Update this timeline entry

History, Complex Hunter-gatherers, and the Mounds and Monuments of Crystal River, Florida, USA: A Ge

Source: History, Complex Hunter-gatherers, and the Mounds and Monuments of Crystal River, Florida, USA: A Geophysical Perspective #1521

Project ID

Chapter

Timeline title

Start date

End date

Filename received

Filename assigned

Content

This article was downloaded by: [University of Georgia]  On: 16 February 2015, At: 07:00  Publisher: Routledge  Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
 
The Journal of Island and Coastal  Archaeology  Publication details, including instructions for authors and  subscription information:  http://www.tandfonline.com/loi/uica20  History, Complex Hunter-gatherers,  and the Mounds and Monuments  of Crystal River, Florida, USA: A  Geophysical Perspective  Victor D. Thompson a & Thomas J. Pluckhahn b  a Department of Anthropology , The Ohio State University ,  Columbus, Ohio, USA  b Department of Anthropology , University of South Florida ,  Tampa, Florida, USA  Published online: 09 Apr 2010.  To cite this article: Victor D. Thompson & Thomas J. Pluckhahn (2010) History, Complex Hunter-gatherers, and the Mounds and Monuments of Crystal River, Florida, USA: A Geophysical Perspective, The Journal of Island and Coastal Archaeology, 5:1, 33-51, DOI: 10.1080/15564890903249811  To link to this article: http://dx.doi.org/10.1080/15564890903249811
 
PLEASE SCROLL DOWN FOR ARTICLE  Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.  This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/ terms-and-conditions  Downloaded by [University of Georgia] at 07:00 16 February 2015  Journal of Island & Coastal Archaeology, 5:33–51, 2010  Copyright © 2010 Taylor & Francis Group, LLC  ISSN: 1556-4894 print / 1556-1828 online  DOI: 10.1080/15564890903249811  History, Complex  Hunter-gatherers, and the  Mounds and Monuments of Crystal River, Florida, USA: A Geophysical Perspective  Downloaded by [University of Georgia] at 07:00 16 February 2015  Victor D. Thompson1 and Thomas J. Pluckhahn2  1Department of Anthropology, The Ohio State University, Columbus, Ohio, USA 2Department of Anthropology, University of South Florida, Tampa, Florida, USA  ABSTRACT  Crystal River (8CI1) is one of Florida’s most famous archaeological  sites. Yet, after over a century of investigations, its place in the history of  FloridaandthesoutheasternUnitedStatesisnotwellunderstood.Crystal  River is an important example, in terms of world archaeology, of a  monumental landscape constructed by complex hunter-gatherer-fishers  along the coast of the southeastern United States. Here, we present the  results of our remote sensing program at the site. This research includes  topographic mapping, a resistance survey, and ground-penetrating  radar transects over various architectural components at the site. These  data lend insight into the scale and rapidity of landscape modification at  the site, as well as information on the location of previous archaeological  excavations and modern disturbances. Further, the data illustrate the  potential of shallow geophysical survey to the investigations of shell  architecture.  Keywords ground penetrating radar, resistance survey, shell architecture, Southeastern  United States 
A central theme in the study of socio political complexity is the role that mon umental architecture plays in structuring social relations. Recently, archaeologists working in coastal and wetland areas around the world note the precocious appearance of  Received 20 Jan 2009; accepted 13 July 2009.  monument construction by hunter-gatherer fishers as well as their role and relation ship to coastal resources (e.g., David and Badulgal 2006; Gaspar et al. 2008; Russo 1994, 2008; Sassaman 2004; Thompson and Turck 2009). Thus, to understand how 
Address correspondence to Victor D. Thompson, Department of Anthropology, The Ohio State University, 4048SmithLaboratory,174W.18thAvenue,Columbus,OH43210,USA.E-mail:thompson.2042@osu.edu  33  Victor D. Thompson and Thomas J. Pluckhahn 
monuments structure and are structured by social relations within and among hunter gatherer societies, archaeologists must exam ine the construction histories of individual structures, as well as their relationship to one another across the landscape (Thomp son 2009). The southeastern United States, specifically Florida, is one region of the world where hunter-gatherers constructed monuments of shell and earth since at least the Late Archaic period (ca. 3000 to 1000 BC) (Milanich 1994; Russo 1994; Sassaman 2004, 2008). While well known in Florida, the scale and type of monument construction  Downloaded by [University of Georgia] at 07:00 16 February 2015  has not been widely published outside of local and regional journals (e.g., Florida Anthropologist, Southeastern Archaeology, etc.). Since many of the monuments in this area are constructed of shell, they should be of interest to archaeologists working in other regions of the world where shell was used as a construction material.  Based on the above points, this paper has two specific goals. First, we wish to spotlight the site of Crystal River (8CI1) as one example of a monumental landscape in the coastal zone of the southeastern United States. A subsistence base primarily depen dent upon hunting, gathering, and fishing supported the labor force that constructed these monumental works of shell and earth, as well as the concomitant ritual activities. Thus, understanding the role and history of these monuments will lend insight into the broader world history of monuments in coastal areas and socio-political complex ity in general among hunter-gatherer-fishers. Our second goal is methodologically based and is to illustrate the potential of shallow geophysical survey, particularly with regards to the investigations of shell architecture.  In what follows, we first present a brief description of Crystal River and the history of research at the site. Next, we define our theoreticalperspective,researchagenda,and how they articulate with the methods used in ourmostrecentresearchatthesite.Afterthis, we describe the results of our geophysical survey and topographic mapping of the site andoffersomepreliminaryinterpretationsre garding the construction of the mounds and monuments. Finally, we put forward some  suggestions for future research at Crystal River and the implications and importance of the site for the study of complex hunter gatherers in coastal environments.  THE MOUNDS AND MONUMENTS OF CRYSTAL RIVER: A BRIEF HISTORY OF RESEARCH  The Crystal River site is located in west central Florida and represents one of the most important Woodland (1000 BC to AD 1000) sites in the region (Figure 1). A quick perusal of the artifact plates in the sections concerning Crystal River from C. B. Moore’s work during the early twentieth century indicate why this site holds so prominent a place in Florida archaeology. Elaborately dec orated painted and incised pottery, worked copper, plummets, as well as a variety of shell artifacts, are found on these pages (Mitchem 1999; Moore 1903, 1907, 1918). In addition,thesite’shypothesizedconnections to Mesoamerica (Bullen 1966; Ford 1966), the speculation that it functioned as a so lar observatory (Hardman 1971; Williamson 1984), and finally its large shell and earthen monuments and stone stelae make Crystal River deserving of the designation a “fa mous” Florida site (see Bullen 1953; Milanich 1999:1).  The site, at a minimum, covers 6.9 hectares and comprises numerous architec tural features. Depending on how you count them (some of the mounds are actually complexes of architectural features), Crystal River contains at least six mounds: two burial mounds (Mound G and the Mound C-F com plex), and four platforms mounds (Mound A, H, J, and K). The largest of these eminences is Mound A at over nine meters tall. There is also a curvilinear shell feature more midden like than intentional architecture (Feature B). In addition to the mounds, three stone monuments are also located at the site.  C. B. Moore (1903, 1907, 1918) con ducted the earliest archaeological excava tions at Crystal River. Despite the limitations in his field methods and reporting, Moore’s excavations remain the most intensive work ever conducted at Crystal River and the 
34 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida  Downloaded by [University of Georgia] at 07:00 16 February 2015  Figure 1. Map showing the location of Crystal River and other important Woodland Period sites in the southeastern United States. Inset map shows Crystal River and related sites and their location along the Gulf Coast. 
baseline for the interpretation of the site (Weisman 1995:12–14). Most importantly, Moore produced the first map of the site, assigning the letter designations that are still used today for the major features of the site; however, this map does not mention the two earth/shell works known today as Mounds J and K, nor the presumed stelae (Moore 1903; Weisman 1995:12–13).  Moore’s 1903 work at Crystal River fo cused on the main burial mound complex,  which he labeled Mounds C-F. While work was conducted in all parts of the complex, he concentrated on the central sand mound (Mound F). The excavations here produced many of the exotic artifacts for which Crystal River has become famous. Returning in 1906 (Moore 1907; Weisman 1995:13), he con tinued excavating the main burial complex, focusing his efforts to the “elevation” or platform (Mound E) surrounding the central burial mound and to a lesser extent in 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 35  Victor D. Thompson and Thomas J. Pluckhahn 
the circular embankment (Mound C). Both locations contained a number of burials; however, most lacked the exotic artifacts of copper and quartz crystal found in Mound F. In 1917, Moore made his final visit (Moore 1918; Weisman 1995:13) and continued working in the circular embankment (Mound C), where he identified more burials, shell, and limestone rubble.  No archaeologists conducted investiga tions at Crystal River for more than three decades following Moore’s work. Neverthe less, the site was occasionally visited and de scribed in print (Weisman 1995:25). During  Downloaded by [University of Georgia] at 07:00 16 February 2015  the 1930s and 1940s, archaeologists began assessing the significance of the site through studies of its material culture, chronology, andapparentconnectionswiththeHopewell phenomenon, an archaeological complex concentrated in the midwestern states of Ohio and Illinois and renowned for its geometric mound centers, burial mounds, and elaborate and exotic artifacts. Greenman (1938) recognized the affinities between artifacts from Crystal River (as reported by Moore) and those from Hopewell sites in Ohio.  Willey (1948a; Willey and Phillips 1944) eventually would help partially clarify the temporal assignment of Crystal River pottery and identify it as belonging to the Deptford, Santa-Rosa Swift Creek, and Weeden Island complexes (Willey 1949). Further, he would suggest that Crystal River pottery is ancestral to Mississippian types (ca. AD 1000 to 1500) (Willey 1948b). Despite these revelations, the dating of Mound A was still unresolved as Willey’s investigations were limited to sur face collections of Mounds C and F (Weisman 1995:28; Willey 1949).  A complicating factor in these early attempts to situate Crystal River in the de velopmental sequence of the southeastern United States was the shortened chronology of the day and the related assumption that Weeden Island was contemporaneous with fully developed Mississippian (ca. AD 1000 to 1500) cultures in the interior. Indeed, as Knight and Schnell (2004:3–4) have pointed out, in the 1940s the Woodland and Mis sissippian sequence for the Gulf Coast was compressed into an interval of around 1500  years (ca. AD 500 to 1500), compared to the 3000 years it is now known to span. A second complicating factor in the dating of Crystal River was the presence of flat-topped mounds. The perceptions of archaeologists at this time was that such mounds dated predominantly or exclusively to the Temple Mound, or Mississippian period (Phillips et al. 1951). Only within the last twenty years has the existence of pre-Mississippian plat form mound construction become widely accepted (e.g., Jefferies 1994; Knight 1990).  To resolve questions about the relative ordering of the pottery series and mound construction at Crystal River, Hale Smith conducted limited work at the site in 1951 (Smith 1951; Weisman 1995:14, 28–29). His investigations included excavations in the midden area (Mound B), Mound H, Mounds C and E, and a surface collection of Mound A. Smith’s analysis suggested that at least a portion of the Mound C embankment was constructedlateintheWeedenIslandperiod, refiningWilley’searliertemporalassignment.  Shortly after this in 1951, Ripley Bullen initiated the first of several seasons of field work at Crystal River (Bullen 1953; Weisman 1995:28–29). These investigations included two stratigraphic excavations in the midden (Area B) to test his idea that the site (and par ticularly the burial mound complex) was in use for more than one period (Bullen 1951). Based on this work, Bullen postulated three periods of occupation and mound construc tion: Santa-Rosa Swift Creek (lower levels of Mound F), Weeden Island (the Mound E platform and Mound C embankment), and late Weeden Island or Safety Harbor (the upper levels of Mound F).  Bullen completed extensive excava tions at Crystal River in 1960 (Weisman 1995:37–38). Perhaps most significantly, these investigations included topographic mapping that led to the identification of two additional mounds and an extension of the midden area (Mound B) to the north of Mound A. Bullen described Mound J as an “irregularly shaped imminence of shell” (Weisman 1995:37), while Mound K was described as a flat-topped deposit resembling a small temple mound. Tests were excavated into these two mounds. Another test was 
36 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida 
excavated into Mound G, where 35 burials were dandified in a 10 foot by 20 foot trench. Finally, Bullen identified undisturbed burials in the Mound F platform and Mound C embankment. Unfortunately, the 1960 investigations by Bullen have never been thoroughly reported.  In 1964, as the site was being cleared  for the creation of the state park, two lime stone stelae were discovered south and east of the main burial complex (Bullen 1966; Weisman 1995:31–32). Bullen excavated the area around Stelae 1, a 2.15 m long upright irregular block that contains a pecked and  Downloaded by [University of Georgia] at 07:00 16 February 2015  incised representation of a human face. The identification of these and a third possible ste lae (Hardman 1971) have fueled speculation aboutconnectionsbetweenCrystalRiverand Mesoamerica(Bullen1966;Ford1966,1969). Contemporary fieldwork at Crystal River  has been limited. In 1985, Brent Weisman and Jeffrey Mitchem excavated core samples and two 2 × 2 m test units in the midden north of Mound A, with the goal of obtain ing samples from the Mississippian (Safety Harbor) component on the site (Weisman 1995:35–36). These excavations have never been thoroughly reported. More recently, Gary Ellis has conducted work at Crystal River in response to natural disasters and general park maintenance (Ellis 2004; Ellis et al. 2003).  THEORETICAL PERSPECTIVE AND  RESEARCH GOALS  As illustrated in the above discussion, despite numerous minor investigations, the Crystal River site remains poorly understood. The paucity of systematic, comprehensive inves tigations, coupled with the lack of adequate reporting, has confounded interpretation of the site and diminished the importance of Crystal River outside of Florida. Our research was designed to address these deficiencies and restore Crystal River to its rightful prominence.  The overarching theoretical framework  for this and our ongoing research is a modified form of historical processualism (sensu Pauketat 2001, 2007; Pauketat and  Loren 2005) or what we refer to as syncretic processualism. Such a framework seeks to understand the historical trajectories and genealogiesofagivenlocalareaandarticulate these local histories with larger regional ones with regards to the actions of both group and individual agents. Critics of strict historical processual approaches argue that such inquires are largely unsystematic and insufficiently generalizing (O’Brien and Ly man 2004). Indeed, interpreting landscape histories is not without difficulties, and de mands more than a superficial acknowledge ment of the spatial distribution of monu ments and other features across the land (cf. Tilley 1994). As such, a fine-grained temporal understanding of places, including theirdevelopmentalandconstructionhistory is required to fully understand their role in regional social relationships, as well as gain insight into overarching socio-political trajectories, so that comparisons between past and present societies can contribute to a global understanding of larger processes. Such a perspective differs from more strict versions of historical processualism, as one of its main goals is comparative. Further, methodologically it requires the formulation and evaluation of explicit research questions rather than inductive reconstructions of spe cific histories.  In order to implement our approach for the Crystal River site, our first research objective must be to understand the site’s developmental sequence. Specifically, we need to understand the construction history of many of the mounds and monuments at the site as well as their spatial relationships at a given time in the site’s history. This research goal, however, has one specific methodological problem. Like many large mounds sites around the southeastern United States, Crystal River is a state park and large scale excavations, especially in the mounds, are strongly discouraged.  Increasingly, archaeologists are turning to the use of geophysical methods and high resolution topographic mapping as a way to evaluate not only large-scale architecture, but also the space around these structures (e.g., Hargrave et al. 2007; Johnson 2006; Thompson et al. 2004). For example, the 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 37  Victor D. Thompson and Thomas J. Pluckhahn 
recent mapping work by Kidder (2002) and the geophysical investigations by Hargrave et al. (2007) at Poverty Point illustrate how such research sheds light into the nature of even some of the most well-known archaeo logical sites in the United States. Thus, while not an end in and of itself, such surveys provide a first line of inquiry into the nature of the built environment. We would also like to point out that when such techniques are clearly articulated with a theoretical perspective and research goals, what we term inquiry-based archaeogeophysics, the chance of a successful research project is  Downloaded by [University of Georgia] at 07:00 16 February 2015  greatly enhanced. We do not wish to dis abuse archaeologists of the utility of simple geophysicalprospection.Instead,wesuggest that recent advances in software allow us to use such techniques in more sophisticated and nuanced ways. In other words, we suggest that the use of this technology should be driven by research questions, not by the technology itself.  It is with these ideas in mind that we began our research at Crystal River. For our initial research, we have several major goals. Specifically we wanted to: first, assess the impact of various recent historic activities (house construction, mining, etc.) at the site prior to and since its development as an archaeological park; next, identify if the mounds evidence different construction stages and/or techniques; third, define spatial relationships vis-`a-vis the mounds and mon uments; and finally, evaluate the utility of geophysical techniques for this area.  Our research questions seek information that is vital for any archaeological project. However, our second and third research questions specifically provide the prelimi nary information that is required by our theoretical framework. In order to achieve these research goals we surveyed the site using a total station instrument, conducted a resistance survey over a large portion of the site, and finally, used ground-penetrating radar (GPR) over select areas of the site.  SURVEY METHODS  Descriptions of the geophysical survey meth ods used in this study can be found in  numerous recent publications (e.g., Gater and Gaffney 2003; Johnson 2006; Kuvamme 2003). While the reader is directed to these publications for more in-depth overviews, we provide a brief description of our sur vey methodology and how each instrument aids archaeological research in this specific context.  The research team conducted the resis tance mapping portion of the geophysical survey at Crystal River in 20×20 m collection grids set in using a total station. Collection gridsizesfortheGPRvaried.Thesegridswere often located on the tops of mounds that have irregular shapes. In these cases, rectangular collection grids of various sizes were shot in using the total station; however, many were not oriented along the axis of the site grid.  The use of total station mapping is now common in archaeology and does not warrant description here. Resistance survey and GPR are less well known and necessitate an overview. Both resistance survey and GPR work to identify local physical differences in the ground that may or may not indicate buried archaeological deposits, depending on soil characteristics and the nature of human induced disturbance (e.g., hearth, burned house, buried shell filled pit). Often the detected disturbances in the geophysical data are termed anomalies. Usually, archae ologists are hesitant to make interpretations based solely on geophysical data; however, in certain cases when the quality of the data is exemplary, initial interpretations may be put forth (see Thompson et al. 2004).  After the initial establishment of our baseline grid, we began topographic map ping, resistance survey, and the collection of GPR data. Thompson (Thompson et al. 2004) has described both GPR and resistance elsewhere. What follows is taken, in part, from these overviews.  Our resistance survey used a Geoscan RM-15 Advanced Resistance meter. This ma chine induces a known electrical current and detects the ease of flow or resistance. These values (measured in ohms) are recorded in a data logger along with their spatial location (Somers2006).Human-induceddisturbances such as pits, house basins, and shell middens, can either be of higher or lower resistance 
38 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida 
than the surrounding soil matrix. As we will show at Crystal River, where shell is one of the primary archaeological deposits, higher resistance values indicate, in part, this material. Similar resistance surveys at other sites in the southeastern United States and elsewhere, corroborate this observation (Dalan et al. 1992:51; Thompson et al. 2004). We collected data with the RM-15 using  a twin electrode array, which uses two pairs of current and potential electrodes (Somers 2006). One pair of electrodes is mounted to a mobile frame, which the archaeologists move along at evenly spaced intervals within  Downloaded by [University of Georgia] at 07:00 16 February 2015  the collection grid. In our case, data were collected at 50 cm intervals along transects spaced 1 m apart. The other pair of probes is inserted in the ground 20 to 30 m away from the grid. The probe spacing on the mobile probes is directly related to the depth the machine can detect below the surface. For our survey, we used a spacing of 50 cm for the mobile probes that allows the record ing of information up to a depth of approx imately 50 cm below the surface. In total, we surveyed 25 20 × 20 m collection grids, which resulted in coverage of 1 ha over the site’s core area.  ArcheoSurveyor was used to process  all resistance data following the procedure outlined in Gater and Gaffney (2003:104, fig. 49). The raw resistance data were reviewed, then a high pass filter was applied and the readings despiked for outliers. Finally, we enhanced the data for presentation by smoothing and interpolating the values.  The GPR survey at Crystal River was  used to complement the resistance survey. In contrast, to resistance data, GPR provides bothhorizontalaswellasverticalinformation regarding the distribution and thickness of archaeological features below the surface. Readers are directed to Conyers (2004, 2006) for a detailed explanation of GPR. Briefly, the GPR propagates radar pulse from a surface antenna. The waves then travel though the neargroundandarereflectedbackwhenthey encounter physical differences in the earth, which may represent buried archaeological deposits (Conyers 2006).  A Geophysical Survey Systems Inc. SIR  3000 GPR with a 400 MHz antenna was  used to complete the survey. GPR data were collected in transect lines that were spaced 50 cm apart; however, collection grid size varied as previously mentioned. Following collection, data were processed using GPR SLICE and GPR Viewer software. Radar data are presented either as individual profile slices or as a series of plan view slice maps showing how anomalies vary according to depth. Amplitude slices, here, are shown as both profile and plan view images that are based on the thickness of anomalies and the wave travel time. As space limitations restrict the number of images, in most cases the authors chose the GPR slice that best represents the phenomena under discussion.  RESULTS  Our fieldwork covered a sufficient portion of the Crystal River site to allow for the produc tion of the first comprehensive topographic map of the site, showing all the mounds, monuments, and relevant features (Figure 2). This map is based on over 18,000 elevation readings collected over a two-week period with three total stations. There are several differences between the location and the representation of certain features between our map of the site and previous maps. We will not dwell on these differences here, as it is the subject of another publication (Pluckhahn and Thompson 2009); however, we present our topographic map as a way of orienting the reader in space with regards to our geophysical survey vis-`a-vis the site’s architectural features.  Resistance Survey  The resistance survey covered approx imately one hectare of the site’s core (Figure 3). Our 20 ×20 m survey blocks were generally contiguous, allowing for a broad view of distribution of archaeological de posits at the site. Comparisons between the resistance survey and the topographic map show a striking correspondence. Thus, it ap pears that topographic relief corresponds to the higher resistance readings (i.e., the dark grey to black areas) on our resistance map. 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 39  Victor D. Thompson and Thomas J. Pluckhahn  Downloaded by [University of Georgia] at 07:00 16 February 2015  Figure 2. Topographic map of the Crystal River site showing all architectural features. The dark grey lines and blocks are modern park sidewalks and buildings. 
One of the interesting characteristics of Crystal River is that it lacks a clearly defined central plaza, in contrast with many other major Woodland ceremonial centers across the southeastern United States (e.g., Milanich et al. 1997; Pluckhahn 2003). Indeed, plaza oriented sites have considerable time depth in the Southeast, extending as far back as the Late Archaic (e.g., Kidder 2002; Russo 2004; Thompson 2007). Therefore, one of our main research goals was to evaluate if Crystal River did indeed have a plaza.  Plazas are important, as they are not merely empty spaces, but rather are “one of the central design elements of community planning and intrasite spatial organization” (Kidder 2004:515). For the southeastern UnitedStates,plazasareidentifiedasflatareas that evidence no domestic occupation and are usually, but not necessarily, flanked by some form of architecture (e.g., domestic or monumental) (Kidder 2004:515–516). The area southwest of Mound H and flanked by Mound G and the main burial mound 
40 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida  Downloaded by [University of Georgia] at 07:00 16 February 2015  Figure 3. Map showing the location of and correspondence between the resistance survey and topographic mapping at Crystal River. Darker (black) areas represent high resistance. 
complex (Mounds C-F) is suggested by Bullen (1965:225) to be a plaza. In order to test this idea we completely covered this area in our resistance survey (Figure 4). The resistance survey produced no obvious geophysical anomalies in this area except for a very small anomaly that appears halfway between Mounds G and C-F. This anomaly could possi bly represent the large posts that are typical features in plazas of later time periods in the Eastern United States (e.g., Cook 2008:39) or possibly another buried stelae. However, for  now, this is speculative Further, this does not mean that no archaeological deposits are present in the area that comprises the rest of the plaza, but rather they are not detectable by this machine. However, based on the available data, we suggest that our work supports the identification of this area as a plaza.  If, indeed, the area south of Mound H represents a plaza, it is interesting that it is offset from some of the other main architec tural elements at the site, specifically Mound 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 41  Victor D. Thompson and Thomas J. Pluckhahn  Downloaded by [University of Georgia] at 07:00 16 February 2015  Figure 4. Resistance map showing the featureless plaza. Dark areas represent high resistance areas associated with Mound G to the west and Mound C to the south. 
A (the largest platform mound at the site). While interesting, an offset plaza such as this is not out of the ordinary for Woodland sites in Florida. Fort Center, another Woodland center, also has an offset plaza (see Sears 1982). The placement of offset verses central places surely would have implications for structuring social relations as well as the historical trajectory of the site in general. We, however, only have space to note this aspect of the site; further comments and implica tions must await future investigations.  Throughout the course of our survey, we covered most of Mound G. This section of the survey is important as it reveals the excava  tions conducted by Bullen in the 1960s, the precise locations of which were previously unknown (see Figure 4). This information informs one of our research goals, to define the impact of past historic activities at the site and thus provides a context for Pluckhahn’s current collections-based research at the site.  We covered a number of other archi tectural features during our survey. These features included Mounds J and K, which are clearly defined in the resistance data (Figure 5). These features are highly resis tant and evidence well-defined semi-straight line boundaries from the surrounding ma trix, suggesting purposeful construction. 
42 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida  Downloaded by [University of Georgia] at 07:00 16 February 2015 Figure 5. Resistance map showing Mound J, Mound K, and the remnant of Feature B. 
Weisman (1995:60) notes that there was some speculation that these features may have been bulldozer piles of shell based on the idea that they were not located on Moore’s earlier maps. Given that these archi tectural features are so clearly defined and are not smeared, as one would expect with bull dozer piles, we argue, following Weisman (1995:60), that these are of Native American origin.  In addition to these mounds, the burial mound complex Mounds C-F was also cov ered (Figure 6). As this portion of the site was heavily excavated and then reconstructed, we did not cover the entire architectural complex. Our results, however, do show that  perhaps a small portion of the circular mound C remains intact.  The final insight provided by our resis tance survey is with regards to Feature B (see Figure 5), a long, curvilinear midden deposit. Modern construction activities prior to the establishment of Crystal River as an archae ological park heavily impacted this feature. Our goal was to evaluate the degree to which archeological deposits were present in this area. Based on our survey, it appears that there are several portions of Feature B intact; however, unlike previous descriptions and maps that show this feature as contiguous, deposits seem to be segmented and discreet. We suggest that this is due, in part, to the 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 43  Victor D. Thompson and Thomas J. Pluckhahn  Downloaded by [University of Georgia] at 07:00 16 February 2015  Figure 6. Resistance map of the Mound C-F complex. Topographic inset with lines shows the reconstruction of the mound based on Moore’s 1903 description. 
historic construction and modification to this area for a trailer park.  GPR Survey  We selected four areas for our GPR survey to evaluate if certain architectural fea tures remained intact, as well as to investigate the nature of construction activities at the site. Two of these areas are located along the tops of Mounds K and H, another located between Mound A and K near Feature B, and one more in the vicinity of where the ramp for Mound A should be located.  Mound H is an elongated mound with a ramp leading to the southwest into our  tentative plaza area. This mound, with its long, narrow summit and extended ramp, is in our opinion virtually unique for its shape in the southeastern United States. It is reminiscent of Prehispanic structures found in parts of Mesoamerica; however, we leave such speculations aside for now as our GPR survey is intriguing on its own without contact from afar.  We surveyed two collection blocks on Mound H, a western and eastern grid. The GPR results of Mound H re veal differential layering in the mound. In both areas surveyed, a highly reflective horizon is indicated between 10 and 20 nanoseconds (na) (ca. 45–50 cm) below the surface of the mound (Figure 7). Yet another 
44 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida  Downloaded by [University of Georgia] at 07:00 16 February 2015 Figure 7. GPR sample profiles from Mound H. 
highly reflective horizon is indicated in the western grid at around 30 ns (ca. 90 cm). This layer is also present in the eastern collection grid, but is less clear as other highly reflective anomalies are noted in some of the GPR profiles just above this layer. Although we cannot say for sure at this time, we suggest thattheseanomaliespossiblyrepresentstruc tural remains or features, perhaps limestone blocks, on a previous mound surface. On a final note, the easternmost portion of the GPR profiles from the western grid indicate a strong anomaly that extents from the surface  of the mound to a depth of over a meter. We suggest that this represents one of Bullen’s previous excavation units as it is in the general vicinity of his test in this mound.  Based on our GPR survey, we suggest that Mound H was constructed in at least three stages. We propose that the highly reflective layers represented in the GPR data represent layers containing higher quantities of shell and/or limestone boulders. This interpretation in based on our knowledge of shell layering at other shell bearing sites (e.g., Thompson et al. 2004) and photographs of 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 45  Victor D. Thompson and Thomas J. Pluckhahn 
Bullen’s excavation that clearly show layers of dense shell deposits in this mound. We argue that the layers that contain less reflec tive material are areas that contain greater amounts of sand than shell. Furthermore, the indication that these levels are present in both the west and east collection blocks support a view that, in terms of the length of the platform, Mound H was conceived as we see it today. Such information is important regarding the size and rapidity of architectural construction.  In contrast to Mound H, our GPR survey of Mound K provides an altogether different  Downloaded by [University of Georgia] at 07:00 16 February 2015  view of architectural construction (Figure 8). The radar profiles of this mound indicate that it is composed primarily of high reflectors, which we interpret to be high-density shell deposits. Thus, it seems that this mound was  constructed using a different technique or under different circumstances than Mound H. If the construction material is indeed mostly shell which was collected during one season, then this mound indicates a more rapid construction than Mound H. Future researchusingisotopicstudiesoftheshellfish should be conducted to evaluate this hypoth esis (see Thompson 2006 for methods).  Our GPR survey of Feature B further indicates that many of the architectural fea tures at Crystal River formed as a result of varying depositional histories (Figure 9). Our survey results suggest that the upper layers of this feature appear to be similar to what we might expect for a dispersed sheet midden; however, there are anomalies that suggest deeper subsurface deposits. In particular, we located a large basin-shaped 
Figure 8. Sample GPR profile of Mound K. The topographic inset shows the survey block. Note the difference between the GPR profiles from Mound H and K. Mound K most likely is comprised of fill that contains more shell than those of Mound H.  46 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida  Downloaded by [University of Georgia] at 07:00 16 February 2015 Figure 9. GPR results for Feature B vicinity. Topographic inset shows location of the survey block. 
feature in the southwest corner of our survey block. Regardless, this area stands in marked contrast to Mounds K and H and thus should be thought of as midden deposits rather than planned architecture.  Our final survey area was in the vicinity of the ramp for Mound A (see map inset of Figure 9). Our purpose here was to identify, if present, the remnants of mound A’s ramp. Nearly one third of the fill that comprised Mound A was removed during the recent  past for various construction projects, an un fortunately common early practice in Florida and many other parts of the world (Claassen 1998:81). While our GPR survey of this area did indicate several anomalies in the area, none suggest sufficient structural formality to indicate that they were part of the ramp. Further,moderndisturbancesinthisareaalso complicated our results. Future geophysical survey and testing in the area may reveal that some of the deposits do indeed represent the 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 47  Victor D. Thompson and Thomas J. Pluckhahn 
basallevelsofMoundA;however,fornowwe can only speculate as to the nature of these anomalies.  CRYSTAL RIVER, COMPLEX  HUNTER-GATHERERS, AND THE  MONUMENTS OF COASTAL RIVER  Returning to some of our earlier statements, we argue that the results of this geophys ical survey lend insight into the historical trajectories of monument construction and landscape use at the Crystal River site. Based  Downloaded by [University of Georgia] at 07:00 16 February 2015  on our survey we have identified differential construction techniques between some of the platform mounds at the site, such as Mounds K and H. Furthermore, these results indicate that Mound H was constructed in stages. We have also identified potential differences in the formation processes of other features at the site, such as Feature B which appears to be more of an accretional middenratherthanaplannedarchitecturalel ement. In addition, we have provided added support for the interpretation of the plaza; however, additional coring in the area will be necessary to verify this idea. While we now have many more questions than answers, we suggest that the quality of our data and the information gleaned from this work is, in part, because we began our survey with clear questions. Thus, our use of these techniques was explicit (what we termed “inquiry-based archaeogeophysics”). We argue that this is the most productive way to incorporate geophysics, as well as any other specialized method, into archaeological research.  The survey allows us to make some preliminary statements regarding the nature of ceremonial centers during the Woodland period in Florida and perhaps the south eastern United States more generally. As noted in the introduction to this paper, the mound construction at Crystal River must be understood in terms of the meaningful practice of individuals and groups. Southeast ern archaeologists often conceive of Wood land mound sites as ceremonial centers for single, autonomous, and rather homogenous social groups (typically individual lineages) (Milanich et al. 1997). To the contrary,  we believe that the scale and diversity of mound constructions at Crystal River argue for the social practice of multiple and varied social groups. As Dillehay (1992) has argued for the Mapuche in Chile, we suggest that the spatial layout of Crystal River—with its similarities to other prominent Woodland sites—served to facilitate participation in extra-local ceremonies and social networks in Florida and beyond. Indeed, participation in at least extra-local exchange networks is evident in the number and variety of non local, Hopewellian artifacts recovered from the site.  While the overarching similarities of Crystal River to other sites with offset plazas and platform mounds would have given it a sense of familiarity to outsiders, the tempo and scale of construction of these monuments would have been determined largely by the local group. Key architectural distinctions between Crystal River and other similar Woodland sites, such as the unique shape of Mound H and the presence of stelae, appear to have been intentionally integrated into the overall site plan, perhaps to underscore the uniqueness of the com munity and the ceremonial practices that took place there. Thus, on the one hand, modification of the landscape by local groups and individuals provided a familiar setting for outsiders coming to the site for ceremonies, while on the other hand, it also served to distinguish Crystal River from other regional centers and provide a unique experience for visitors to the site (see Dillehay 2004 for parallels in Formative Peru).  While not an end in itself, our survey adds to understanding the historicity of the site and marks a point of departure for future considerations of the place of Crystal River in the larger region. Thus, we view these data as the beginning of a long-term research pro gram aimed at understanding the sociality of themonumentsduringtheWoodlandperiod. Our future work will include excavations, as well as radiocarbon dating of the mounds to place them within their proper sequence at the site, thus helping to understand the social history of Crystal River.  In terms of sites with monumental works of shell and earth constructed by complex 
48 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida 
hunter-gatherers, the southeastern United States, and Florida in particular, has one of the highest densities of such structures in the world. Yet, few archaeologists outside of Florida, much less the southeastern United States, recognize this important aspect of the southeast’s archaeological record (see Weisman 2003). This is changing, partic ularly with regards to some of the work that has been done regarding Archaic Pe riod complexity (e.g., Russo 1994, 2004, 2008;Sassaman2004,2008;Thompson2007; Thompson et al. 2008); however, the full breadth and diversity of these sites is far from  Downloaded by [University of Georgia] at 07:00 16 February 2015  complete.Indeed,whileArchaicperiodmon uments are gaining more attention, the later elaboration of such traditions in coastal areas during the Woodland period remains grossly underrepresented in the broader literature outside the southeast.  Our survey and description of the  mounds and monuments of Crystal River could be easily mistaken for a description of an interior ceremonial complex supported by intensive agriculture. Perhaps, it is for this very reason that southeastern scholars have not explicitly framed sites like Crystal River in terms of hunter-gatherer studies. The key point, however, is that Crystal River is the result and actions of people who supported their labor force, derived their ideology, and emerged from a history that was rooted in the coastalenvironment.Thus,theprocessesand histories of the coastal regions of Florida and other areas of the southeast have much to inform the world regarding the emergence and role of monuments and monumental landscapes in coastal environments.  ACKNOWLEDGEMENTS  A number of individuals and institutions made this research possible. First, we thank Nick Robins (Park Manager, Crystal River Archaeological State Park) for his sup port.For permission to work at the park, we also thank Parks Small (Chief, Bureau of Natural and Cultural Resources), Dr. Ryan J. Wheeler (Chief, Bureau of Archaeological Research and State Archaeologist), and William Stanton (Archaeologist, Bureau  of Natural & Cultural Resources). We are indebted to Chris (Paula) Carpenter, Jamie Gridwain, Mike Petellat, and Leroy Smith, the staff of Crystal River State Archaeolog ical Park. Rich Estabrook of the Florida Public Archaeology Network provided cru cial logistical support. The University of South Florida Office of Research through the New Researcher Grant Program, in part, supported this work. Both the Uni versity of South Florida and University of West Florida Departments of Anthro pology provided additional support. The Bureau of Natural and Cultural Resources of the Florida Department of Environmen tal Protection provided space for us to camp. Our field crew Amanda Roberts, Nick Laracuente, Sarah Mitchell, Adrienne Sams, the USF Field School, and the UWF Field School all, ultimately, made the research possible. We thank Dr. Lawrence Conyers and an anonymous reviewer for their thoughtful comments and critique of this article. As always, the authors are solely responsible for all errors, omissions, and mistakes.  REFERENCES  Bullen,R.P.1951.TheenigmaticCrystalRiversite. American Antiquity 17:142–143.  Bullen, R. P. 1953. The famous Crystal River site. The Florida Anthropologist 6:9–37.  Bullen, R. P. 1965. Recent Additional Information [p. 10], in Crystal River Indian Mound Museum. Mimeograph prepared for the opening of the Crystal River State Archaeological site. On file, Ford Library, Florida Museum of Natural History.  Bullen, R. P. 1966. Stelae at the Crystal River Site, Florida. American Antiquity 31:861–865. Claassen, C. 1998. Shells. Cambridge: Cambridge University Press.  Conyers, L. B. 2004. Ground-Penetrating Radar for Archaeology. Walnut Creek: AltaMira. Conyers, L. B. 2006. Ground-penetrating radar. In Remote Sensing in Archaeology: An Explicitly North American Perspective (Jay K. Johnson, ed.):131–160. Tuscaloosa, AL: University of Alabama Press.  Cook, R. 2008. Sunwatch: Fort Ancient Develop ment in the Mississippian World. Tuscaloosa, AL: University of Alabama Press. 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 49  Victor D. Thompson and Thomas J. Pluckhahn 
Dalan, R. A., J. M. Musser, Jr., and J. K. Stein. 1992. Geophysical exploration of a shell midden. In Deciphering a Shell Midden. San Diego (Julie K. Stein, ed.):43–59. Orlando, FL: Academic Press.  David, B. and M. Badulgal. 2006. What hap pened in Torres Strait 400 years ago? Ritual transformations in an island seascape. Journal of Island and Coastal Archaeology 1:123– 143.  Dillehay, T. 1992. Keeping outsiders out: Public ceremony, resource rights, and hierarchy in historic and contemporary Mapuche society. In Wealth and Hierarchy in the Intermediate Area (F. W. Lange, ed.):379–422. Washington  Downloaded by [University of Georgia] at 07:00 16 February 2015  DC: Dumbarton Oaks.  Dillehay, T. 2004. Social landscape and ritual pause: Uncertainty and integration in for mative Peru. Journal of Social Archaeology 4:239–268.  Ellis, G. D. 2004. Storm Damage Assessment on Mound G, Crystal River State Archaeological Park, Crystal River, Florida. Gulf Archaeolog ical Research Institute, Crystal River, Florida. Submitted to Crystal River Preserve State Park, Crystal River, Florida.  Ellis, G. D., J. Dean, and R. Martin. 2003. Dis placed Midden Recovery Project, Crystal River Mounds State Park. Gulf Archaeological Re search Institute, Crystal River, Florida. Submit ted to Crystal River Preserve State Park, Crystal River, Florida.  Ford,J.A.1966.EarlyformativeculturesinGeorgia and Florida. American Antiquity 31:781–99. Ford, J. A. 1969. A Comparison of Formative  Cultures in the Americas. Smithsonian Contri butions to Anthropology 11. Washington DC: Smithsonian Institution Press.  Gaspar, M. D., P. DeBlasis, S. K. Fish, and P. R. Fish. 2008. Sambaqui (Shell Mound) societies of coastal Brazil. In The Handbook of South American Archaeology (H. Silverman and W. Isbell, eds.):319–335. New York: Springer.  Gater, J. and C. Gaffney. 2003. Revealing the Buried Past: Geophysics for Archaeologists. Wiltshire: Tempus.  Greenman, E. F. 1938. Hopewellian traits in Florida. American Antiquity 3:327–32.  Hardman, C., Jr. 1971. The primitive solar obser vatory at Crystal River and its implications. The Florida Anthropologist 24:135–168.  Hargrave M. L., T. Britt, and M. Reynolds. 2007. Magnetic evidence of ridge construction and use at Poverty Point. American Antiquity 72:757–769.  Jefferies, R. W. 1994. The Swift Creek site and Woodland platform mounds in the Southeast  ern United States. In Ocmulgee Archaeology 1936–1986(D. J. Hally, ed.):71–83. Athens, GA: University of Georgia Press.  Johnson, J. K. (ed.). 2006. Remote Sensing in Archeology: An Explicitly North American Per spective. Tuscaloosa, AL: University of Alabama Press.  Kidder, T. R. 2002. Mapping Poverty Point. Amer ican Antiquity 67:89–102.  Kidder,T.R.2004.Plazasasarchitecture:Anexam ple from the Raffman Site, Northeast Louisiana. American Antiquity 69:514–532.  Knight, V. J., Jr. 1990. Excavation of the Trun cated Mound at the Walling Site: Middle Woodland Culture and Copena in the Ten nessee Valley. Report of Investigations 56. DivisionofArchaeology,AlabamaStateMuseum of Natural History, University of Alabama,  Tuscaloosa  Knight, V. J., Jr. and F. T. Schnell. 2004. Silence over Kolomoki: A curious episode in the history of Southeastern Archaeology. Southeastern Ar chaeology 23:1–11.  Kuvamme, K. L. 2003. Geophysical surveys as landscape archaeology. American Antiquity 68:435–457.  Milanich, J. T. 1994. Archaeology of Pre columbian Florida. Gainesville, FL: University Press of Florida.  Milanich, J. T. 1999. Famous Florida Sites: Crystal River and Mound Royal. Gainesville, FL: Uni versity Press of Florida.  Milanich, J. T., A. S. Cordell, V. J. Knight, Jr., T. A. Kohler, and B. J. Sigler-Lavelle. 1997. Archaeol ogy of Northern Florida, A.D. 200–900. New York: Academic Press (Orig. pub. 1984).  Mitchem, J. M. 1999. The West and Central Florida Expeditions of Clarence Bloomfield Moore. Tuscaloosa, AL: University of Alabama Press.  Moore, C. B. 1903. Certain aboriginal mounds of the Central Florida west coast. Journal of he Academy of Natural Sciences of Philadelphia 12:440–494.  Moore, C. B. 1907. Crystal River revisited.Journal oftheAcademyofNaturalSciencesofPhiladel phia, Second Series 13:406–425.  Moore, C. B. 1918. The northwest Florida coast revisited. Journal of the Academy of Natu ral Sciences of Philadelphia, Second Series 16:514–581.  O’Brien, M. J. and R. L. Lyman. 2004. History and explanation in archaeology. Anthropological Theory 4:173–197.  Pauketat,T.R.2001.Practiceandhistoryinarchae ology: An emerging paradigm. Anthropological Theory 1:73–98. 
50 VOLUME 5 • ISSUE 1 • 2010  Mounds and Monuments of Crystal River, Florida 
Pauketat, T. R. 2007. Chiefdoms and Other Ar chaeological Delusions. Walnut Canyon, CA:  AltaMira Press.  Pauketat, T. R. and D. DiPaolo Loren. 2005. Alternative histories and North American ar  chaeology. In North American Archaeology  (T.R.PauketatandD.DiPaoloLoren,eds.):1–29.  Oxford: Blackwell Publishing.  Phillips, P., J. A. Ford, and J. B. Griffin. 1951. Ar chaeological Survey in the Lower Mississippi  Alluvial ValleyArchaeological Survey in the  Lower Mississippi Alluvial Valley. Cambridge:  Harvard University, Papers of the Peabody Mu  seum of American Archaeology and Ethnology  25.  Downloaded by [University of Georgia] at 07:00 16 February 2015  Pluckhahn, T. J. 2003. Kolomoki: Settlement, Ceremony, and Status in the Deep South, A.  D. 350 to 750. Tuscaloosa, AL: University of  Alabama Press.  Pluckhahn, T. J. and V. D. Thompson. 2009. MappingCrystalRiver:Pastandpresent.Florida  Anthropology 62:3–22.  Russo, M. 1994. A brief introduction to the study of Archaic mounds in the Southeast. Southeastern  Archaeology 13:89–92.  Russo, M. 2004. Measuring shell rings for social in equality. In Signs of Power: The Rise of Cultural  Complexity in the Southeast (J. Gibson and P.  Carr, eds.): 26–70. Tuscaloosa, AL: University of  Alabama Press.  Russo,M.2008.LateArchaicshellringsandsociety in the Southeast U.S. The SAA Archaeological  Record 8:18–22.  Sassaman, K. E. 2004. Complex hunter-gatherers in evolution and history: A North American per  spective. Journal of Archaeological Research  12:227–280.  Sassaman, K. E. 2008. The new Archaic, it ain’t what it used to be. The SAA Archaeological  Record 8:6–9.  Sears, W. 1982. Fort Center: An Archaeological Site in the Lake Okeechobee Basin. Gainesville,  FL: University Press of Florida.  Smith, H. 1951. Crystal River revisited, re visited, revisited. American Antiquity 17:  143–144.  Somers, L. 2006. Resistivity survey. In Re mote Sensing in Archaeology: An Explicitly  North American Perspective (J. K. Johnson,  ed.):109–130. Tuscaloosa, AL: University of  Alabama Press.  Thompson, V. D. 2006. Questioning Complexity: The Prehistoric Hunter-Gatherers of Sapelo Island, Georgia. Ph.D. Dissertation. Lexington, KY: University of Kentucky.  Thompson, V. D. 2007. Articulating activity areas and formation processes at the Sapelo Island shell ring complex. Southeastern Archaeology 26:91–107.  Thompson, V. D. 2009. The rhythms of space time and the Making of Monuments and Places during the Archaic. In Trend, Tradition, and Turmoil: What Happened to the Southeastern Archaic? (D. Hurst Thomas and M. Sanger, eds.):in review. New York: Anthropological Papers of the American Museum of Natural History.  Thompson, V. D., M. D. Reynolds, B. Haley, R. Jefferies, J. K. Johnson, and L. Humphries. 2004. The Sapelo shell rings: Shallow geophysics on a Georgia sea island. Southeastern Archaeology 23:192–201.  Thompson, V. D., W. D. Stoner, and H. D. Rowe. 2008. Early hunter-gatherer pottery along the Atlantic Coast of the Southeastern United States: A ceramic compositional study. Journal of Island and Coastal Archaeology 3:191–213.  Tilley, C. 1994. A Phenomenology of Landscapes: Places, Paths and Monuments. Oxford: Berg. Weisman, B. R. 1995.Crystal River: A Ceremonial Mound Center on the Florida Gulf Coast. Tallahassee, FL: Florida Department of State, DivisionofHistoricalResources,FloridaArchae ology Series, Number 8.  Weisman, B. R. 2003. Why Florida archae ology matters. Southeastern Archaeology 22:210–226.  Willey, G. R. 1948a. The cultural context of the Crystal River negative-painted style. American Antiquity 13:325–328.  Willey, G. R. 1948b. A prototype for the Southern Cult. American Antiquity 13: 328–30.  Willey, G. R. 1949. Archaeology of the Florida Gulf Coast. Washington DC: Smithsonian Mis cellaneous Collections 113.  Willey, G. R. and P. Phillips. 1944. Negative painted pottery from Crystal River. American Antiquity 10:173–185.  Williamson, R. A. 1984. Living the Sky: The Cosmos of the American Indian. Boston: Houghton Mifflin Company. 
JOURNAL OF ISLAND & COASTAL ARCHAEOLOGY 51