From Regional Center to Mound-Residential Compound (Phase 3)

Recent archaeological work suggests that people began moving away from Crystal River in Phase 3, which probably began between around AD 500 and 600 and lasted until sometime between AD 650 and 750, during the Late Woodland Period. Nevertheless, the site seems to have continued to serve as a ceremonial center. The village contracted to the area north of Mound A, which was expanded during this interval; perhaps the continuing presence represents a caretaker population or a compound occupied by a leader and his or her family. Some of the former residents of Crystal River may have moved the short distance downstream to Roberts Island, where settlement was initiated in this interval. Shifts in settlement such as this abandonment and collapse are typical of the Gulf Coast at this time, and may be at least in part a response to a more variable climate and lowered sea level associated with the interval known as the Vandal Minimum.

From Early Village to Regional Center (Phase 2)

The village at Crystal River expanded greatly in size and permanence in Phase 2, which began sometime between around AD 200 and 300 and ended by around AD 500. This growth may have owed partially to a rise in sea level associated with the warmer temperatures of the Roman Warm Period, which might have made life on the seaward islands more difficult. The exchange of Hopewell exotics faded in this interval, but the societies of the Gulf Coast appear to have witnessed a fluorescence, as indicated by the widespread exchange of Swift Creek pottery and Weeden Island pottery. Crystal River was peripheral to these pottery traditions, but it may have been an important nexus between these and the Glades tradition of southern Florida, specifically with regard to the exchange of craft goods manufactured from marine shell. The gulf coast fluorescence is also indicated by a heightened pace of the construction of mounds. At Crystal River, three small platform mounds were initiated in this interval, clearly differentiating it from its peers in the region.

A Center Emerges

Current radiocarbon evidence suggests that monument construction at Crystal River began sometime around 1000 BC, based on dating of human remains excavated from the circular embankment of the Main Burial Complex. Construction of the two burial mounds began a few centuries later, but likewise predates the earliest occupation of the village. Thus, the site began as a vacant ceremonial center, probably a place where small family groups dispersed on small islands in the surrounding landscape came together at certain times of the year. This pattern is typical for burial mound sites on the Gulf Coast, but Crystal River exhibits a unique degree of elaboration of architecture and burial treatments that suggest it had already emerged as a regional center. Likewise, the presence of large quantities of exotic Hopewell culture artifacts in a few burials suggest that certain people were already differentiated from others, perhaps owing to their roles as ritual specialists.

New Centers Emerge (Phase 4)

Phase 4, the final phase of occupation at Crystal River and Roberts Island, started sometime between around AD 780 and 870 and ended between 900 and 980, spanning the transition between the Late Woodland period and Mississippian period, and the beginning of the Medieval Warm period. Occupation at the former site waned during this interval, while Roberts Island emerged as a major ceremonial center that included three small platform mounds tightly clustered around a small plaza. The pattern is reminiscent of earlier spatial layouts at Crystal River, but other features--such as linear ridges of shell and a water court--suggest greater influence from the Caloosahatchee tradition to the south. Isotopic studies of oyster shells from Roberts Island display less variability in habitat relative to those from earlier contexts, perhaps consistent with ownership of particular resource locations. Possibly, larger corporate groups (such as matrilineages) began working cooperatively to target specific resource locales that they owned and managed for themselves, to the exclusion of other such groups.

Context

The importance of the archaeological site of Crystal River has been known since at least 1859, but it was excavations in the site’s burial mounds by C.B. Moore in the early twentieth that made the site famous among archaeologists. Later, Ripley Bullen provided additional insight on several of the other mounds and the village at Crystal River, and he and Adelaide Bullen supplied the first account of the nearby site of Roberts Island. Unfortunately, however, the excavations of both Moore and Bullen are underreported, and there has been little work at the sites using modern archaeological methods. Recent work under the auspices of the Crystal River Early Village Archaeological Project rectifies this with research program that combines the analysis of previous collections with minimally invasive new field work. The latter included detailed topographic mapping, coring and shovel testing, geophysical resistivity survey using ground-penetrating radar and electrical resistance, Bayesian modelling, and small-scale test excavations. As a result of these investigations, the sites are among the most thoroughly dated of any Woodland-period sites in eastern North America.

Crystal River and the Archaeology of Early Village Societies in the American Southeast (and Beyond)

In the archaeology of the American Southeast, the Woodland period (from around 1000 B.C. to A.D. 1050) is not conventionally understood as an interval marked by significant “firsts.” But it was marked by a dramatic change in the way people related to one another, as indicated by the earliest widespread appearance of sedentary villages, often associated with large-scale public works like mounds of earth and shell. Crystal River and Roberts Island are examples of these “early villages,” a term archaeologists have used to describe similar societies around the world, typically in reference to societies making a transition from hunting and gathering to farming. However, the people of Crystal River and Roberts Island faced many of the same social and ecological pressures. Early villages are important for what they can tell us about the role of cooperation, collective action, and conflict in the historical process and development of larger and more complex societies.

From Vacant Center to Early Village (Phase 1)

Radiocarbon dates from the lowermost levels in midden excavations suggest that people began to live at Crystal River sometime in the first century AD, during the Middle Woodland period. Phase 1 was relatively short, probably lasting only a century or two. Evidence from features and GPR suggest that the village was comprised of a scatter of small houses. Isotopic analysis of the shells of oysters suggest that the settlement was impermanent. Ceramics (including Deptford pottery) from this interval show disparity in temper suggestive of two communities of practice. It would thus seem that the dispersed groups that had congregated at the site to participate in ceremonies for hundreds of years began residing for longer intervals.

The Early Village at Crystal River in Broader Perspective

The pattern and historical process of early village development at Crystal River are contextualized with regard to regularities that might be typical of this transition on the Gulf Coast, focusing particularly on a handful of sites where archaeological investigations provide sufficient spatial and temporal control to reconstruct the lived experience of early village formation. These case studies permit the identification of eight postulates that can be evaluated for broader applicability to complex hunter-gatherer-fisher coalescent societies in other areas of the world.