas in prehistoric times (up to about 450 years ago) was by a number of Caddo-speaking groups, including ancestors of the Hasinai, Kadohadacho, Nadaco, Nasoni, Hainai, Nacogdoches, and other tribes and alliances, that lived in settled horticultural and agricultural communities after about 1,150 years ago; they grew maize, beans, squash, and other plants. Their sites had wood and grass-thatched houses, arbors, ramadas, and outdoor activities, as well as sacred areas where the dead were buried in cemeteries. Burials of adults and children in these cemeteries were often accompanied by funerary offerings, among them ceramic vessels of many forms and decorative styles, quivers of arrow points, and large bifacial tools. The ancestral Caddo were well-versed in the manufacture from local clays of a variety of ceramic vessel forms (bottles, bowls, carinated bowls, compound bowls, jars, and effigy bowls) and decorative elements, and ceramic pipes of different styles: long-stemmed Red River pipes and elbow pipes made after about 600 years ago. Caddo ceramics are considered the best-manufactured and decorated ceramics in eastern North America.
These ancestral Caddo communities were composed principally of farmsteads and small hamlets, but larger villages with constructed earthen mounds (used as burial mounds for the social elite, platforms for temple structures, and to cover ritually-burned buildings) and plazas were situated along the major rivers and tributaries during much of the prehistoric and early historic era. Peaks in the prehistoric population of Caddo groups were reached after 600 years ago and did not decline because of introduced epidemic diseases until after European contact in the late seventeenth century. Caddo archeological sites in the region are known to be primarily located on elevated landforms (alluvial terraces and rises, natural levees, and upland edges) adjacent to the major streams, as well as along spring-fed branches and smaller tributaries with dependable water flow. They are also located in proximity to arable sandy loam soils, presumably for cultivation purposes with digging sticks and stone celts.
These Caddo groups were powerful theocratic chiefdoms that built earthen mounds—like the thirty-foot-tall mound with eight levels of burned and buried temple structures built after 500 years ago at the Hatchel site (see HATCHEL-MITCHELL SITE) on the Red River in Bowie County, and the platform mounds and burial mound (where the elite were buried) at the George C. Davis site (also known as Caddo Mounds State Historic Site) in Cherokee County—for political and religious purposes, functions, and rituals. They also traded goods and products extensively across the region with Caddo neighbors and with near and far-flung non-Caddoan speaking groups in Central Texas, the southern Plains, and as far as Cahokia on the Mississippi River in the Midwest, and developed intensive maize-producing economies.
Southeast Texas. The inland parts of Southeast Texas have been occupied by aboriginal peoples from as early as ca. 13,400 years ago during the Paleoindian period, but mainly sites are known that were used during Archaic (ca. 10,000–2,000 years ago), Woodland or Early Ceramic (ca. 2,000–1,000 years ago), and Late Prehistoric (ca. 1,000 years ago) periods. During historic period times, the region was home to the Bidai and Atakapa groups in inland areas as well as the lower reaches of the Sabine, Neches, and Trinity river basins.
The archeological record for both the Paleoindian and Archaic periods is marked by the recovery of a variety of chipped and ground stone tools from sites that were apparently occupied on short term and seasonal bases by hunting-gathering groups. Recent work in Liberty County has shown that there was a strong connection between the region and Central Texas, as many of the artifacts recovered of Paleoindian and Early Archaic age were made from Central Texas Edwards Formation chert. However, by the latter part of the Middle Archaic, there was a shift to the use of poorer quality and more local lithic resources (petrified wood and quartzite), which has been suggested relates to reduced group mobility and more tightly-defined group territories. The Paleoindian period lanceolate projectile points include Clovis, Dalton, San Patrice, Pelican, and Scottsbluff types, while the later Archaic period assemblages include a variety of expanding and parallel-stemmed forms in Early, Middle, and Late Archaic period occupations, and Kent, Ensor, Godley, and Gary point types were predominantly manufactured and used throughout the latter part of the Late Archaic and then in the Woodland periods.
The Mossy Grove Woodland or Early Ceramic period artifact assemblages in inland Southeast Texas have ceramic vessels, particularly sandy paste wares of the Goose Creek Plain series, including a few Goose Creek Incised vessels and other decorated vessels; and Lower Mississippi Valley Tchefuncte and Marksville wares occur in low frequencies in inland Southeast Texas Woodland period sites. Sites of this period are part of the Mossy Grove tradition. It has been suggested that the Mossy Grove culture sites represent settlements of ancestral Bidai, Atakapa, and Akokisa groups. By around 1,200–1,300 years ago, straight-stemmed to expanding-stemmed arrow points, such as the Alba and Catahoula types, began to be made by these groups, and by ca. 1,000 years ago, Late Prehistoric ceramic wares were made with grog temper. These grog-tempered ceramics, as well as later Perdiz arrow points, have stylistic and cultural affiliations with ancestral Caddo groups in the Neches/Angelina and Sabine River basins as well as with coastal Texas groups that made grog-tempered Baytown and San Jacinto ceramic wares. The grog-tempered ceramics have incised, incised-punctated, and punctated decorative elements, and engraved fine ware sherds (probably from vessels obtained from ancestral Caddo groups) have been reported from a few sites in the region. There was also communication with indigenous groups living in western Louisiana.
Both Woodland and Late Prehistoric sites (dating after ca. 1,200 years ago) tend to be situated on upland landforms overlooking stream floodplains as well as on elevated alluvial landforms along the stream floodplains. Archeologists have suggested that aboriginal groups in the inland parts of Southeast Texas had residential base camps with either a wide or limited range of onsite activities as well as task-specific extraction sites with low densities of artifacts that were the product of the procurement and/or processing of particular resources.
Upper and Central Coast. Archeological investigations along that portion of the Texas coast extending from the Sabine River to just below Corpus Christi began in the late 1880s and early 1900s when local residents and early archeologists became interested in the region’s prehistoric artifacts and shell middens and discussed their findings in local publications, newspapers, and nascent archeological journals. Eventually, with the culmination of more modern studies, beginning in the late 1930s and lasting through today, archaeologists have been able to piece together a picture of the various cultures that once occupied the region.
Although there are tantalizing hints of people entering North America 17,000 to 20,000 years ago, the earliest proof of occupation on the Texas coast comes from scattered occurrences of certain Early Paleoindian period projectile points, principally the Clovis type. A major concentration of such points, along with the remains of extinct megafauna (mammoth, mastodon, giant ground sloth, etc.) that presumably served as prey for these Paleoindian hunters, occurs as wave-deposited material at the McFaddin Beach site (41JF50) along the Gulf shoreline in Jefferson County. These early sites, although on the coast today, were well inland 13,000 years ago when sea level was much lower and the shoreline was 50 to 100 miles farther out in the Gulf of Mexico.
The same may be said for most subsequent occupations related to the Middle and Late Paleoindian and Archaic periods, since sea level did not reach its current level until about 3,000 years ago during the Late Archaic period. Nevertheless, some Early and Middle Archaic period sites are known from today’s coastal zone, including the ca. 7,000-year-old cemetery at the Buckeye Knoll site (41VT98) in Victoria County, and a ca. 8,000-year-old shell deposit at the Kendrick’s Hill site (41JK35) in Jackson County. Buckeye Knoll is especially intriguing due to the presence in several burials of potentially exotic artifacts (bannerstones and a large fish-tailed biface) suggesting possible contact with cultures in the Midwestern U.S. and Central or South America.
Once sea level stabilized ca. 3,000 years ago, occupations within the coastal area increased dramatically, as people took advantage of the tremendously rich array of marine and estuarine resources, particularly oysters, marsh clams, and numerous fish species such as red and black drum, sheepshead, and hardhead and gafftopsail catfish. Migratory waterfowl, including mallard and black ducks and sandhill and whooping cranes, also became available at certain times of the year and complemented a diet of terrestrial mammals like whitetail deer and bison.
Based on the seasonality analysis of shell midden sites within the Galveston Bay region, Lawrence Aten hypothesized that people lived in scattered bands in coastal areas during the summer and traveled inland during cooler months to congregate in large villages, thus taking advantage of different environments and the resources they offered. In contrast, Robert Ricklis hypothesized a different “annual round” in which people along the central coast, especially between Matagorda and Corpus Christi bays, occupied large winter villages during the cooler months of the year to take advantage of the “drum runs” that occurred when drum fish entered the bays and estuaries to spawn. During the summer months, these same people dispersed into small, family-sized groups and traveled inland to collect newly-sprouted plant foods and to hunt deer and bison. In both scenarios, the annual movement between coastal and inland locations could be traced back thousands of years into the Late Archaic, from the historically-known Akokisa of the Galveston Bay area and the Karankawa of the central coast.
One of the more salient aspects of Late Prehistoric residents on the central and upper coasts was development of a vibrant ceramic technology. In addition to the Caddo of East Texas and the Pueblo groups of Southwest Texas, the coast was one of the main regions where ceramics were manufactured in abundance. While the use of stone tools (projectile points, knives, scrapers, and perforators) had continued since Paleoindian times, pottery manufacture did not appear until about 2,500 years ago in the Sabine Lake area, until ca. 2,250 years ago in the Galveston Bay region, and even later, at ca. 1,000 years ago, along the central coast. The upper and central coasts also created pottery along two different trajectories. Pottery on the upper coast first was manufactured in a manner similar to Tchefuncte culture vessels from coastal Louisiana (during the Orange period in the Sabine Lake area, ca. 2,500–2,100 years ago and the Clear Lake period in the Galveston Bay region, ca. 2,250–1,750 years ago). These were poorly-made vessels in which non-sandy clay was removed from the ground and fashioned directly into pots and bowls. In the Sabine Lake area, strong influences from Louisiana continued as many vessels were manufactured in a similar, if not identical, fashion to subsequent Marksville culture pottery during what has been identified as the Big Hill period, ca. 2,000–1,550 years ago. Sites with components containing Tchefuncte and/or Marksville ceramics have been found on Adams Bayou in Orange County, Hillebrandt Bayou and Taylor Bayou, both south of Beaumont, and three locales on the edge of the marsh in southern Jefferson County. Black Hill Mound on Hillebrandt Bayou is particularly interesting, as it likely served as an Orange period burial mound similar to Tchefuncte mounds in Louisiana.
Eventually, the local tendency to use a sandier clay led to the development of the Goose Creek ceramic series. Goose Creek pottery became prevalent in the Sabine Lake and Galveston Bay regions and consisted of vessels with minimal decoration that usually was confined to the lip or rim area, mostly using incised designs done when the vessel clay was still malleable. Goose Creek ware became the main pottery during the Mayes Island, Turtle Bay, early Round Lake, and late Old River periods in the Galveston Bay area (ca. 1,750–900 years ago), then again after ca. 450 years ago. As time went by, tiny pieces of crushed sherds (called grog) were added to the sandy clay for some vessels, and different designs were incorporated as decoration. This grog-tempered pottery has been identified as the San Jacinto ceramic series and was prevalent in Galveston Bay during the late Round Lake and early Turtle Bay periods (ca. 900–450 years ago).
Pottery manufacture on the central coast also employed sandy clays, but often with the addition of other inclusions, such as grog, crushed bone, small bits of caliche, and/or a combination of these elements. This produced a vessel that was harder and more compact than typical Goose Creek ware. These vessels often were decorated with asphaltum (found naturally along the Gulf shoreline) applied as a paint to create straight lines, wiggled lines, dots, dabs, and blotches. This is Rockport ware, found at such sites as Guadalupe Bay (41CL2) adjacent to San Antonio Bay and McGloin Bluff (41SP11) on Corpus Christi Bay, and the people who created the pottery were part of the Rockport phase (dating ca. 800–350 years ago). Incised decoration also occurred on some Rockport vessels, as well as the practice of coating the vessel with a white or gray wash (presumably created out of crushed calcium carbonate nodules) prior to applying the asphaltum designs. Sometimes the interior of a Rockport vessel was entirely coated with asphaltum, which likely served as a waterproofing agent.
Overall, prehistoric cultures on the central and upper Texas coasts were significantly different than in other areas in the state. They produced ceramics unique to their two areas, employed settlement systems that incorporated an annual round between the coast and inland locations, and they utilized an adaptive strategy that relied heavily on marine and estuarine resources.