The Hypsithermal is frequently referred to in archaeological works, but little hard data are usually offered in these reports. Pielou, in her excellent volume After the Ice Age (1991:269), has stated that the Hypsithermal has no absolute beginning and end, but rather was time-transgressive. For the purposes of this paper we have selected the period between 6500 and 3000 B.C. to define the Hypsithermal and provide us with a starting and ending point. These dates encompass the latter portion of the Early Archaic period through the latter portion of the Middle Archaic.
Many writers refer to the Hypsithermal as a period of great warmth and dryness. Droughty conditions caused the extension of prairies into the oak-hickory forests of the Central Mississippi Valley (Smith 1986:5). To the south, the weather patterns changed, inducing thunderstorms and lightning-started forest fires that replaced the predominantly hardwood forests of the Coastal plain with fire-tolerant pines (Delcourt and Delcourt 1979; Smith 1986:6). Unfortunately, for a major portion of the Southeast there are relatively few pollen core samples to provide evidence for this dramatic change. Yet these few samples are generally used to paint a picture of the Southeast as a whole. This paper will present a few of the more recently observed and, in our opinion, more compelling facts that suggest a major climatic shift during this time in the Southeast, especially when viewed from Mississippi. In addition to palynological studies, the Quaternary geologic history, floodplain evolution of surrounding areas, and archaeological data will be examined to discern if such a drastic climatic shift is reflected in the material culture of the appropriate time periods. We contend that the evidence suggests two things: (1) western and northern peoples migrated into the Southeast during the early part of the Hypsithermal; and (2) the shift away from the use of local gravels to tabular cherts, limonites, sandstones, and quartzites for tool manufacture can be directly correlated with the environmental effects of this period.
Pielou (1991:269–290) suggests that the Hypsithermal climatic shift varied greatly from one location to another. Furthermore, she asserts (Pielou 1991:270) that all biogeographical zones were shifted hundreds of kilometers northward. Annual rainfall is believed to have decreased, and Pielou describes barren areas with tremendous dust storms caused by high winds (Pielou 1991:271–287). Oak-hickory forests were gradually replaced by pine forests, and they in turn by prairie grasslands. The grasslands were replaced by ragweed, sagebrush, pigweed, and similar species (Pielou 1991:275). Aggravating the situation, the western prairies were devastated by huge wildfires and plagues of grasshoppers, driving out the herbivores (Pielou 1991:289) and therefore, we contend, the humans.
The bulk of Pielou’s volume discusses the northern United States and Canada, upwards into the Arctic regions. However, if the Hypsithermal had such an effect on the northern portion of the United States, it stands to reason that its effects must have been even greater in the Sunbelt. Some dramatic evidence of a climatic shift in the southern states (by now starting to sound biblical in its proportions and severity) is to be found in Saucier’s (1994) opus. In addressing the “pimple mounds” of Texas, Oklahoma, southern Missouri, Arkansas, and western and southern Louisiana, he presented an interesting explanation for these unusual features (Saucier 1994:162–165). Pimple mounds are typically 50 feet in diameter and 2 to 3 feet high. They occur in clusters and vary in number from 100 to 300 per square mile. They are found upon a variety of soils and landforms but not in low-lying, flood-prone areas. They approach the Mississippi River but are not found on the eastern shore. Finally, they do not touch each other, but occur as single, discrete mounds. Saucier believed this ruled out a geomorphological explanation and suggested territorial animal behavior as being responsible. Ants and/or termites were the suspected culprits. Furthermore, evidence suggests that these features were constructed during the Hypsithermal. However, recent research by Seifert and associates (2009:1–11) suggest that aeolian deposition in the middle to late Holocene was responsible for the pimple mounds, and that new data reinforce the concept of greatly decreased moisture and a shift to prairie vegetation during the middle Holocene (Seifert et al. 2009).
Additional confirmation of the Hypsithermal climatic shift comes from southeastern Missouri, where pollen core samples suggest a period of changing plant communities during this period (King 1981, as reported in Morse and Morse 1983:100–101). The lowest levels indicate the presence of a bottomland hardwood forest with open-water swamps. Beginning around 6700 B.C., the forest began to decline, being replaced by grasses. As time passed, grassland species began to dominate the assemblage, comprising approximately 85 percent of the species present by 5000 B.C. After this (by approximately 4500 B.C.) the grasses declined, and percentages of swamp species increased, although relatively dry conditions persisted until 3000 B.C.
Palynological studies in Florida and the coastal plain of Georgia and South Carolina indicate that the replacement of predominantly hardwood forests with pine occurred relatively slowly from north to south. “The expansion of pine to a new stability was achieved first in South Carolina by about 8000 B.P., last in southern Florida by about 4500 B.P.” (Watts et al. 1996:37). Furthermore, Watts et al. (1996:32) note that Florida (in fact, many parts of the Southeast) “today contains species representative of widespread prairie genera . . . which are isolated from the great prairies of the central United States, but suggest a biogeographic relationship at some remote time.” They suggest that these prairie areas are remnants of the oak/herb plant communities that were replaced by pine. If this were the case, it would have provided human and herbivore immigrants with an environment somewhat similar to the one being devastated behind them, lasting until 3000 B.C. or so in southern Florida. Brown (2003:15) notes that during the Hypsithermal, a prairie corridor extended from the Great Plains to the upper Atlantic coast as evidenced by disjunct species of insects, bison, and prairie chickens in eastern “prairies” in historic times. There is evidence for lowered water tables and the drying out of lakes and swamps during this time (Peacock 2008:71–73). For example, Peacock and Seltzer (2008) examined mollusks (freshwater mussel shells) from the accretional, Middle Archaic-period Vaughn Mound (22LO538) on the central Tombigbee River using a combination of taxonomic, morphometric, trace element, and light isotope data. Contrary to an earlier suggestion (Peacock 2008:95), they found that, on balance, these data suggest the Tombigbee River was “smaller during the Middle Holocene” (Peacock and Seltzer 2008:2562), indicating a hot, dry climate during this time (see also Delcourt and Delcourt 2004; Stein 2005).
The drier climate and sparser vegetation would have had another effect upon the environment. When rain did come, the desiccated and relatively denuded surface would have been subjected to tremendous erosion. This is supported by research on the evolution of the lower Illinois River floodplain at the Koster and Napoleon Hollow sites (Hajic 1981; T. Styles 1984; B. Styles 1986:148–151) and Joseph Schuldenrein’s (1996) geoarchaeological research on Mid-Holocene landscape history. These researchers agree that “the Hypsithermal was heralded by increasing sedimentation rates as erosional conditions were accelerated and the climate assumed a warming-drying aspect” (Schuldenrein 1996:24).
Such stream siltation would have buried gravel sources in the streambeds (Bense 1987:397). Not only would the gravel bars be covered with soil, but this soil would have become hard-packed and eventually covered with shrubs and grasses (David Dockery, personal communication 1996). An exception to this would be the pre-loess gravel in areas near steep runoffs, such as along the loess bluffs on the eastern margin of the Lower Mississippi Valley. We contend that the material culture of this time period supports the occurrence of such an erosional episode.
The first evidence we have of a movement of peoples from the West into the Southeast occurs around 7000–6500 B.C. with the appearance of the Cody complex. Scottsbluff points and Cody knives appear in Louisiana and Arkansas, and several are known from Mississippi (Brookes 2007:4–7; Jeter et al. 1989:89; McGahey 2000:80–84; Morse and Morse 1983:107). Most of the Scottsbluff points are made of non-local cherts believed to derive from Oklahoma and Texas (Schambach and Early 1982; Webb 1981:7). As these styles move from west to east, the Scottsbluff point is modified into a Hardin point, a common form in the Mississippi Valley. Also, the stemmed Cody knife loses the stem and becomes what is known as a Cobbs triangular knife (Overstreet and Peake 1981:182–186; Perino 1985:79). Almost all Hardin points and Cody/Cobbs knives are manufactured from local Citronelle Gravel. The modification of the form of Hardin and Cody artifacts is probably due to the smaller size and generally inferior quality of this raw material. The movement of a Plains bison-hunting culture into the Southeast at a time when the prairies were supposed to be expanding suggests that the Hypsithermal was having an effect. Further supporting this idea is the fact that few other artifactual remains are present, especially diagnostic projectile point/knives, that can be assigned to this period, giving one the impression that local populations have moved. The use of local Citronelle Gravels for 95 percent of the tools of this period suggests that gravel bars were still accessible and that trade or travel to western sources, as suggested by Schambach and Early (1982), was not occurring, at least not in the Mississippi Valley and eastern valley margins.
At approximately the same time, another point form appears in Mississippi. Pine Tree points are plentiful in northeast Mississippi and are the dominant point form of the terminal Early Archaic period (McGahey 2000:80–84). Pine Tree points are unusual in several respects. They are the only Early Archaic corner-notched point/knife form in the region that is not typically resharpened by alternate beveling. Alternate beveling is the predominant resharpening technique in the Southeast, but it is almost nonexistent in the Atlantic Coast region. Distributional and technological evidence suggests that the Pine Tree may be a late variety of Kirk corner notched. Another very unusual feature of the Pine Tree point type is that Kosciusko Quartzite was a preferred material. Kosciusko Quartzite is found in a narrow band in the north-central portion of Mississippi. This material has to be quarried, and it is the opinion of most researchers that thermal alteration is a must for this extremely tough rock to be worked at all. Kosciusko Quartzite is a distinctive greenish-gray, fine-grained material. It was avoided by all Archaic groups except the makers of Pine Tree points. In Mississippi, 51 percent of Pine Tree points are made of something other than local gravels (McGahey 1999:1–11). Kosciusko Quartzite makes up 39 percent of this total, with other exotic material comprising the remaining 12 percent. Kosciusko Quartzite was traded or exchanged up to 80 km from its source. At the Hester site in north Mississippi, Pine Tree points were found in a layer just above Hardin/Lost Lake points and are believed to date from 6500–5000 B.C. (Brookes 1979:109).
At present, few Pine Tree components have been excavated, but the idea of a southward movement of people at the onset of the Hypsithermal is certainly supportive of the argument that the Hypsithermal was a disruptive event in the lives of prehistoric people. Furthermore, the fact that an upland raw material, assiduously avoided by all other groups because of the difficulty in obtaining and working it, was not only being quarried, but actively sought, suggests that the Citronelle gravel beds were being silted over. Supporting this idea is the fact that of all point/knife forms in Mississippi, the Pine Tree is the most curated form, with many examples resharpened until most of the blade has been worn away (Samuel O. McGahey, personal communication 1996). Similarly on the Gulf Coast, Bolen points are the dominant Early Archaic form. There are several varieties of Bolen, and some appear to be earlier than others. Bolens are made from Tallahatta Quartzite 43 percent of the time, and from other quartzite 4 percent of the time, with the remainder, 53 percent, being fabricated from Citronelle Gravel (McGahey 1999:78–80). When the later Bolens are sorted out (the ones that appear to be Hardin/Lost Lake), the percentage of local gravel used drops to 33 percent.
Further evidence comes from the succeeding Middle Archaic period of 5000–3000 B.C. Numerous projectile point forms are present that are radically different from those of the preceding eras. Unlike earlier points, Middle Archaic forms are thick and broad. The fine pressure flaking of earlier times disappears. In the earlier part of the Middle Archaic, corner-notched forms are most common, while in the latter part, broad-stemmed forms predominate. A broad stem as defined here is usually in excess of 20 mm wide.
This period of time, especially ca. 4000 B.C., is when the Hypsithermal was at its maximum. At just this time there is a shift in raw material selection from local gravel to several other sources. In the north, Fort Payne chert is the preferred material, while in central Mississippi, Tallahatta Quartzite is the dominant stone. In south Mississippi, Middle Archaic points are sometimes made from a purplish sandstone quartzite and from limonite. This is a high-quality material composed of iron oxide cements with some polycrystalline quartz and small amounts of chert (Heinrich 1988:A-5). Although quarries of this material have been found in Alabama, it is believed that there are potentially many sources of this material throughout southeastern Mississippi and southwestern Alabama (Rebecca Lumpkin, personal communication 1995). Outcrops of this material have now been documented on Camp Shelby in south Mississippi (McCarty 2011). This material is not a purely occurring quartzite substance, forcing would-be users of this quartzite to search through a lot of poor-quality, friable sandstone to find this higher quality material (Rebecca Lumpkin, personal communication 1995).
As part of the research design for the Midden Mound Project on the Tombigbee Waterway (Bense 1987), several hypotheses were put forward to explain events in the Middle Archaic. One of these hypotheses dealt with the Archaicperiod shift to Fort Payne chert for biface manufacture. Prior to the Benton phase in northeast Mississippi, less than 5 percent of tools were made from Fort Payne chert, even though this fine-grained, blue-gray tabular chert is the finest raw material available within 150 km. During the Benton phase, 50–80 percent of all tools were made from Fort Payne chert (Johnson and Brookes 1988:60; Meeks 1999:35). This phenomenon, the almost exclusive use of Fort Payne chert to manufacture a specific point type (Benton), was first noted by Gerald P. Smith (1972) and was formally recognized shortly thereafter by Rafferty et al. (1980). Bense (1987:393) suggested hypothetically that the shift to the use of Fort Payne chert may have resulted from gravel bars being silted over during the Hypsithermal. Vaughn points (Atkinson 1974:132) are another type of broad, thick point from this time period. McGahey (1999:106–107) states that almost 100 percent of Vaughn points are made from Tallahatta Quartzite. Although there is evidence for floodplain gravel bars being buried in the Tombigbee River at that time, Bense (1987:397) rejected her own hypothesis because she thought that gravel would have been available in the tributary stream valleys. However, David Dockery, geologist with the State of Mississippi Geological Survey, has stated that the same factors at work in the Tombigbee would have been at work in the uplands (David Dockery, personal communication 1996). Thus, the shift to Fort Payne chert and other raw materials can be explained by the effects of the Hypsithermal.
Most researchers agree that both Tuscaloosa and Citronelle Gravel were procured from streambeds where all the sorting had already taken place. There are no known prehistoric quarry sites in Mississippi where gravels were being dug. Most quarry sites are upland sites where tabular material was being procured in Middle Archaic times (Johnson 1981; McGahey et al. 1992). The shift to what has been considered second-line material sources can be documented within the Middle Archaic.
The Benton phase of ca. 4500–3000 B.C. marks the period in the northern part of Mississippi where this shift takes place. In the southern portion of the state, Giliberti (1995) has studied the Beaumont site in the Leaf River floodplain with similar results. The Middle Archaic forms at Beaumont are made of local gravel only 46 percent of the time. The data suggest that the shift from local gravel to tabular chert began in the latter part of the Early Archaic and increased in the Middle Archaic. A serendipitous aspect of this shift to alternate lithics is that people were able to manufacture larger bifaces, because this tabular material is so much larger than the gravel cobbles. These large bifaces occur across most of Mississippi, and oversized blades cross-cut several types (Benton, Castroville, Turkey Tail, and Pickwick). At the Eva site in Tennessee, an oversized Benton blade was found at the pelvic area with a female burial (Lewis and Lewis 1961:115). Many of these Middle Archaic forms were knives, used by both males and females, and do not necessarily represent a male activity.
In addition to the shift to previously unacceptable (or unobtainable) lithic materials, something else unusual happened in the Middle Archaic. Exchange networks appeared over a wide area of the Southeast (Brookes 2004; Crawford 2003; Jefferies 1996; Johnson and Brookes 1987, 1988). Most of the artifacts from sites involved with the networks date to the Middle Archaic period with Cypress Creek, Pickwick, and Benton point types predominating. The former two point types are made from Tallahatta Quartzite, while the latter is made from Fort Payne chert (although a few specimens made of Tallahatta Quartzite have been found in south Mississippi). In the northern part of the state, Pickwick points made from Tallahatta Quartzite have been found in association with caches of Benton points, thus firmly dating Pickwicks to the 3700–3000 B.C. period in this area. Investigations by the Mississippi Department of Archives and History have yielded several large Middle Archaic sites in the Yazoo Basin with Middle Archaic points and classic Bentons made from Arkansas novaculite (Samuel O. McGahey, personal communication 1996). A more recently described cache from Mississippi consists of points and bannerstones, a classic Benton point made from novaculite, and a hypertrophic Johnson point made from Burlington chert (Brookes 2002b). The fact that these lithic exchange networks appear over a large area of the region at the same time suggests that a need for suitable raw material was present. The dust storms and erosion of the Hypsithermal, coupled with reduced stream discharge and velocity, took their toll upon the lithic resources of the region, causing people to modify their behavior accordingly.
As we mentioned earlier, many of these oversized blades appear to be large hafted knives rather than spears. In the course of examining collections of Middle Archaic points, we have noted a peculiar breakage pattern. Many of these points exhibit deep, conchoidal fractures on the edges of the blade. This type of breakage has seldom been noted on earlier points. Middle Archaic bifaces also have a much lower rate of impact fractures than do the earlier forms (McGahey 1999:88–138). This blade damage appears to be the result of the biface being inserted into an unyielding object and twisted. It is entirely possible that most or all of these so-called projectile points of the Middle Archaic period are indeed not points at all but knives, and that projectile points of this period were made from wood, antler, or bone.
Another change in the archaeological record manifests itself in the Middle Archaic period. Throughout the earlier periods, one of the most common artifacts was the unifacial end scraper. Unifacial end scrapers on Middle Archaic sites are nonexistent. At the Middle Archaic Denton site in the Yazoo Basin, none are reported in an analysis of several hundred tools (Connaway 1977:20–21). In fact, the only Middle Archaic sites that yield end scrapers are those containing earlier components. End scrapers were a major part of the toolkit of Eskimo peoples up until very recent times. The Eskimo used these tools primarily to process hides used in their clothing. When people came into America at the end of the last Ice Age, their clothing most probably resembled Eskimo-style clothing. With the retreat of the glaciers and the warming of the Hypsithermal, this type of clothing was not a necessity. Therefore, the disappearance of unifacial end scrapers is suggested to be yet another sign of the Hypsithermal.
If the environment underwent such radical changes during this time, we could also expect to see a shift in settlement patterns. Evidence from northeast Mississippi (Sparks 1987:50) indicates that during the Middle Archaic there is an increase in the use of prairie bottoms and a decrease in use of terraces and uplands. Similarly, Reams (1995:77) found that Middle Archaic sites in south-central Mississippi were seven times more likely to be located within one kilometer of water than were sites of other time periods. The Midden Mound Project of the Tombigbee found that the large base camps of this period were either in the floodplain or on the edge of the first terrace (Bense 1987:3). Peacock (1988) also found this to be true of Middle Archaic Benton sites. Similar site placement has been noted by other authors (Claassen 1996; Dye 1996). All evidence suggests that flooding was not a concern at this time period, although recent work by Kidder (2006) suggests the same cannot be said for the terminal Late Archaic period in the Lower Mississippi Valley.
The Middle Archaic is also the time period of the Shellmound Archaic (Claassen 1996; Ford and Willey 1941; Jennings 1952; Phillips et al. 1951). Although most shell mounds are found north of Mississippi, one has to wonder how much preservation has to do with this phenomenon. The Denton site is apparently located on an old channel. The deep, conchoidal edge fractures mentioned earlier were first noticed by Brookes on biface blades recovered from Denton, and we suggest that the breakage was caused by using the bifaces to open mussels. Dean Snow (1980:177–181) has been a critic of the idea of people subsisting primarily on shellfish, but numerous researchers have noted an increase in the exploitation of freshwater mussels during the Middle Archaic (Claassen 1996; Neusius 1986; B. Styles 1986:156). The evidence indicates that mussels became very important during this period over much of the eastern United States. Coupled with a shift in settlement patterns to the edge of streams, this fact suggests that people wanted to be as close to the water as possible. Several authors (e.g., Brown 1986; Brown and Vierra 1983) have suggested that aquatic enhancement of riverine environments pulled people to the bottoms rather than people being forced out of the uplands because of harsh conditions there. It appears to us that both factors were at play. As recently noted (Peacock and Jenkins 2010), there are thus far only two sites in Mississippi that have produced unequivocally Middle Archaic-period shell: the Vaughn Mound and the Trice site, both in the Tombigbee drainage. This suggests that something other than preservation bias is at work. Actually, this fact may feed into the idea that gravel bars—i.e., mussel habitat—were silted over.
A totally new type of site also appears during this time period (4500–3000 B.C.), mound sites (Russo 1994). These sites can be both single mounds, as at Banana Bayou (Gagliano 1967), or mound groups such as Watson Brake (Jones 1983). Most occur in Louisiana, but some were constructed in Mississippi. These sites represent the earliest mounds not only in the New World, but they predate any such constructions in the Old World as well. Archaeologists are currently in the process of attempting to determine the function of these mounds. Whatever the function, two things are immediately apparent when one considers these early mounds. These first mounds appear ca. 4500 B.C. and mound building ends around 3000 B.C. This time period brackets the most severe portion of the Hypsithermal (4500–4000 B.C.). The Hypsithermal appears to end rather abruptly around 3000 B.C., and so too does this early mound building. It is also of interest that mound building does appear again in the same region with the Late Archaic Poverty Point culture, but there appears to be a hiatus of some 13 centuries between these two periods. Thus the Middle Archaic mounds are not antecedent to the Late Archaic mounds of the Poverty Point culture, but rather a disparate event occurring some 1,300 years earlier. It can now be stated that there are at least four Middle Archaic mounds in Mississippi. We have already mentioned the Vaughn mound on the Tombigbee River, but this mound is likely more akin to the accretional “midden mounds” on that waterway rather than being an intentionally constructed earthwork. There is, however, a single mound (22LI504) in Lincoln County (Peacock et al. 2010) and a group of three mounds, the Grant Mounds (now destroyed), in what is now downtown Pascagoula (Brookes 2004:111). Botanical materials from the former mound produced a number of radiocarbon dates verifying that it “is an Archaic period construction well over 5,000 years old” (Peacock et al. 2010:364). The latter group of mounds yielded a cache of stone beads, including a classic Middle Archaic zoomorphic effigy form (Blitz 1993b:26; Crawford 2003:153).
In a paper delivered at the William G. Haag Honorary Symposium, Schambach (2005) suggested that Gulf Coast shell was moving north. This is certainly true for the Middle Archaic period as well as for Late Archaic Poverty Point groups. Schambach (2005) makes the point that the toolkit for manufacturing shell beads and gorgets is found at the three primary Poverty Point sites (Poverty Point, Jaketown, and Claiborne). A similar toolkit, along with prepared cores and microdrills, is found at several Middle Archaic sites, and while many of these sites exhibit stone bead making complexes (Hadley and Carr, this volume), shell bead manufacture could have been occurring as well. The exchange networks that move rocks have been discussed in an earlier paper (Brookes 2004). It was noted in that paper that what was often moving in the networks was not needed raw material but special objects: effigy pendants, beads, and bifaces as well as oversize bifaces. Effigy beads, probably all made in Mississippi, are one of the most widely dispersed of the objects moving in the exchange networks (Crawford 2003). These effigy beads are interpreted to be fetishes (Connaway 1977:118–129). Fetishes are powerful objects and are connected with ritual and magic. This ties in with the mounds and mound groups, which are most likely constructed for ritual purposes. I have noted the importance of effigy bifaces in dance ritual in Middle Archaic, Mississippian, and contemporary Creek groups (Brookes 1997:55–70).
In conclusion, if our vision of the Hypsithermal as an episode of climatic disruption is valid, we would expect to see an amelioration of these effects beginning in the Late Archaic period. While Fort Payne chert is the finest raw material available in the state of Mississippi, it was used extensively only during the Middle Archaic. After 3000 B.C., this material is almost never used. The same phenomenon holds true for the sandstones and quartzites of coastal Mississippi. While they are very poor as pertains to flaking quality, they were extensively used during the Middle Archaic. After this period, utilization of both these materials drops to less than 5 percent. There is continued use of Tallahatta Quartzite, but in a greatly reduced quantity.
According to Pielou (1991:269–290), after 3000 B.C. the climate began to cool and rainfall increased. This would have served to increase stream velocity and discharge, once again exposing the gravel bars. Johnson and Brookes (1988:53–63) noted a drop-off in the use of Fort Payne chert in the Late Archaic but were unable to explain it. The drop-off is not so much a factor of leaving an available source of quality chert, but of returning to a preferred (and more convenient, if lesser quality) source, the stream gravel. The exchange networks disappeared, and almost no raw material was exchanged for a period of 1,500 years. In addition, settlement patterns shift back to pre-Hypsithermal disruption locales (Reams 1995:83; Sparks 1987:50). Unifacial end scrapers, however, are still lacking in Late Archaic assemblages. Although the climate was cooler, the conditions were nowhere near those of the Late Pleistocene. In fact, in Mississippi end scrapers do not reappear until the Little Ice Age, post A.D. 1400, when they were likely related to the deerskin trade with Europeans.
So, in summation, one sees a number of interesting phenomena coincident with the Hypsithermal event in Mississippi and the Midsouth. Exchange networks moving ideotechnic artifacts, an otherwise inexplicable shift to heretofore unexploited lithic resources, site settlement shifts to areas previously avoided, mound and earthwork construction, and the suggestion of craft specialists all appear at this time. During the thermal maximum of 5000–4000 B.C. this behavior reaches its maximum intensity. With the end of the Hypsithermal at 3000 B.C., all these behaviors either disappear or revert to their previous states. After an absence of 1,300 years some begin to appear again. It is entirely possible that all of this is fortuitous. We think not, however. It is our opinion that the rise of cultural complexity in the Southeast was a long process that began with the “great warmth” and intensified as the Hypsithermal intensified. Cultural complexity arose as a response to a world that was rapidly changing and to ensure their survival, people changed with it. As is most always true when dealing with the unknown, ritual and religion move to the forefront. We feel there is a certain grain of truth in the old maxim, “There are no atheists in foxholes.” Faced with a changing environment and ecosystem, the likes of which are totally foreign to us, people did what they felt they had to do to survive, and this involved much more than changing hunting techniques and strategies.
We suggest that the Late Pleistocene was a time of successful expansion of prehistoric people into the Land of Plenty. However, the great warmth of the Hypsithermal period that followed was a time of intense stress. The people who had prospered during the Ice Age found themselves in a changing world that required profound changes in the structure of their societies.