Watson Brake is among the oldest and best preserved Middle Archaic period (6000–2000 BCE) earthworks in the Southeast. The site was built by the Evans culture. The massive site contains eleven earthen mounds connected by ridges. The two curved ridges link groups of earthworks that are referred to as the north and south mounds. Together these features form a large enclosure measuring 984 x 656 feet. The site was occupied by fisher-hunter-gatherers who resided on the earthworks in all seasons of the year (although perhaps not year-round). Artifacts recovered provide information on fishing and hunting tools, cooking practices, foods consumed, and jewelry created at the site.

The Site

Watson Brake is located in northeast Louisiana, on the east edge of an alluvial terrace overlooking the swampy floodplain of Watson Brake, a minor tributary to the Ouachita River. This was an ideal location, with access to both upland and lowland resources. When it was inhabited, the site was about one mile west of the Ouachita River (which has shifted east since the site was occupied), and Watson Brake was a clearer, faster stream.

Avocational archaeologist Reca Jones identified the site and, along with Harvard graduate student John Belmont, sketched the first map of the earthworks in the early 1980s. At that time, Watson Brake was considered a Poverty Point culture (1700–800 BCE) site. It was not until the 1990s that the remarkable antiquity of Watson Brake, and twelve other Middle Archaic mound sites in Louisiana, was recognized. In fact, prior to this time, many archaeologists were resistant to the idea that mounds were built during the Archaic period (8000 BCE–800 BCE) at all, with the Poverty Point site being the looming exception. Long-standing theories about the way cultures must operate insisted that large-scale earthworks could not be built until hierarchical status systems developed, with chiefs (or someone with power) to compel labor. In addition, archaeologists theorized that a mound-building society had to be supported by agriculture because, it was thought, hunting and gathering could not provide enough of a food surplus to support mound builders, who were not hunting and gathering for themselves. Evidence from Watson Brake, the LSU Campus Indian Mounds, and other Middle and Late Archaic sites have toppled these ideas. Almost everyone now agrees that natural resources were abundant enough to support mound builders and that permanent status hierarchies were not necessary to get the job done.

Radiocarbon dates indicate that Watson Brake was occupied for some time before any of the mounds or ridges were constructed. This occupation was sporadic, with an initial occupation at circa 4000 BCE. Mound construction took place over six hundred to seven hundred years, from 3500 BCE to sometime after 2800 BCE. Many of the mounds exhibit multiple stages of construction, with extended periods of occupation on each stage. The largest, Mound A, is twenty-five feet tall and has a basal diameter of 230 feet. It may have had as many as seven stages of construction. On the other hand, the second largest mound, Mound E (thirteen feet tall) was constructed in one continuous episode. The other nine mounds range between 1.6 to 11 feet in height. In general, mounds close to the terrace edge were built in multiple stages, while those farther from the bluff were built in a single construction episode.

Evidence of occupation on the earthworks is widespread. Lacking pottery, the inhabitants heated gravel to cook with. Hot rocks were either used in hearths to provide a long-lasting heat source, or they were used to “stone boil” liquids in containers like leather bags or gourds. Heated gravel tends to fracture into small pieces, and pre-pottery sites sometimes contain huge amounts of the debris. At Watson Brake, the density of fire-cracked rock is very high on the earthworks but virtually absent from the level interior of the enclosure. This suggests that cooking was done on the mounds and ridges.

Watson Brake Artifacts

Artifacts at the Watson Brake site were made on-site from locally available resources. A distinctive and characteristic stone point at Watson Brake (and other Middle Archaic mound sites in the area) is the Evans point. Archaic period toolmakers typically created a single pair of indentations (notches) along the side of a stone point, which helped fasten the point to a spear. The Evans point has a curious second pair of notches on the point blade. These additional notches could have provided a more secure fastening, but they were not used before or after the Middle Archaic period, so they don’t appear to have been necessary. Debris from shaping points and other stone tools is common at the site.

Other characteristic artifacts include fired-earthen blocks, many of which are cubes about 1.5 inches square. These may have been precursors to the better-known Poverty Point Objects (ceramic cooking balls), which were used somewhat like hot rocks. The actual purpose of the blocks, however, is unknown. Bone tools were scarce, but included eight awls and a fishhook, along with one finished and five unfinished beads. An antler flaker (used for fine edge sharpening on stone tools) is the only antler artifact. These were found at the base of Mound B, where a shell deposit may have added to the pH of the soil, aiding in bone preservation.

People at Watson Brake were some of the first in Louisiana to make stone beads from local chert gravels. The gravel was flaked to about half an inch long, and then laboriously ground smooth, presumably with sand and water. Holes for stringing the beads were created using tiny stone “microdrills,” made from stone flakes or blades. To drill completely through a bead, the drill must have been hafted onto a wooden or bone shaft.

Microdrills were the most abundant type of stone tool at the site—154 finished and 93 unfinished drills (or drill bits) were recovered from the site. The majority of these (73 drills and 51 unfinished drills) were found on what appears to be a workshop floor in Mound D. Seven chert beads were found, although none of these came from the workshop floor. Only three of the beads were finished to the point of drilling a hole; all three beads displayed problems with the drilling and are considered “rejects.” Where did all the whole, drilled beads go?

Daily Lives

Archaeologists can reconstruct some of the diet of ancient American Indians by looking at food remains deposited in middens (trash piles). Unfortunately, preservation of organic material is rare at Watson Brake. However, a midden at the base of Mound B contained a lot of shell. This shell “sweetened” the soil and aided in the preservation of dietary items like food bone and seeds. These remains reflect the desirability of the Watson Brake site placement, which was at the intersection of lowlands and uplands. This “ecotonal” location provided immediate access to a diverse range of plants and animals; overall, fifty-six different kinds of plants and animals were represented. Upland species included turkey, fox squirrel, cottontail, pocket gopher, ruffed grouse, and white-tail deer. However, riverine and bottomland animal remains dominated the assemblage. Fish, mussels, and aquatic snails were the mainstays of the diet. These were supplemented with other lowland animals, including small mammals (beaver, raccoon, muskrat, and otter); waterfowl (mostly ducks and geese); aquatic turtle; water snakes; and amphibians. Wild plant foods included goosefoot, marsh elder, fruits, and nuts. The same diet persisted throughout the occupation of the site.

Food remains can also provide information on the seasons a site was occupied. Fish otoliths (ear bones) record seasons much like tree rings do. Those at Watson Brake indicate that the site was used in every season. Seasonal information from other animal and plant remains at the site support this conclusion. This does not necessarily mean that people lived at the site continuously throughout the year; there are a number of smaller sites in the Watson Brake area that might have been used for brief periods as well.

It is unclear why Watson Brake and other Middle Archaic mounds were built. The mounds were not used for burial. The first stages at Watson Brake were quite small, so, initially anyway, the mounds did not make a grand statement. Nevertheless, there appears to have been a formal plan for the site from the beginning of mound construction. Researchers looking at the way Watson Brake and other contemporaneous mound sites were laid out suggest that many Middle Archaic sites shared a system of measurement and geometry. This shared “blueprint” implies more technical skill and more intercommunication among Middle Archaic mound builders than researchers previously had thought.

The End of Watson Brake and the Middle Archaic

Mound construction at Watson Brake and at all other Middle Archaic sites ceased after 2800 BCE. It is possible that the El Niño–Southern Oscillation (ENSO) contributed to the decline. The latest mound construction date is near the end of a relatively calm ENSO interval. Subsequently, the weather was much more severe. There could have been a catastrophic flood almost every four years in the Lower Mississippi River Valley; sufficient dietary mainstays may not have been able to recover from floods of this frequency. After 2800 BCE, no mounds were built in Louisiana for more than one thousand years, until the Poverty Point site arose and brought mound building to new heights.