Historical Archaeologies of the Caribbean: Contextualizing Sites through Colonialism, Capitalism, and Globalism 9780817392482

Citation preview

Introduction Contextualizing Caribbean His­tori­cal Sites through Colonialism, Capitalism, and Globalization Todd M. Ahlman and Gerald F. Schroedl

Archaeological studies of colonialism, capitalism, and globalization are intertwined (Hodos 2017a; Leone and Knauf 2015a; Feinman 2017) and reflect the interconnectedness of these concepts in the study of the last 500 years. This is especially true of the Caribbean, where the colonization and exploitation by European powers from the sixteenth through the nineteenth century created one of the epicenters for historic globalization and the expansion of capitalism through­out the world. This is not a new or revolutionary perspective by any means and probably understates the role that the Caribbean played in geopo­liti­cal contact, conflict, and exploitation. As Sidney Mintz (1985) suggested, the industrial organization of the Caribbean sugar plantation laid the groundwork for the social, po­liti­cal, and labor organization of the industrial revolution and the subsequent expansion of capitalism. Studies of colonialism, capitalism, and globalization are of­ten Eurocentric in nature, and postcolonial studies have moved beyond the study of Europeans and Euro-­Ameri­cans to take cross-­cultural and “bottom-­up” approaches (see chapters in Hodos [2017b] and Leone and Knaupf [2015b] for examples) that bring a broader analy­sis and understanding of the past. This in turn accounts for the exercise of agency within and between a diversity of cultural expressions and allows for a variety of disenfranchised groups to assert a distinct identity (Hawley 2015). The Caribbean is an ideal region to study colonialism, capitalism, and globalization, as the major players are European powers, sugar monoculture, and enslaved Af­ri­cans. Often left out of Caribbean histories is a broad range of actors and places with stories that show that there is a diversity of narratives beyond plantations that played important roles in colonialism, globalization, and capitalism. Colonization and exploitation of the Caribbean in the sixteenth century began with the search for riches in gold and spices followed by cash crops like tobacco and indigo in the early seventeenth century. These goods were quickly supplanted by sugar monoculture before the third quarter of the seventeenth century. In fact, the popu­lar history of Caribbean colonization, exploitation, globalization, and capitalism is most closely tied to sugar

2 / Ahlman and Schroedl

cultivation and the enslavement of Af­ri­cans. The Caribbean landscape and memory are replete with reminders of the plantation period because most islands are littered with the remains of sugar and other plantations that once made the region one of the wealthiest and contested in the world. David Watts’s research into the his­tori­cal geography of the Caribbean provides an excellent insight into how pervasive sugar was in the region. Watts (1987:337) found, for example, that by the 1770s approximately 88% of the cultivated land on St. Kitts was in sugar production with the remainder used for grazing or mountain land, and similar patterns were present across the Englishand French-­speaking Lesser Antilles. Small cotton, coffee, and spice plantations existed on some islands, but like St. Kitts, sugar was supreme and symbolizes the height of seventeenth-­, eighteenth-­, and nineteenth-­century globalization and capitalism. Dan Hicks’s study (2007) of St. Kitts plantations led him to conclude that the early period of development on the island was feudal in nature because the spatial arrangement of estates was transplanted from familiar European models of agricultural and landscape development. Although this familiar landscape allowed many early settlers to feel comfortable with their surroundings, the agricultural landscape of the Caribbean was capitalist in nature from the beginning with a focus on production and profits. Marco Meniketti’s landscape study (2015) of Nevis found that the island’s plantations were soon connected with shipping centers and road networks. The island’s villages, towns, and farms were founded as a response to the growth and location of the shipping centers and plantation support. Military sites were established to protect commercial interests rather than the residents’ interests. Meniketti’s study shows how capitalism drove the landscape formation on Nevis and likely did so on most, if not all, Caribbean islands. Cossin and Hauser (2015) argue that the spatial organization of efficiency and surveillance built into the Caribbean plantation landscape and exported to the industrial society of Europe played a crucial role in the rise of the Industrial Revolution. Sugar’s dominance of Caribbean society, agriculture, and trade resulted in a concomitant large labor force needed to work the cane fields. Caribbean historians and archaeologists are all too familiar with the role that indentured and enslaved labor played in providing the workforce in the sugar fields. Indentured servants were of­ten the first labor force employed on Caribbean coffee, tobacco, indigo, spice, cotton, and sugar plantations and are of­ten overlooked in his­tori­cal overviews because by the mid-­seventeenth century they had been replaced by enslaved Af­ri­cans who were forcibly brought to work in the sugar cane fields (Beckles 2011). Historian Barry Higman (1995) notes that by 1830 over 70% of the enslaved laborers in the British Carib-

Introduction / 3

bean outside Jamaica worked on sugar plantations and, even on Jamaica itself, over half of the enslaved laborers worked on sugar plantations. David Watts (1987) also shows that as sugar monoculture grew in the Caribbean so did the number of enslaved Af­ri­cans, while the Caribbean’s European population decreased. For instance, on Barbados in 1645 there were 18,130 white people and 5,680 enslaved Af­ri­cans whereas in 1833 there were 12,797 whites and 80,861 enslaved laborers (311). Similar patterns of population growth and decrease were also present in the French Caribbean, especially Martinique and Guadeloupe, the centers of the French presence in the Caribbean (Boucher 2011). The movement of people and the movement of goods from around the world to and from the Caribbean are indicative of the globalization phenomenon. Tamar Hodos (2017a) defines globalization as a “wide-­scale flow of ideas and knowledge alongside the sharing of cultural customs, civil society, practices and the environment” (4). Hodos acknowledges that globali­ za­tion is asymmetrical in the rate and impact of these flows with continual and complex changes in the direction of development and interactions. As a process, globalization has led to complex similarities and awareness in self and placeness in the modern world. Using a postcolonial approach, his­tori­cal archaeologists first turned to the creolization process of how Amer­indian, European, and West Af­ri­can cultures interacted and developed in the modern world. This has been refined through consideration of hybridity or the manner in which transformation is effected through negotiation of the colonizers and colonized. The complexities of cultural interactions, creolization, and hybridization are acutely evident in the Caribbean where global cultures and nations interacted and changed in response to global economies and armed conflicts. The Caribbean’s sugar economy, history, and culture are of­ten the primary foci of his­tori­cal and archaeological research at the expense of research on other aspects of the region’s development. As historian Verene Shepherd (2002) notes, “Caribbean socioeconomic history displayed a rather totalizing tendency, with historians focusing on the sugar sector and virtually ignoring other agricultural sectors” (5). Boucher (2011) echoes this sentiment and points out that many of the smaller French and Dutch islands were not directly tied to the sugar plantation model. Shepherd (2002) further states that “the study of the experiences of property owners who were not members of the sugar-­planting elite is among the most underresearched areas of Caribbean history” (6). The same sentiments can be stated for archaeological research in the Caribbean, especially regarding the British islands. Chapters in the volume Island Lives, edited by Paul Farnsworth (2001) express the need for archaeological research to focus beyond the plantation while si-

4 / Ahlman and Schroedl

multaneously not ignoring the role that sugar plantations played in Caribbean history. Although he focuses on the British Caribbean, David Watters (2001) notes that urban, military, contact-­period, freedman, and European (e.g., white) sites are poorly represented in Caribbean research and hold great potential for future research venues. Delle’s work (1998) in Jamaica and Armstrong’s research (2003) in the Virgin Islands were among the first to meet some of these deficiencies with explicit examinations of the post­eman­ci­pa­ tion period. Previous studies of urban contexts include Smith and Watson’s work (2009) in Barbados and Barka’s findings (2001) on St. Eustatius. More recent research has brought additional focus to the postemancipation period, for example in St. Lucia and Nevis (e.g., Meniketti 2015; Seiter 2016), and also to poor white and mixed-­race communities in Barbados and Montserrat (e.g., Reilly 2016; Ryzewski and Cherry 2015). Although much research remains centered on the lives of workers, this volume and other recent publications (see Bates et al. 2016) broaden the inquiry to account for the diversity in the Caribbean his­tori­cal archaeological record. The archaeologies of colonization, capitalism, and globalization of­ten focus on worldwide and regional scales, but Gary Feinman (2017) advocates a multiscalar approach to these studies. Large-­scale studies of­ten miss the subtleties of localized impacts and contexts by focusing too much on social and material inequality and ignoring the negotiated space of cultural development (Hauser 2011a; Horning 2011). It is important to examine islands of different nationalities and a variety of site types to better understand the negotiated material and social inequalities in the historic Caribbean indicative of colonialism and capitalism. It is also important to take a postcolonial approach to examine these factors from a local perspective that includes a variety of his­tori­cal and archaeological viewpoints from which to construct different narratives of Caribbean cultural complexity. Following Feinman’s multiscalar recommendation (2017), this volume examines plantation, frontier, freedman, urban, military, and postemancipation sites from Antigua, Domi­nica, Martinique, Nevis, St. Eustatius, St. Kitts, and St. Thomas. The following chapters cover a wide range of time periods and social situations, demonstrating the Caribbean’s diversity that is seldom explored in archaeological and his­tori­cal research. Furthermore, we embrace the view now prevalent among his­tori­cal archaeologist that situates their discourse within the broader unifying contexts of colonialism, capitalism, globalism, and the postcolonial critique (Croucher and Weiss 2011; Hauser 2011a; Leone and Knauf 2015a). Each essay in this volume is intended to expand the range and depth of this dialogue by exploring new topics, new materials, and new interpretations. The chapters begin by examining the frontier situation on Domi­nica followed by studies on a freed-

Introduction / 5

man site on St. Eustatius, urban sites and trade on St. Thomas, St. Kitts, and Nevis, a look at slave and postemancipation slave villages on Antigua and Martinique, and then a review of institution and military sites on St. Kitts and Nevis. In the first chapter, Stephan Lenik notes that Caribbean frontiers were present in many different ways and forms over several centuries. Frontiers occurred on all Caribbean islands, but their study is nearly invisible in the research of Caribbean archaeology and history (Deagan 1995; Lenik 2009a, 2010). Examining the frontier situation of French settlement on Domi­nica, considered a Neutral Island and safe haven for the Amerindian Kalinagos, Lenik shows at two sites how Amerindians adapted to the European presence by situating their settlements in places that afforded them protection as well as surveillance points from which to monitor European activities and movements. Lenik notes that French Jesuits on Domi­nica, which was not formally settled by Europeans until 1763, established not only churches but also plantations to help fund religious and personal projects. Lenik’s research shows how a capitalist-­colonial foothold was achieved in Domi­nica, reminding Caribbean his­tori­cal archaeologist not to exclude Amerindians from regional culture formation in the early colonial period. This study may encourage similar research on other islands and provide one avenue for Kalinago people to participate in the construction of a Caribbean his­ tori­cal narrative. West Indian historian Elsa Goveia (1965) suggests that over 50% of the white population across the Caribbean lived in towns and urban areas. Her research shows, for example, that in 1788 over 80% of the white population on St. Kitts lived in towns, and Welch (2002) believes that a large proportion of the Caribbean free black population lived in urban areas by the end of the eighteenth century. R. Grant Gilmore III’s chapter on the Congo Free Black Village on St. Eustatius (Statia) provides a rare look into not only the lives of free blacks but also of a village established and occupied by free blacks. As Gilmore notes, Statia is a Dutch island that once was one of the most important capitalistic and globalization centers in the New World. The island, especially the warehouses and dwellings in the urban area of Oranjestad, have been the focus of archaeological investigations since the 1970s, making Statia one of the most intensively studied Caribbean islands. Gilmore shows that through­out the period of slavery and into the postemancipation era the Free Black village inhabitants of one particular housing lot built a series of structures. The earliest is comparable with structures occupied by enslaved Af­ri­cans, and the latest is similar to many of the Dutch-­ occupied houses on Statia. The lot’s occupants, however, maintained a use of space reflecting traditional West Indian/Af­ri­can cultural traditions. The

6 / Ahlman and Schroedl

material culture reflects European-­style consumerism of the time but also strong adherence to Caribbean/Af­ri­can customs and traditions—all within the norms of Dutch colonial culture. Gilmore’s analy­sis contrasts sharply with research that is singularly focused on European and enslaved Af­ri­can lives on Statia as well as across the Caribbean, and there is no doubt from Gilmore’s account that free and enslaved Af­ri­cans found numerous avenues for participating and thriving in Statia’s capitalist economy. With the introduction and expansion of the plantation economy, urban centers quickly developed and grew across the Caribbean. Port Royal, Jamaica, or Havana, Cuba, for example, are commonly known West Indian cites, but because of their size and complexity these places are more the exception than the rule. Smaller villages and towns like Jamestown, Nevis, and Charlotte Amalie, St. Thomas, that formed at coastal anchorages and shipping centers are more typical of Caribbean urban centers. These towns and villages were microcosms of capitalism and globalization where warehouses held locally produced sugar, coffee, spices, and other commodities before being shipped to North and South Ameri­can and European markets as well as markets, stores, and mercantiles that supplied island residents with European and Ameri­can goods. The town and urban centers were also home to the merchants, doctors, lawyers, clergy, and others who supported and profited from their connections to the plantation network. Some villages, towns, and cities have a long history, while others, like James­ town, Nevis, reflect the fortunes and distress of the sugar trade and slowly vanished from the landscape. Jamestown, however, has a larger place in Nevisian lore because oral accounts and scant written records concerning the town maintain that it was the victim of a tsunami and was swept into the ocean. Carter L. Hudgins, Eric Klingelhofer, and Roger H. Leech explore the fate of Jamestown, and by dispelling some of the postcolonial myth making about the site, contextualize the town in two related ways. Most obvious is that Jamestown played a vital role in the development of Nevis’s colonial plantation enterprise, making it a key element in both the economic and the po­liti­cal life of the colony. Second is how the town’s architecture and its material culture were altered as the island’s economy declined from demographic collapse and geopo­liti­cal conflicts, while experiencing subsequent building resurgence as the economy improved in the nineteenth century. The sugar trade represents an external economy that, as mentioned, dominates the his­tori­cal studies of Caribbean economies. Since the early 2000s, archaeologists have moved beyond the sugar-­based economy and have begun to focus on island economies and the trade of locally made pottery. The role of Afro-­Caribbean ware, a pottery produced and exchanged by enslaved and freed Af­ri­cans, has become a criti­cal aspect in the study of internal is-

Introduction / 7

land economies because it is a direct tie to enslaved Af­ri­cans and a tangible marker that survives in the archaeological record. As Todd M. Ahlman, Gerald F. Schroedl, Barbara J. Heath, R. Grant Gilmore III, and Jeffery R. Ferguson show in chapter 4 on the trade of Afro-­Caribbean ware, pottery production and exchange were not limited to internal economies and reveal important information on intra-­and interisland trade in which enslaved and freed Af­ri­cans were engaged. From studying Afro-­Caribbean ware from 24 sites on seven islands and 23 clay samples from five islands, they show that pottery was widely transported through­out the Caribbean and that people had differential access to these pottery wares. This makes it clear that enslaved Af­ri­cans participated in the capitalist economy as they manufactured, distributed, and sold these wares, and that at least in the north­ern Leeward Islands, Nevis likely was the center of this industry. The influence of European merchants is seen through­out the Caribbean; however, the material culture of these people is poorly known. In their chapter on the Bankhus site on St. Thomas, Christian Williamson and Douglas V. Armstrong describe the material record of a Danish colonial merchant property. They show that the material culture of urban properties in the Caribbean of­ten has a complex history tied to changes in residents, owners, and hurricanes and that merchant properties represent complex urban contexts where people of different ethnicities and socioeconomic standing of­ten lived in close proximity. The material culture from the Bankhus illustrates the intersection of colonialism, ethnicity, and consumerism in a Caribbean urban setting. Although this volume primarily focuses on nonplantation contexts, we would be remiss if worker villages from sugar plantations were not presented. The term worker village is used here because the sites examined in Antigua and Martinique cover both slave and postemancipation eras. Diane Wallman and Kenneth G. Kelly examine five occupation areas from the Habitation Crève Cœur, an eighteenth-­to nineteenth-­century sugar plantation on Martinique, to address slave and freedman sharecropper subsistence and purchasing habits. Samantha Rebovich Bardoe studies the slave village and postemancipation laborer village at the Green Castle Estate on Antigua. These chapters show the similarities and differences in how slavery and the postemancipation periods were perceived by French and British Caribbean governments and elites. Wallman and Kelly’s assessment finds intrasite differences in access to goods and foodstuffs as well as diachronic changes in ceramic use that reflect greater free­doms in purchasing power, especially in the postemancipation era. Bardoe found similar patterns on Antigua, but her research suggests differences in how the government and the freedmen themselves adjusted to life in the postemancipation period. Both chapters demon-

8 / Ahlman and Schroedl

strate that enslaved Af­ri­cans, despite the restrictions imposed on them, found ways to successfully participate in the capitalist market economy and how plantation owners encouraged them and benefited from the cost savings of provisioning their slaves. Emancipated slaves, as a result, were better positioned to participate and profit once wage labor conditions were established. The multinational nature of the Caribbean islands of­ten meant that Caribbeans were involved in the conflicts and disagreements of their European colonizers, which meant that to protect themselves on many islands people constructed extensive and of­ten massive forts and armaments along the coasts. Forts, such as Charles Fort on St. Kitts and Fort Charles on Nevis, protected the towns and villages where commodities were stored before shipment and guarded the ports and anchorage where cargo was placed on ships. Charles Fort on St. Kitts was established in the 1670s by the British to protect the sugar port and anchorage of Sandy Point, a valuable shipping harbor in the British sugar world that also happened to be adjacent to the French-­ occupied lands. Charles Fort played an important role until it was eclipsed by the fortifications at the nearby Brimstone Hill and was occupied until the 1850s. In 1890 the fort was repurposed as the leper asylum later known as the Hansen Home. Of the numerous leper asylums in the Caribbean, St. Kitts’s Hansen Home was one of the largest and longest-­used facilities. As Gerald F. Schroedl and Todd M. Ahlman show, the history and archaeology of the Hansen Home are representative of late nineteenth-­and early twentieth-­century institutions devoted to the care of lepers. In their chapter they describe the buildings and fort/asylum landscape that are representative of military and institutional attitudes toward health and confinement in the eighteenth to mid-­twentieth century on St. Kitts. Their evidence suggests that attitudes toward the leper asylum “inmates,” as they are called in his­tori­cal records, are consistently aimed toward easing the effects of institutionalization through housing, meals, tasks, and gardening. The repurposing of Charles Fort was consistent with postemancipation health care and demography as societal organization became less plantation centered. Perhaps more importantly, their essay has relevance for the St. Kitts postcolonial narrative of contemporary social status and health care. Edward González-­Tennant and Diana González-­Tennant take a new approach to the study of Fort Charles on Nevis. Fort Charles was established in the seventeenth century to protect the Charlestown anchorage, which at the time was an important harbor because of Nevis’s role in the sugar trade (see also chapter 3 by Hudgins, Klingelhofer, and Leech). While Fort Charles played an important role in Nevis’s history, its place in the Nevisians’ collective memory is not well positioned. The González-­Tennants discuss how they have forged a relationship with the Nevis His­tori­cal and Conservation

Introduction / 9

Society to incorporate local interests into their research. The focus of their chapter is on the development of new digital recordation and interpretation techniques that aim to bring the history and importance of Fort Charles to a wider local and international audience. As the González-­Tennants point out, the future of Caribbean research will only be meaningful in the long term once archaeologists begin incorporating the research interests of local groups within their own research interests. Their technological approach to site interpretation allows the Nevisian pub­lic and culture heritage interests to construct a variety of site narratives from which to interpret their own colonial and postcolonial history. The volume concludes with Paul Farnsworth’s review and discussion of each chapter. His remarks tie together how each contribution provides new insights into the roles urban, military, institutional, and plantation sites played in the Caribbean’s his­tori­cal development. Farnsworth closes his summary with important recommendations for the future of Caribbean area studies by addressing some of the differences in research interests of North Ameri­ can and European archaeologists and the interests of local residents. His advice on working closely with local communities should be heeded by all archaeologists working in the region. The essays in this volume examine diverse geographical, chronological, and archaeological contents and contexts. The contributors explore aspects of social and economic differences, relating mostly to the transformation from slavery to emancipation when former slaves, their former masters, and colonial governments were compelled to create patterns of economic relationships, social orders, and cultural understandings that had never before existed in the Caribbean. The unfolding of this process has significance for contemporary Caribbean societies and is fundamental to today’s vision of a post­colonial consciousness. These essays contribute to the emphasis on the intersection of colonialism, capitalism, and globalization with the postcolonial critique that is shaping his­tori­cal archaeology in the region. In this way the essays both expand and complement research elsewhere through­ out the Caribbean.

1 Kalinagos and Catholics in Domi­nica before 1763 Archaeology and History of Caribbean Frontiers Stephan Lenik For the colonial Caribbean, frontiers have primarily been approached according to their economic aspects, as territories that have not yet been, or are in the process of being, developed for industrial agriculture and administered as formal colonies. The lure of potential wealth extracted from the natural resources and fertile soils of island frontiers attracted European nations, trading companies, and planters, who introduced enslaved and indentured labor regimes. This refashioned most of the region into a major contributor to the Atlantic trade and a source of wealth for the metropole. When studying frontier sites his­tori­cal archaeologists have explored questions of landscape, interaction, and exchange at fifteenth-­ and sixteenth-­ century Spanish frontier towns in Hispaniola (Deagan 1995), in the east­ern Caribbean at Kalinago settlements (Honychurch 1997; Lenik 2012a), and at early colonial forts and plantations founded by British, French, Dutch, and Danish settlers (see Gilmore, chapter 2 herein, and Schroedl and Ahlman, chapter 8 herein) (Hauser and Armstrong 2012; Hauser and Kelly 2011; Hicks 2007; Klingelhofer 2008; Lenik 2010, 2012a, 2012b; Ryzewski and Cherry 2015:363–367). Where archaeologists directly engage with the idea of the frontier, it is economic exploitation that fig­ures most prominently, such as Lesser Antillean “frontier societies” (Hicks 2007:26–33), the “agro-­industrial frontier” and related social relations in places like Nevis (Meniketti 2015:51–52), or permeable frontiers revealed by ceramics reflecting circumvention of trade restrictions (Hauser and Kelly 2011). A similar tendency is evident among some historians, with a “frontier era” in early French colonization of Martinique and Guadeloupe (Boucher 2008:138–143), a Lesser Antillean frontier attracting immigrants to engage in exploitative “frontier activities” (Pérotin-­ Dumon 1999:152–156), the expanding “sugar frontier” of English Barbadian planters (Roberts 2016), or Suriname as “the last frontier” satisfying thirst for new lands (Candlin 2012). Undoubtedly, economic exploitation profoundly shaped the Caribbean islands as the setting of the world’s most ancient colonies (Mintz 2010), and many archaeologists are guided by this

Kalinagos and Catholics in Dominica / 11

narrative by tracing the advance of intensive economic exploitation into frontiers that are developed into colonies. This chapter shows how his­tori­cal archaeology may provide a more comprehensive understanding of Caribbean frontiers to reveal the range of dynamics among people who inhabited or passed through these spaces, besides or in addition to their economic attributes. Some starting points for such a discussion already exist. For example, Sidney Mintz points out that while plantations and slavery were sensible developments in certain Caribbean contexts, concurrently there were internal frontiers where small farmers could succeed in alternative lifeways in interior countrysides, mountainous regions, or liminal parts of lesser islands separate from the plantation sys­tem (Mintz 2010:32–36). Pirates and maroon populations further illustrate the possibilities of frontiers, either within or beyond colonies. These examples suggest alternate processes or ordering regimes that were operating in tandem with, alongside, or in spite of the agro-­industrial concerns for Caribbean frontiers and that these are worth thinking about. This is further supported by recent archaeological scholarship about frontiers and related concepts of border, borderland, and periphery, which approach frontiers as permeable zones of multiethnic interaction and creativity that typically are contested spaces removed from direct po­liti­cal control (Lightfoot and Martinez 1995; Parker 2006; Ylimaunu et al. 2014). The material rec­ord of frontiers may account for this porosity and depart from overly deterministic, macroscale models that assume discrete cultural categories with certain material or spatial correlates. Rather, the fluidity of frontiers challenges and reevaluates social relations, architecture, modes of subsistence, and economics, and it is these qualities that help make these zones so interesting (Alconini 2008; Deagan 1995; Lightfoot and Martinez 1995; Naum 2010, 2012; Parker 2002, 2006; Rodseth and Parker 2005; Russell 2001:12– 13; Ylimaunu et al. 2014). Vital to understanding frontiers are the specific multiscalar qualities of each, or “boundary processes,” as the spatially and temporally specific circumstances of each set of complex interactions in frontiers (Parker 2006; also Meniketti 2009:47). Accounting for all these processes requires careful consideration of the multiple variables present in Caribbean frontiers. With a frontier era spanning more than 250 years before official colonization, Domi­nica is an appropriate locale for applying this thinking to Caribbean frontiers, using archaeological data from three sites outlined in this chapter. The island was a homeland of the indigenous Kalinago, or Carib, who first encountered Europeans on Columbus’s sec­ond voyage. Domi­nica is noted for its mountainous topography and heavily forested landscape, a setting that was not amenable to rapid agricultural development but offered

12 / Lenik

a refuge as nearby islands in the Lesser Antilles yielded to formal colonization. A series of treaties between Britain and France from as early as 1660 declared Domi­nica’s neutrality and stipulated that only Kalinago should­ inhabit it (Baker 1994:44; Boromé 1967:9; Lenik 2010:93). Nevertheless, restrictions were ignored and by the late seventeenth century French-­speaking settlers and enslaved Af­ri­can laborers had begun to build wood-­cutting, tobacco, and provisioning estates on the windward and north coasts (Hony­ church 1995:49–51). By 1730 there were 351 Europeans, 30 free Af­ri­cans, and 425 slaves in Domi­nica, and in 1753 the population had grown to 1,260 Europeans and 3,530 slaves. The number of indigenous inhabitants was not recorded (Boromé 1967:16). Among the frontier residents after 1747 were French Jesuits from Martinique who built a plantation and church at Grand Bay (Figure 1.1) (Lenik 2010, 2012b) as well as French families in places like Bois Cotlette plantation (Hauser 2015). The Ka­linago were marginalized in the interior and on the isolated windward coast near Salybia, which would become the core of the Carib Territory declared in 1903 (Honychurch 1997, 2000; Hulme 2000). This frontier status ended, at least po­liti­cally, when Britain conquered Domi­nica in 1761 during the Seven Years’ War, followed by the 1763 Treaty of Paris that ceded the Neutral Islands to Britain (Pares 1936:195–216). Even as Domi­nica became a colony, the Kalinago and many French planters and free and enslaved Af­ri­ cans who remained were joined by settlers from Britain, thousands of new enslaved laborers, and maroons, continuing the multiethnic setting albeit in a formal colony. I begin by placing the frontier within the context of questions of boundedness in the Caribbean, a region characterized by cases of admixture or creolization as opposed to neatly bounded cultural units, as anthropologists and archaeologists have noted. Viewing frontiers as porous zones of encounter conducive to creativity and hybridity within multiethnic populations resonates with this perspective, and thus a fuller account of frontiers captures the multifaceted interactions and imprecise or imposed boundaries in a way that has wider utility for the region. Yet many archaeological and his­tori­cal analyses of the frontier focus mostly on economic prospects, as new realms to be exploited for purposes of profit-­making. More can be done to capture the range of activities and ordering regimes in these places, as shown by a review of the literature for east­ern Caribbean frontiers, and discussions of frontiers, borders, and borderlands among archaeologists. Supporting this are archaeological data from Domi­nica (see Figure 1.1), which reflect the range of processes besides economic exploitation. First, a Kalinago site at Indian River illustrates the permeability of frontiers, as the site’s coastal location and artifact assemblage suggest it was situated to facilitate observation of ocean traffic and landings, as well as to permit trade and exchange,

Kalinagos and Catholics in Dominica / 13

Figure 1.1. Map of Domi­nica showing site locations mentioned in the text. (Stephan Lenik)

though acquired objects appear to have been used elsewhere as few appear archaeologically (Honychurch 1997; Lenik 2012a). A sec­ond Kalinago site at Grand Fond, located on a remote north coast hillside adjacent to Domi­nica’s first recorded petroglyphs, suggests that these markings and other rock formations influenced decisions about which specific frontier areas were chosen for concealment among many options for refuge, suggesting a “mythic geog-

14 / Lenik

raphy” for the Kalinago (Honychurch 2002; Mol 2011:70). As well, Grand Fond’s assemblage indicates the lack of congruency between cultural categories and material culture in frontier zones removed from plantation development as formal colonization began. Lastly, excavations at a Jesuit mission in Grand Bay composed of a plantation and a parish record material traces of the church and reconstruct the spatial extent of a community of free and enslaved parishioners at small plantations. Collectively, these data reveal perceptions of frontier space among the Kalinago and French Catholics that illustrate the multiple possibilities these porous zones of interaction offer.

Frontiers and Borderlands in Archaeology The idea of the Caribbean frontier resonates with anthropological concerns regarding the degree of boundedness of cultural, geographic, or po­liti­cal units of analy­sis, considering the region’s low occurrence of discrete “cultures” (Meniketti 2009; Palmié 2013; Slocum and Thomas 2003; Trouillot 1992; also Green and Perlman 1985a:3–4). Without many well-­defined homogeneous cultural units of analy­sis, scholars confronted the islands as venues of cultural admixture via models like creolization, trans­cul­tura­tion, syncretism, and the like, built on the realization that “the Caribbean is nothing but contact” (Trouillot 1992:22). Some archaeologists have struggled with boundedness too, as it applies to geographical space and material things. Meni­ ketti argues there is no bounded space in the peripheral zone of the Caribbean, and that any “geo-­social boundaries” are in fact “relics of an economic cartography of imperial ambitions” (Meniketti 2009:45–46). In relation to the Precolumbian period, Torres and Rodríguez Ramos (2008) write that even though the insular quality of the Caribbean seemingly offers discrete boundaries where land meets sea, in actuality there is inter­island visibility that greatly expands how people experience boundedness. Hence the Caribbean is a “relational space, which is continental in nature,” so that things like viewshed, as a way that humans perceive the islands, rather than physi­ cal space, best approximate indistinct boundaries (Torres and ­Rodríguez Ramos 2008:22). These works suggest a porosity or fuzziness of boundaries, rather than discrete forms, that are his­tori­cally situated. It is this insight that fits with current archaeological thinking about frontiers, in that these are not homogeneous spaces. But first it is necessary to briefly trace theorization of frontiers among archaeologists. Much of the Ameri­canist scholarship on frontiers and borderlands can be traced to the “frontier thesis” of historian Frederick J. Turner (1938 [1893]). In Turner’s view, there was “a continuously advancing frontier line” that constituted a “meeting point between savagery and civilization,” moving

Kalinagos and Catholics in Dominica / 15

west­ward as an outlet for settlers of the United States, which explains the unique character of denizens of the frontier (Turner 1938 [1893]:187). Turner’s adaptationist view directed scholars to examine in situ frontier development, which in his model was vital to the United States’ formation, as opposed to the transfer of ideals from outside sources, that is, Europe. Many further studies of the Ameri­can West were inspired by Turner’s frontier thesis, whereas others assembled a range of criticisms, too extensive to recount here, that essentially point out the one-­sided, Eurocentric, and triumphal nature of the frontier thesis (Klein 1996; Waselkov and Paul 1981:309– 311; Weber 2003). Nevertheless, the frontier, both as a geographical space and a locale of culture change, has guided studies among North Ameri­can his­tori­cal archaeologists who are concerned with intertwined pattern and process in frontiers as spaces of adaptation (Green and Perlman 1985b; Lewis 1984; South 1977:141–164; Steffen 1980; Waselkov and Paul 1981). Among the goals of these studies is to differentiate between types of frontiers, such as Steffen’s distinction (1980) of cosmopolitan frontiers versus the insular frontiers Turner was interested in, Hardesty’s industrial frontier at mining towns in the West (1985), and South Carolina’s agricultural frontier for Lewis (1984), who also defines military, trading, ranching, and other frontier types. Other archaeologists look to world systems approaches to understand relationships in frontiers as peripheries from which wealth was extracted for the benefit of a metropolitan core (Lightfoot and Martinez 1995:475–476). Some of these works also search for wider patterns in archaeological data, such as the activity group analy­sis of South’s “Frontier Pattern” (1977), where percentages of artifact groups reflect specific frontier processes, and the use of settlement patterning or ceramic data to identify standard, generalizable laws or patterns for a given frontier type (Green and Perlman 1985a; Lewis 1984). It is on the basis of these cited works that archaeologists began rethinking how the frontier should be approached in the 1990s, led by an influential paper by Lightfoot and Martinez (1995), who argue against a colonialist perspective that uses colonizer versus colonized categories in frontiers, because these tend to attribute passivity to indigenous actors. Rather, frontiers are “socially charged places where innovative cultural constructs are created and transformed,” so that these zones are open to manipulation and creativity (Lightfoot and Martinez 1995:472). Following this line of thought to pivot away from earlier typologies, a frontier can be “a shifting zone of innovation and recombination, through which cultural materials from many sources have been unpredictably channeled and transformed” (Rodseth and Parker 2005:4). For the Caribbean these transformations occurred within particular contexts that vary spatially and temporally, and even though eco-

16 / Lenik

nomic activity was the chief motivation for colonization it does not tell the complete story. Archaeology can uncover frontier dynamics by isolating specific variables in each context (Parker 2006:78), departing from frontier patterns and typologies in earlier views (Green and Perlman 1985b; Lewis 1984; Steffen 1980). Parker defines “boundary processes” as “the finite, if complex, set of dialectic interactions that take place in contact zones” (Parker 2006:78) that can be arranged on a “continuum of boundary dynamics” composed of geographic, po­liti­cal, demographic, cultural, and economic axes of variation (Parker 2002, 2006). While Caribbean island frontiers may represent “the outer reaches—the front tier—of a more powerful or central state,” these are permeable spaces of interaction in which there are encounters of indigenous and European, as well as Af­ri­can, peoples (Rice 2011:29– 30). Deploying his­tori­cal and archaeological evidence to trace interactions among heterogeneous peoples in hybrid frontier spaces reveals how variable these spaces actually were for the people involved (Naum 2010:105–106; also Russell 2001). Postcolonial approaches to frontiers and borderlands highlight the hybridity or in-­betweenness of these zones, deploying Homi Bhabha’s idea of “Third Space” (Bhabha 1994) to focus on multiple discourses intersecting in the material realm (Naum 2010, 2012; Ylimaunu et al. 2014). In frontiers this is reflected by the rejection of European sewing equipment or mimicry at New England Praying Indian Towns (Naum 2010), and the placement of churches and cemeteries as multivocal processes in Finland (Ylimaunu et al. 2014), for example. The criti­cal engagement with boundedness and interaction in Caribbean anthropology as well as current archaeological conversations regarding the permeable, hybrid nature of frontiers and borderlands illustrate the dynamic nature of these zones that is contingent on localized and multiscalar variables. Examining the material record and spatial attributes should recognize this fluidity and move beyond the focus on economic factors, as the next section shows, revealing routes toward a more inclusive portrayal of Caribbean frontiers.

Archaeology and History of Caribbean Island Frontiers Much of the historiography of the Caribbean effectively tracks the transition of frontiers into colonies, as Europeans dispossessed indigenous peoples to seize and develop lands into plantations based on enslaved Af­ri­can labor (Blackburn 1997; Moya Pons 2007; Watts 1987). Methods ranged from military conquest such as Britain’s seizure of Jamaica and treaties in which islands like St. Croix or St. Lucia changed hands multiple times and included periods of shared occupation like in St. Kitts and Martinique. In this way,

Kalinagos and Catholics in Dominica / 17

new territories were approached as agricultural frontiers to be targeted for plantations that would generate profits (Boucher 2002, 2008:138–143; Candlin 2012; Pérotin-­Dumon 1999:152–156; Roberts 2016). His­tori­cal archaeology contributed at this juncture by combining written records with survey and excavation field methods to identify and record material evidence of daily life at frontier sites, which tend to be poorly recorded, if accounts were kept at all. Archaeologists use the frontier concept in the Caribbean differently, depending on time period. In the Precolumbian period, frontiers formed as migration into the island chain occurred, at first identified by lithic and marine shell artifacts deposited by preceramic peoples, followed by successive waves from South America traced by ceramic types as cultural frontiers or boundaries advance over space and time (Rouse 1992). According to Rouse’s model (1992), frontiers were temporal-­spatial boundaries between cultures that could be traced in material culture, especially ceramics. Newly arrived migrants might halt their advance when encountering existing inhabitants, leaving a porous frontier through which interaction could occur (Wilson 1997:173). Also, frontiers need not move linearly, since islands in the chain could be skipped (Haviser 1997). Among his­tori­cal archaeologists and historians, the idea of frontiers or borders in the east­ern Caribbean shifts focus to unexploited areas for expanding production of sugar, tobacco, coffee, and other crops via plantation agro-­industry (Baker 1994:47–56; Barbier 2011:314; Boucher 2002, 2008; Deagan 1995; Garraway 2005:39–70; Hauser and Kelly 2011; Hicks 2007; Sheller 2003). Some of these studies (Candlin 2012; Meniketti 2015:14; Roberts 2016) resemble the agricultural frontiers of Lewis (1984). For Hicks (2007), the frontier is the period of initial development before official colonization, with a zone of economic interaction characterized by the timber trade and “nucleated plantations” in “frontier societies” (Hicks 2007:26– 33), so that frontiers are viable areas for economic exploitation as reflected in their spatial organization. In cases when colonies and frontiers are connected by ocean passages, Hauser and Kelly (2011) explore interisland maritime trade that violated trade restrictions such as the exclusif and Navigation Acts imposed by European nations in the eighteenth century, with the goal of exclusive colony-­metropole exchange. Archaeological evidence in the form of ceramic sherds traces how restrictions were ignored, so the material remains track this process as frontier activities were motivated by economic opportunities (Hauser and Kelly 2011). Other his­tori­cal archaeological studies go further by thinking about characteristics of the societies linked to the plantation sys­tem developing in east­ ern Caribbean frontiers. For Domi­nica, Baker (1994:47–56) describes a “fron­ tier society” in socioeconomic terms that transitioned to a more “ordered

18 / Lenik

society” from around 1700 to 1763 (Baker 1994:50–51). Based on his study of seventeenth-­century French Martinique and Guadeloupe, Boucher (2002, 2008) proposes a “frontier era,” which precedes the “mature” period when sugar plantations came to dominate the colonial landscape, when French settlers held middle-­level skill positions, relying on small-­scale cash crops and resisting impositions of power from the monarchy and church, which sent priests and missionaries, established courts, and divided islands into parishes or administrative units (Boucher 2008:138–143). Frontier-­era Marti­ nique and Guadeloupe is divided into two phases. First, the initial “pioneer­ ing” phase covers the 1620s to the 1660s (Boucher 2002:211) when settlers cleared forests for agricultural development and set up habitations (Boucher 2008:122–125), followed by the “pre-­plantation era” from the 1660s to circa 1700 (Boucher 2008:115–116). Early eighteenth-­century Domi­nica resembles Boucher’s frontier era as settlers in coastal areas relied on cash crops and began establishing administrative and social order, while the interior and windward areas were sparsely settled and uncontrolled by Europeans. By 1745, for areas in Domi­nica they controlled, the French collected export duties, settlers paid taxes on their enslaved laborers and were compelled to provide enslaved Af­ri­cans for pub­lic works projects (Boromé 1967:19). An added factor for Domi­nica is the opportunity for refuge the rugged terrain offered. Based on ethnographic fieldwork in the 1980s, Trouillot’s study (1988:27– 28) shows how Domi­nica’s thick vegetation and topography restricts settlement to a “patchwork of enclaves,” so the primary settlement zone arose along the coasts as people “had to accommodate themselves within the marginal areas left between or around the steeps” (Trouillot 1988:28). Social relations shaped by this topography that Trouillot traces would likewise ­apply to the frontier. Besides the extraction of resources, some historians and archaeolo­gists identify other processes characteristic of frontier spaces. For example, Garraway applies the term “border” in studying French-­Kalinago relations in seventeenth-­century Mar­ti­nique (Garraway 2005:39–92). Since French colo­ nists and the Kalinago lived in discrete communities that did not in­ter­mingle, Garraway defines a “moving border that shifted across time and space,” as depicted on Du Tertre’s 1667 island map (Garraway 2005:42–43). This was a permeable “interstitial zone” of contact among Caribs, Europeans, and enslaved Af­ri­cans (Garraway 2005:43–44). Peabody adopts a similar approach, with a “cultural frontier” between missionaries and enslaved Af­ri­ cans contributing to a “shift in racial discourse” that sof­tened perceptions of enslaved Af­ri­cans (Peabody 2004:114). Archaeological investigations at Puerto Real in Hispaniola reveal characteristics of diet and spatial organization that illustrate the interaction of imported Iberian ideals and life­ways

Kalinagos and Catholics in Dominica / 19

that are adapted to Spanish-­Caribbean frontier life (Deagan 1995:447–455). Drawing from Hardesty (1985), Meniketti’s approach to agro-­industrial frontiers in Nevis looks at the role of material goods in defining social behaviors (2015:50–52).

Archaeological Evidence of the Frontier in Domi­nica Following the arguments outlined, and building from accounts of Domi­nica’s heterogeneous frontier setting, the archaeological data reveal ­aspects of the geographic, cultural, and economic diversity among the island’s indigenous, Af­ri­can, and Euro-­Ameri­can inhabitants. Domi­nica has been among the least archaeologically explored Caribbean islands. Systematic research since the early 2000s has greatly improved understandings of settlement patterning, material culture, and intra-­and interisland interactions, with studies of the Precolumbian period (Shearn 2014; Bérard 2006), French and British plantations, French Jesuit missions, and British fortifications (Beier 2014; Hauser 2015; Lenik 2010, 2012a, 2012b). This ongoing work has uncovered intact subsurface resources despite the precipitous terrain, heavy rainfall, agro-­industrial development in the colonial period, and scarcity of level topography. With a lengthy frontier period Domi­nica is a logical place to explore frontier dynamics, as the island harbored a Kalinago population through the historic period to the present, and by the late seventeenth century attracted missionaries and French Catholic settlers who brought in enslaved Af­ri­can laborers. The pre-­1763 period presents challenges: written accounts and maps offer some direction, yet documentary references to certain bays or coastal settlements (e.g., Boromé 1966, 1967; Honychurch 1997) of­ten are imprecise and difficult to isolate in the field. Nevertheless, careful reading of the sources combined with pedestrian survey and systematic shovel test pit surveys have identified and sampled frontier sites at Indian River, Grand Fond, and Grand Bay (Lenik 2010, 2011, 2012a, 2012b, 2012c). These data portray the range of ways of engaging with frontier space and material objects, other than for agro-­industrial purposes, among Domi­ nica’s frontier enclaves. As shown here, archaeology identifies arrangements of domestic and religious occupations and sites of trade and exchange that reveal strategies guiding placement of frontier sites. Additionally, the artifact assemblages illustrate porous or fluid boundaries among handmade and imported objects that are subject to local variables at each site. Indian River Prince Rupert Bay (see Figure 1.1) has many hallmarks of frontiers in the Americas as a zone of indigenous-­European interaction in a coastal envi-

20 / Lenik

ronment where exchange of material goods and foodstuffs occurred. After 1635, Domi­nica was a po­liti­cally neutral island frontier surrounded by the French colonies of Guadeloupe, Marie Galante, and Mar­ti­nique that were visible to the north and to the south, and connected by navigable ocean passages. Emptying into Prince Rupert Bay, Indian River was a locale of intermittent Kalinago-­European frontier encounter, as it was a stopping point for European vessels for trade, supply, and rest after transatlantic voyages (Hony­ church 1997, 2000:44–47; Lenik 2012a). Honychurch identifies accounts of Kalinago in this bay and along the Indian River from the sixteenth century at least until the 1640s, and possibly as late as 1700, when French squatters began settling the west coast (Honychurch 1997:297; 2000:44–47). Whereas Europeans had stereotyped the Kalinago as warlike cannibals as justification for enslaving or eradicating them, Honychurch finds that the Kalinago traded fruits and vegetables with ships in the bay, and he interprets the accounts to suggest that they grew tobacco for trade with Europeans in order to acquire “iron cutting tools” (Honychurch 1997:297, 2000:127) as well as “glass beads, knives, hatchets, saws, copper jewels, and brass pendants” (Honychurch 1997:297). A sixteenth-­century account describes an inland settlement on the Indian River as “a poore Towne . . . of some twenty cottages rather than Houses, and yet there was a King, whom they found in a wide hanging garment of rich crimson Taffetie, a Spanish Rapier in his hand, and the modell of a Lyon in shining Brasse, hanging upon his breast” (Hulme and Whitehead 1992:60). Clearly, there were economic boundary processes at work in the vicinity of Indian River. Yet earlier discussions of the frontier would urge attention to geographic, po­liti­cal, demographic, and cultural processes (Parker 2006), such as the meanings of these items in possession of the “King” and the uses of iron commodities among the Ka­linago, not only at Prince Rupert Bay but also through­out the island. Following Honychurch’s identification (1997) of a post-­1492 Kalinago site on the Indian River’s north bank, I directed archaeological excavations in a 6,500 m² area about 1–2 m (3–6 ft.) above sea level, about 50 m west of the bay and bordering the river (Lenik 2009a, 2012a). A systematic survey of 44 shovel test pits (STP) plotted artifact distributions and three 1-­m-­square test units searched for remains of structures, midden deposits, or layers reflecting Kalinago occupation. Some crab holes and tree growth caused localized disturbances to subsurface contexts, and British colonial period artifacts were recovered because this area was used for different activities in­clud­ing repair of sails. Even so the findings support a Kalinago presence as shown by indigenous handmade earthenware sherds (n = 422; 4.94 kg) from through­ out the site (Lenik 2012a:90–91). Fragments of six vessels shown in Figure 1.2 possess distinctive decorations and forms consistent with Cayo ceram-

Figure 1.2. Cayo pottery fragments recovered from Indian River, Domi­nica. IR = Indian River artifact number. (Stephan Lenik)

22 / Lenik

ics (Lenik 2012a:86–87), which Boomert (1986, 1995, 2004) has shown to match written accounts of Kalinago pottery and archaeological collections in the Windward Islands and Guianas. The Indian River assemblage yields little evidence of the European commodities recorded to have been exchanged here prior to British colonization. Some iron objects, in­clud­ing 65 nail fragments and undiagnostic fragments (n = 83; 0.75 kg), were recovered but none appear to have been cutting tools. One blue glass bead was found on the surface, a compound cylindrical bead with eight vertical white lines. The European ceramic collection (n = 126) yields only a few fragments of tin-­glazed wares (n = 3), Staffordshire slipware (n = 3), and agateware (n = 1) with production dates extending before 1763. A ware produced in France known as Vallauris (n = 11) has been found in early eighteenth-­century contexts, in­clud­ing the Jesuit mission at Grand Bay. Otherwise no recovered European artifacts can be confirmed to overlap with dates of Kalinago occupation. Written sources that indicate Indian River as a locus of frontier exchange would seem to prioritize economic boundary processes in characterizing this frontier setting, and from a Eurocentric perspective Prince Rupert Bay was a place where commodities such as tobacco and food and water could be acquired. However, the archaeological findings, in terms of material culture dating before formal colonization began in 1763, reveal virtually no evidence of such items. Elsewhere I have argued that the Indian River was not a full-­time settlement and that trade goods were taken to other contexts removed from the bayfront. This was a coastal area prone to flooding that can be infested with mosquitoes, and from a Kalinago perspective it offered versatility to observe traffic at sea, to choose to engage and possibly trade with visitors, but also to remain concealed and flee if desired (Lenik 2012a). Thus Indian River permitted a variety of economic and social choices and appears to have been a transshipment point or observation post rather than a permanent settlement. Newly arrived goods were selected and adapted to Kalinago material life, such as the garment, personal adornment, and sword in possession of the “King” described earlier. Thinking about Indian River from the perspective of the Caribbean as “relational space” (Torres and ­Rodríguez Ramos 2008:22), the archaeological deposits capture only one segment of the acquisition, use, and deposition of trade goods. The frontier in this part of Domi­nica had blurry boundaries, formed from the collectivity of the land areas, the river, bay, and open ocean. Grand Fond A sec­ond Kalinago site more removed from the Caribbean frontier’s economic processes is a remote refuge on the north coast at Grand Fond (Lenik

Kalinagos and Catholics in Dominica / 23

2012a:91–94). Looking to discussions of boundedness cited earlier, incongruencies and blurred boundaries would be expected in east­ern Caribbean frontiers, and Grand Fond illustrates two aspects of this. First, petroglyphs and rock formations as features of a mythic landscape shaped Kalinago perceptions of space as they sought refuge while Europeans initiated a plantation regime based on enslaved Af­ri­can labor. Second, the artifact assemblage does not fully match what would be expected at an indigenous site, illustrating the tenuous relationship between material culture and identities in the frontier. A Kalinago settlement at Grand Fond on the north coast appears on two maps dating to the 1760s. On “A Plan of the Island of Domi­nica” (The National Archive [TNA], Lon­don, England; MFQ 1 1173), “Caraibes” is written between the Tafia River and Reposoir, a spit of rock jutting into the sea. Between these same landmarks on the Thomas Jefferys’s West India Atlas map of Domi­nica (TNA, CO 700 / DOMINICA5), “Carabbe Indians” appears alongside several shapes indicating a habitation. Kalinago may have remained at Grand Fond into the colonial period, since the 1776 Byres map (TNA, CO 700 / DOMINICA6) shows neither settlements nor surveyed land plots here. This isolated coastal region offers few locations for human settlement since it consists of a series of steep ridges and ravines running down to the sea, terminating in cliffs above the Guadeloupe Passage. The most likely area for settlement is the former village of Grand Fond, abandoned after Hurricane David in 1979 (today a village on the east coast shares the same name). Reconnaissance below the former village site in the Sibouli Ravine found a level area atop a small hill near a spring. The site, located about 350 m from the coast at over 76 m (250 ft.) above sea level, was tested with 22 STPs at a 10 m interval and three 1-­m-­square excavation units. Surface scatters of Amerindian pottery as well as British creamware and pearlware sherds indicate a mid-­to late eighteenth-­century date (Lenik 2009a, 2009b, 2012a:91–94). In the assemblage, European tableware ceramics consist only of creamwares (n = 81), pearlwares (n = 108), and tin-­glazed (n = 2). Evidence of structures is shown by 256 nail fragments, with 246 found in the three excavation units. No brick or tile was recovered. A nonferrous metal fragment pierced with evenly spaced holes to create a rough surface for grinding cassava found on the surface cannot be assigned a date. During fieldwork at Grand Fond, Domi­nica’s first petroglyphs were discovered, consisting of six simple faces and one partial face carved into a volcanic stone that are consistent with other indigenous Caribbean rock art sites (Lenik 2009a, 2009b, 2012a:92–93, 2012c). The Grand Fond site reveals two features of the Kalinago experience in east­ern Caribbean frontiers. As was the case with Indian River, the pres-

24 / Lenik

ence of European manufactured goods indicates some forms of exchange were taking place, but this economic process distracts from other ordering regimes alongside or in spite of the agro-­industrial mode of expansion into frontiers. First, as noted elsewhere (Lenik 2012b), the artifact assemblage leaves a gap in the period from circa 1000 to 1750, with late Saladoid phase ceramics (AD 300/400 to circa 600/800; Petersen et al. 2004:25–26) too early to be connected to Kalinago, and the creamwares, pearlwares, and glass bottle fragments that date to the mid-­eighteenth century at the earliest are too “late.” Similar occurrences have been noted at North Ameri­can indigenous sites, where European-­made goods may have been deposited by any number of people (e.g., Silliman 2009:225; 2010). The takeaway here is that eighteenth-­century Kalinago sites may not consist of the ceramic, lithic, and shell artifacts typically associated with Precolumbian indigenous occupation ­(Lenik 2012b:93–94). Second, frontier space for the Kalinago may have been shaped by a “mythic geography” (Honychurch 2002; Mol 2011:70), so that the petroglyphs located in the middle of the archaeological site had a deeper set of meanings that has been attached to that spot for centuries as indicated by the Saladoid ceramics found there. Similarly, a volcanic rock formation known today as l’Escalier Tête-­Chien is found near Salybia. Thus, indigenous people interacted with this Caribbean frontier at a time when they were forced to conceal themselves in the mid-­to late eighteenth century, according to an entirely distinct topography that had virtually nothing to do with this frontier’s economic potential.

Jesuits and Frontier Parishes A third site illustrating the variety of interactions in Caribbean frontiers is a Catholic parish based at Grand Bay, part of a French Jesuit mission lasting from 1747 to 1763 (see Figure 1.1). As a frontier Domi­nica attracted missionaries who unsuccessfully proselytized among the Kalinago before re­ directing efforts toward Francophone settlers who built plantations, since there were few prospects for acquiring land for agricultural exploitation in nearby French colonies that had reached a mature phase of plantation development (Boucher 2008:138–143). But in addition to financial concerns, many of these frontier inhabitants continued practicing Catholicism outside the direct influence of church administration until it became institutionalized by a parish the Jesuits established. In this case archaeology offers two perspectives toward understanding how parishes as administrative and ecclesiastical units shaped frontier social relations. First, excavations identified remains of the church, in­clud­ing building materials and the layout of the mission’s core area, which reveal the materiality of a frontier parish. Second,

Kalinagos and Catholics in Dominica / 25

the spatial extent of a frontier Catholic community composed of free and enslaved inhabitants linked by the parish can be partly reconstructed based on the parish register, maps, and land records. The earliest recorded visit by a Jesuit to Grand Bay occurred in 1747 at the invitation of a free person of color named Jeannot Rolle, who in 1691 had come with his family from Mar­ti­nique to establish a plantation. After conflicts with the Kalinago over the display of crosses made from wood, the Catholic Rolle had his enslaved laborers erect a stone cross on the bayfront in 1692 that led to the expulsion of the indigenous inhabitants. More French Catholic settlers joined Rolle in Grand Bay Quarter (Boromé 1967:24–25; Honychurch 1995:56–58; Moris 1926a–1926e). The Jesuit who eventually came to Grand Bay was Antoine de La Valette, the Mar­ti­nique mission’s financial manager, and despite rules forbidding religious orders from engaging in commerce, he had been borrowing large sums, among other commercial schemes. With the proceeds he restored the Jesuit plantation in Mar­ti­nique and went to neutral Domi­nica to acquire land for a new plantation, where he constructed a large factory building and manned it with enslaved Af­ri­can laborers purchased in Barbados. Income from a variety of cash crops produced in Domi­nica helped repay the mission’s debts. To justify the Jesuits’ incursion a parish church was placed next to the factory (Lenik 2010:114–163, 2012b; de Rochemonteix 1907; Thompson 1976, 1996). Parochial duties were assigned to Cordeliers, or Franciscan priests from Mar­ti­nique ­(Moris 1926b:219–220), while Jesuits managed the plantation. A shipment of La Valette’s products captured in 1755 bankrupted a major lender, causing a series of lawsuits eventually reaching the Paris parlement. Its decision resulted in the entire Society of Jesus being held responsible for La Valette’s sizeable debt, and as he was unable to repay this sum the Society was dissolved and expelled from France and its colonies (Lenik 2010:114–163, 2012b; de Roche­monteix 1907; Thompson 1976, 1996). After the Jesuits’ departure, the site was abandoned and British colonial period plantations were built farther inland, at Berricoa and Geneva (see Figure 1.1). Excavations recorded the remains of four buildings that reconstruct the ­ arish mission’s core compound on the shoreline, in­clud­ing the factory, p church, a manager’s residence, and a fourth structure likely related to the plantation. The site covers 0.3 ha, with 37 STPs and a total of 131 1-­m-­square units excavated. These were near Rolle’s stone cross on the bay­front near the mission, although the precise location of the cross is unknown because it was moved to a nearby Catholic cemetery in 1892 (Moris 1926a:187). Overseen by a Jesuit, Père Lancastre, and a settler, Etienne Picot, the church was “a small, wooden building on a stone foundation” that was consecrated on Christmas Eve, 1749 (Moris 1926a:189, 1926b:219–221; Lenik 2010:126–

26 / Lenik

127). The church was damaged by a hurricane in 1765 and replaced by a new one on a nearby hillside around 1769, until by 1790 only its ruins were visible (Moris 1926b:221). Excavation of 1-­m-­square units traced the church’s buried remains, uncovering a rectangular structure parallel to the bayfront. It is 14.5 m long, and a slight bend in the east wall suggests a transept, so the width is 7.0 m in the southwest and 8.0 m wide in the northeast. Foundations consist of locally available stones gathered from the beach that are bonded with sand mortar. Evidence of stone flooring in the north­east suggests a paved sanctuary. Large quantities of ceramic tile indicate a tiled roof, and pieces of thick flooring tile, in­clud­ing glazed varieties in yellow, green, and brown, that are 25.9% of the floor tile assemblage, would have been featured in the walls or floor. The uniformity of these imported tiles marked the church as sacred space and prominently displayed the role of the Catholic Church, next to Rolle’s stone cross, in a conspicuous manner to project the prestige of the Jesuit order and to justify the incursion into a Neutral Island (Lenik 2012b). This parish also had wider ramifications in the frontier beyond its core at Grand Bay that show the role of Catholicism among free people and enslaved Af­ri­cans who had sacraments performed in this parish. Evidence of the spatial extent of this frontier community can be reconstructed using maps of the island and two preserved portions (1748–1755 and 1780–1782) of the parish register compiled by Father Raymond Proesmans (1943a– 1943f ). This register lists baptisms, marriages, and burials of free and enslaved people, recording 291 individuals, 250 of whom are enslaved, along with 32 free people and nine of unknown status. Aside from the Jesuits, there are 22 other owners (Lenik 2010:129–133). Most notable about this population is the large number of enslaved people of Af­ri­can descent, in­ clud­ing Af­ri­can-­and Creole-­born, who appear in the record as being associated with these sacraments, indicating participation in this community. The durability of these links among the enslaved laborers is shown by the fact that the Catholic population begun in the frontier era survived through the entire period of Domi­nica’s status as a British colony under the Anglican Church and postindependence also. Several French surnames listed in the 1748–1755 register appear in British land records for St. Patrick’s Parish and can be spatially referenced to the 1776 Byres map (Figure 1.3). Surnames like Antheaume, Dubucq, Hirriart, and Pacquet from the 1748–1755 register appear in the list of properties accompanying Byres’s map (Lenik 2009a, 2009b, 2010:129–140). This assumes French settlers who took a loyalty oath to stay in Domi­nica leased the same properties they had occupied before 1763 (Atwood 1791:3–5; Honychurch 1995:73–75; Murdoch 1984:556). The geographical distribution of people who had sacraments

Kalinagos and Catholics in Dominica / 27

Figure 1.3. Map of Grand Bay Quarter, Domi­nica. Note the origi­nal French spelling of Grand Bay (Grande Bay) on this map. (Adapted from the Byres map, 1776. Stephan ­Lenik.)

performed in this parish shown in Figure 1.3 illustrates the network of enslaved and free people in the frontier and approximates the small plantations where they lived. It is unrecorded whether people traveled to visit the priest at the church in Grand Bay, or if the priest personally visited plantations. Nevertheless, this evidence reflects a form of social relationships in the frontier, such as the role of godparents and conversion to Catholicism. Weekly Mass and other observances at the church provided a schedule that would have ordered daily life in Grand Bay and more broadly in south­east­ ern Domi­nica.

Conclusions By viewing frontier Domi­nica as a porous zone of interaction that offered multiple possibilities alongside an economic mode of ordering frontier space, these three sites illustrate in a more comprehensive manner the potential for Caribbean frontiers and the multiple boundary processes that charac-

28 / Lenik

terize these spaces. First, these three sites cannot be completely removed from economic factors. Indian River was a site of trade, and the Kalinago possessed European goods at Grand Fond. The Jesuits built a plantation at Grand Bay alongside the parish. Imported commodities appear archaeologically in all three instances, which is not unexpected, as this also occurs in assemblages at frontier towns in Hispaniola (Deagan 1995) and early colonial forts and plantations (Hauser and Armstrong 2012; Hauser and Kelly 2011; Hicks 2007; Klingelhofer 2008; Lenik 2010, 2012a, 2012b; Ryzewski and Cherry 2015:363–367). But the economic activities suggested by these objects must not distract from other frontier processes, and the three sites make this point through the modes of sociospatial ordering and material culture. For the Kalinago, the Indian River was a flexible locale in a permeable border zone, allowing trade and exchange, but also concealment. Cayo pottery verifies a Kalinago presence, yet trade goods were not deposited here. This indicates the complexity of indigenous use of space (for example, a flexible locale like Indian River along the permeable and unbounded zone of interaction on land and sea surrounding Prince Rupert Bay). At Grand Fond, a Carib village that colo­nists documented in maps from the 1760s, the Kalinago found a remote location allowing concealment, but they also were influenced by this petroglyph site as part of a “mythic geography” in the frontier that influenced the choice of this specific area, as does the l’Escalier Tête Chien rock formation near Salybia. Grand Fond also reflects the possible use of European commodities, suggesting that in this part of the frontier there were flexible linkages of material things and identities that may in fact have been complex and multifaceted, compared to those who frequented Indian River. Moreover, there could be social interaction among groups of Kalinago that do not show up archaeologically. At Grand Bay, Francophone settlers who exploited neutral Domi­nica’s available lands and fertile soil were joined by ­Jesuits from Mar­ti­nique. The community formalized by the Catholic parish diffused through the frontier in a way that was parallel to the network of small estates, while the church and stone cross at Grand Bay introduced a material manifestation of the Catholic Church. While we cannot estimate the specific meanings Christianity entailed, particularly for enslaved Af­ri­cans, long-­lasting shared patterns must have emerged from these frontier era relationships because today there remains a strong Catholic community in Grand Bay. Together, these archaeological findings illustrate the heterogeneity of Caribbean frontiers and suggest a suite of strategies by which frontier inhabitants organized their daily lives and their engagement with frontier

Kalinagos and Catholics in Dominica / 29

space, besides its economic aspects. Thinking about Caribbean frontiers as unbounded permeable space, attuned to the geographic, po­liti­cal, demographic, cultural, and economic axes of variation (Parker 2002, 2006), shows some of the richness of Caribbean frontier zones, so that a more complete picture of the islands emerges alongside of or despite the plantation societies that came to dominate the colonial period.

2 The Congo Free Black Village on St. Eustatius, Netherlands Caribbean R. Grant Gilmore III

Archaeologists working in St. Eustatius are addressing important questions that focus on West Af­ri­can–derived architecture, social, and religious spaces. The objective is to describe and interpret how these cultural elements relate to social, economic, and power structures found in the colonial and postcolonial lives of enslaved Af­ri­cans and free blacks in St. Eustatius (residents refer to themselves as Statian and as being from Statia). This chapter combines archaeological, architectural, his­tori­cal, and photographic evidence to inform these issues. The goal is to address trade relationships and legal regulations affecting enslaved Af­ri­cans, free blacks, and colonial whites. This also considers patterns of provisioning in Statia’s Afro-­Caribbean community. The final section summarizes the results of excavations at the Congo Free Black Village that allows for the careful examination of variability in free black domestic architecture on St. Eustatius and for broader comparisons elsewhere in the West Indies. This chapter demonstrates that free blacks on St. Eustatius enjoyed unfettered access to an economic outlet for their own wealth production activities. The research described here represents one of the few intensive archaeological investigations regarding a his­tori­cal Caribbean free black community. The results illustrate Creole Statian society’s unique characteristics that are manifested on many sites across the island (Gilmore 2004, 2006a, 2013, in press). The excavation also provides interesting clues regarding the resilience of imported cultural traditions in a society that was at once outwardly very cosmopolitan and also more “free” and thus less likely to prohibit or inhibit anomalous social norms.

His­tori­cal and Geographical Setting St. Eustatius is located in the northeast­ern arc of the Lesser Antilles— an ideal position for establishing a trading depot like none seen before or very likely since (Figure 2.1). Both prehistoric and colonial peoples used the proximity of islands in the Caribbean archipelago as stepping-­stones for colo-

The Congo Free Black Village / 31

Figure 2.1. Location of St. Eustatius in the northeast­ern Caribbean and Site SE 132, the Free Black village. (R. Grant Gilmore III)

nization and trade (Armstrong 2006; Armstrong et al. 2009; Hofman et al. 2007, 2008). People, ideas, and products were easily transferred over great distances with relative ease (Schaw et al. 1923). Powered by favorable trade winds, ships carried these items from and to the rest of the Atlantic World— especially during the colonial period (Great Britain Board of Trade 1789; Morgan 1993; Postma and Enthoven 2003). In addition to its ideal geographical location, St. Eustatius is a comparatively low island and as a result receives very little annual rainfall (Westermann and Kiel 1961). For these reasons and the island’s small size it was largely ignored by France and England as a place to establish plantations with extensive agricultural production (Goslinga 1985; Knappert 1932). The evolution of many colonial Caribbean physical and fiscal landscapes was inexorably tied to the economic cycles associated with agricultural production. Plantation sugar monoculture provided the wealth, both urban and rural, through which all levels of society were able to finance landscape

32 / Gilmore

modi­fi­ca­tions. On St. Eustatius, although there were many plantations, they contributed little to the economic position of the island’s inhabitants. One of the largest trading networks in the colonial world was centered on St. Eustatius in the latter quarter of the eighteenth century (Emmer 1998; Gilmore 2013; Goslinga 1990; Klooster 1998; Oostindie and Roitman 2014; Postma and Enthoven 2003; Tuchman 1988). This affected each level of society from the ultrawealthy merchant/planters to the enslaved Af­ri­cans. The result was a society set apart from all others in the colonial Caribbean. Merchant/planters built a society designed to maximize profits through personal contacts reinforced by social structures centered on entertainment and ostentatious displays of wealth. At the other end of the spectrum, although the enslaved and free blacks were kept at the physical periphery of Oranjestad, they were intimately involved in keeping the city’s trading activities running smoothly for slaveholders while at the same time improving their own physical and economic conditions (Gilmore 2009). By the 1680s, with the simultaneous collapse of the North Ameri­can tobacco market and the transfer of Dutch Manhattan to the English, St. Eustatius’s economy focused on trade and transshipment of goods (Goslinga 1971; Klooster 1998; Postma and Enthoven 2003). Slave trading was well established during the 1720s but began a steady decline by the 1740s (Enthoven 2012, Goslinga 1985; Knappert 1930). The Dutch West India Company’s directors, the Heeren XIX, gave permission for St. Eustatius to be the first free trade port in modern times in 1754 (Emmer 1998; Enthoven 2012; Gilmore 2009; Oostindie and Roitman 2014). Thus began the rapid ascent of St. Eustatius to becoming one of the busiest ports in the world by the 1760s with a colonial material cultural heritage to match (Gilmore 2006b, 2013; Smith 1776). However, by the early nineteenth century, the growing economy of the United States of America spelled an end to the prodigious wealth once generated on the island (Gilmore 2013; Jameson 1903). When his­tori­cal archaeologists first came to St. Eustatius in the 1960s, they found a veritable cornucopia of very well preserved sites. Ivor Noël Hume was the first his­tori­cal archaeologist to set foot on the island in 1966, and he later related the uniqueness of St. Eustatius’s archaeology within the Atlantic World (Noël Hume 1991, 2001). In 1982 Edwin Dethlefsen and Norman Barka dubbed the island “The Pompeii of the New World” because of its outstanding preservation of buildings and its terrestrial and submerged archaeological sites (Dethlefsen and Barka 1982). Thus, from the 1970s onward, his­tori­cal archaeologists have continuously worked on the island, with a permanent presence beginning in 2004 with the establishment of the St. Eustatius Center for Archaeological Research (SECAR). Primary research foci for SECAR have been to quantify and understand

The Congo Free Black Village / 33

the character of St. Eustatius’s contributions to capitalism as a global economic force as well as characterizing the lives of St. Eustatius’s inhabitants both enslaved and free in the context of this unique economic environment (Gilmore 2004, 2006b, 2013; Gilmore and Miller 2011; Miller 2008; ­Gilmore and Nelson 2015).

Free Blacks and Trade As a tax-­free port, St. Eustatius became the center for international Atlantic trade in the last quarter of the eighteenth century and represented the apogee of efficient commerce during this period. Island residents, both permanent and transient, gained unimaginable wealth within this perfect economic storm. At the core of this trading activity was a slave and free black community unlike any found in the New World. European settlers and merchants, enslaved Af­ri­cans and free blacks all participated in the brisk trade at Statia’s Lower Town warehouses. As occurred elsewhere, any surpluses realized by the enslaved (either by their own labor or through stealing) were of­ten sold to fellow enslaved people, freedmen, or in markets if permission could be obtained from slaveholders. The profits from these transactions were spent on any number of items that the enslaved required for day-­to-­ day living. In some areas, substantial profits were made through the industriousness of the enslaved (Berlin and Morgan 1991; Beckles 1991; Craton 1982). Enslaved people were so successful in these ventures that other market sellers (freedmen) protested vehemently against their selling goods in markets in many colonies. In fact, the free people were able to get legislation passed that prohibited or at least severely restricted the enslaved ­people’s ability to sell produce in direct competition with their wares. Some authors even believe that the enslaved came to monopolize local trade on some islands (Marshall 1991:56). Some free blacks prospered under Dutch conditions of free trade as well as during more restrictive but brief periods of British and French rule. For example, the free Negro Cloé, formerly the property of a “Mr. Rieboo,” owned a slave named Marian (Nationaal Archief [Dutch National Archives, DNA] OASE 82 Peter Ouckama, Last Will and Testament of Cloe, Free Negro Wench, 1782). Her economic prowess is illustrated by the fact that she purchased her own free­dom and she eventually acquired a number of houses that she rented to other free blacks on the island. Joseph Howe, another free black, owned four slaves (DNA OASE 84 Peter Ouckama, Last Will and Testament of Free Black Joseph Howe, 1784), and the free Negro May Harvis, owned three slaves and her own house in the New Town (DNA OASE 84 Abraham Du Veer Deed [for Glassbottle Fort plantation]. April

34 / Gilmore

17, 1818). In 1818, the free black woman Frances Cuffey accumulated sufficient wealth that she was able to purchase Glassbottle Fort plantation (in­ clud­ing 14 slaves) for $3,000. Olaudah Equiano (I use his Af­ri­can name instead of his legal name, Gustavus Vassa [Equiano et al. 2013]), the eighteenth-­century autobiographer, is perhaps the most extraordinary case of an enslaved Af­ri­can achieving economic success in Statia. Although his origins, education, and personal motivations are debated, Equiano and perhaps many other enslaved and free blacks, consistent with social and economic circumstances on Statia, were literate (Acholonu 1989; Byrd 2006; Carretta 2007; Carretta and Gould 2001; Edwards 1969; Equiano et al. 2013; Gilmore 2009). This gave Equiano and others a distinctive advantage over those who could not read or write (Gilmore 2006a). Equiano’s social, po­liti­cal, and most of all economic self-­determination are evident in the routine and frequent trips he made between Statia and Montserrat, where his trading activities gained profit for himself and his master (Equiano 1999). In his autobiography he carefully describes his process for self-­manumission. He sold excess provisions from his plot to purchase a glass tumbler. He doubled his money on its sale on nearby Montserrat. This process continued over many months until he had saved enough to purchase his own free­dom. Equiano was so successful that he eventually moved to Lon­don, settling in the Marylebone neighborhood where he referred to himself as a gentleman in eighteenth-­century English fashion. There he married Susannah Cullen, and when he died in 1797 his estate was valued near two thousand pounds. His experience is perhaps unusual, but it does reflect a socioeconomic state for the unfree that was permissive and pervasive in Statian society at the time, one that provided ample opportunity for those enslaved who could take advantage of it. With no taxation for imports or exports and little additional governmental regulation, Statia was a great place to conduct trade (Enthoven 2012; Postma and Enthoven 2003). Government workers and business entrepreneurs on Statia were of­ten one and the same with mutually beneficial results for all. The enslaved, free, and Europeans from all nations traded peacefully and freely for decades. Goods from the world over transshipped through the ware­houses on Oranje Bay. It could even be said that every colonial household in the Americas had at least one product that was traded via the Golden Rock. Under such conditions many enslaved were able to earn money to purchase their free­dom in less time than it took in other colonies. It is also evident that not all freed slaves left the island to seek their fortune elsewhere during Statia’s economic boom. The social and economic advantages to continue living on Statia were clearly evident, and thus on manumission many free blacks remained embedded in the Statia economy. Enslaved Af­ri­cans

The Congo Free Black Village / 35

on St. Eustatius were at the core of economic activity—and thus were at the core of the development of capitalism as we know it. Equiano and many other free blacks helped create and contributed to an economic boom that lasted until 1795.

Slave Regulations Despite the economic free­dom that many slaves and free blacks on Statia enjoyed, the colonial government attempted to manage and exercise social control over their lives. For example, in some areas in the Americas such as Charleston, South Carolina, it was customary to have the enslaved wear some sort of identification, such as “slave tags,” and for the emancipated to carry their manumission papers with them at all times. On St. Eustatius, in 1785, legislation mandated that free blacks had to wear an outward sign of their free status in the form of a red ribbon on their breast (DNA OASE 24, No. 10 Herman Brower Free Negroes and coloureds obliged to bear a red ribbon on the breast, 1785). Tellingly, the enslaved were prohibited from wearing red ribbons on their days off. Justification for this legislation was the “insults and licentiousness given by negroes and coloureds” toward whites. Statia was described as “a land of evil,” probably due to the ruthless trading atmosphere that was more akin to a pirates’ den such as Port Royal, Jamaica, than to a refined market town. This law indicates that some enslaved people clearly thought that they were free to address white people in a less than respectful manner, and it also shows that on their days off they felt truly free, believing they could act and behave the same as free blacks, even in­clud­ing wearing a red ribbon. In addition, characterizing Statia as a “land of evil” suggests a social atmosphere where both whites and free blacks mutually tolerated a certain degree of rudeness, insolence, and animosity, thus imparting a greater sense of free­dom to the enslaved. This situation became more acute during French and British occupations of the island between 1795 and 1816 when restrictive trade regulations and taxes, particularly instituted by the French, caused severe economic decline. As homes and warehouses were abandoned across the island (Gilmore 2004), the enslaved of­ten pilfered their contents (Gilmore 2006a) for personal use and resale. Gambling and gaming among the enslaved also became widespread and was considered such a detriment to their work and the pub­lic welfare that French authorities tried to prohibit it in 1797. What made the extent of the problem so great was that free blacks and whites encouraged these behaviors, even providing places to gamble in their houses and yards, both in town and in the countryside.

36 / Gilmore

As Statia settled into economic despair during this period, the unfree took to selling rum and other alcoholic beverages to soldiers stationed on the island. A regulation passed in 1810 forbade this activity (DNA OASE 82 H.W. Pandt, H. W. Proclamation Forbidding military men from procuring [alcoholic] drinks, 1810, p. 175). By 1811, field slaves were illicitly selling yams in the pub­lic market from their fields as well as those allotted to them in their weekly allowance (DNA OASE 82 Thomas Barrow, Proclamation Regarding the Exportation of Yams, 1811, p. 197). Many yams were exported, and some people purchased large quantities to raise their price. This was deemed harmful to the poor and an inconvenience to both slaves and slaveholders. Broad conservation of resources was also encouraged because of fears that impending war with the United States would greatly reduce or cut off supplies to Statia. Additional proclamations in 1811 and 1812 (DNA OASE 82 Thomas Barrow, Trade with slaves, 1811, p. 206) show the extent to which the enslaved were engaged in illicit trade. Enslaved merchants were buying and selling so much stolen property from other unfree and free people that they were required to obtain written lists of the products that slaveholders had authorized them to sell. Nevertheless, these problems persisted as uninhabited houses were illegally demolished to sell the wood in the island’s market, and cattle theft was attributed to runaway slaves (DNA OASE 82 Thomas Barrow, Regarding destruction of uninhabited houses in the town, 1812, pp. 217–218; Regarding the prevention of cattle stealing, 1812, p. 223). These proclamations reflect the economic and social trials that engulfed the island and the enslaved that lived there under French and British occupation through 1816. The strife they were suffering at this point was just a precursor to the wholesale economic collapse that would overcome the island within the decade. Prior to this, unfree and free blacks were integral elements in trading activities on St. Eustatius. The trade on St. Eustatius once was probably similar to larger urban centers such as Charleston, South Carolina (Morgan 1998). In contrast, however, trading conditions were likely easier for unfree merchants on Statia.

Free Blacks and Provision Grounds All residents participated in the free trade on St. Eustatius and thus contributed to the colonial global economy (Emmer 1998; Paula 1993). Enslaved Af­ri­cans were so well integrated into trading relationships that opportunities abounded for self-­manumission (Gilmore 2004). The diversity and proportions of Statia’s population reflect these circumstances where a significant proportion of the population consisted of manumitted slaves. For example,

The Congo Free Black Village / 37

in 1715, the population was 41% white, 15% free black, and 44% enslaved (Table 2.1). By 1790 the white population had declined to 34%, and free blacks had reduced to 6%, while the enslaved represented approximately 60% of the population. By 1795 during the peak decade of Statia’s economy, over 7% of the population was of free Af­ri­can descent (Table 2.1). With the close of the Napoleonic Wars and the collapse of Statia’s economy, much of the white population abandoned the island, taking many enslaved with them. Free blacks did not leave the island in as high numbers. Thus, by 1819, the free black population was 13% of the total population or close to 40% of the entire free population. Similar demographic patterns have been identified for other Caribbean economies (Debow 1849). With the exception of St. Lucia and Mar­ti­nique, in much of the Ameri­ can south and the Caribbean the enslaved far outnumbered free people of color during the later 1700s (Table 2.1) (Armstrong 2003; Armstrong et al. 2009; Craton 1997; Cohen and Greene 1972; Fitchett 1950). In Jamaica in 1785, for example, 87% of the population was enslaved Af­ri­cans. In Barbados in 1786, free blacks made up a miniscule 1.1% of the total population. It is important to emphasize that the larger percentage of free blacks in St. Eustatius directly affected the amount of land and the size of plots set aside as provision grounds on the island. The enslaved could sell any provisioning surpluses, no matter how obtained, privately to other unfree or free blacks or in pub­lic markets if slaveholders granted them permission (Gilmore 2009). A wide variety of provisions was also purchased by slaveholders for their enslaved people. An external supply of provisions was particularly important to the Caribbean colonies where there was insufficient arable land for raising crops or where virtually all available space was devoted to intensive sugar production (Bates 2016). Any interruption to provisioning in the West Indies, especially from warfare or catastrophic weather, created severe complications for slaveholders and deadly consequences for the enslaved. Both the Seven Years’ War and the Ameri­can War of Independence caused considerable consternation if not outright panic among slaveholders in the Caribbean colonies (Sheri­ dan 1976). The effects were especially acute in the British colonies such as Jamaica, Barbados, and other islands, where thousands of unfree people starved because of their dependence on imported foods. Hurricanes that devastated provisioning grounds and disrupted transportation were prevalent in the last quarter of the eighteenth century, further complicating enslaved provisioning (Poey 1855; Sheridan 1976). As a result, some islands relied on neutral colonies, especially St. Eustatius, to supply them with the necessary staples. When the British took the island in 1781, these shipments stopped, further complicating the access slaveholders had to food across much

Table 2.1. Summary of White, Free Black, and Enslaved Populations on St. Eustatius, 1715–1863 Year

White

%

Free Black

%

Total % Free

Enslaved

% Total

Total

Source

1715 1786 1789 1790 1795 1818 1819 1847 1850 1863

524 3,000 2,375 2,886 2,400 501 507 766 782 no data

41.1 39.5 30.3 34.1 29.5 18.8 19.6 40.3 40.5  0.0

189 600 511 548 584 302 336 with whites with whites no data

14.8  7.9  6.5  6.5  7.2 11.3 13.0  0.0  0.0  0.0

26.5 16.7 17.7 16.0 19.6 37.6 39.9  0.0  0.0  0.0

561 4,000 4,944 5,042 5,140 1,865 1,747 1,137 1,150 1,087

 44.0  52.6  63.1  59.5  63.3  69.9  67.5  59.7  59.5 100.0

1,274 7,600 7,830 8,476 8,124 2,668 2,590 1,903 1,932 1,087

Knappert 1932:39 Valk 1889 Goslinga 1985:152 Goslinga 1985:152 Hartog 1976:52, 1978:59 Hullu 1913:433 Stemler 1876:64 Grevelink 1848 DeBow 1849:236 Kuitenbrouwer 1978:77

The Congo Free Black Village / 39

of the West Indies (Sheridan 1976:628). Although some authors (e.g., Marshall 1991) claim that provision grounds supplied the vast majority of enslaved people’s food through­out the eighteenth century in the Caribbean, it was only after the suffering endured as a result of war (the Ameri­can and French Revolutions and the Napoleonic Wars) and the weather that the pro­vision grounds on a majority of plantations were increased (Sheridan 1976). The flurry of legislation enforcing the quantity and quality of provision grounds that was passed after 1800 is not only reflective of abolitionist pressures but also the high death rates of the enslaved and slaveholders during the previous half century. There are no detailed archaeological studies of faunal remains from sites in Statia (primarily due to generally poor faunal preservation conditions), but preliminary observations at Pleasures Estate and the Duinkerk House provide some insight to slave provisioning and diet on the island. At Pleasures Estate, faunal evidence included a variety of fish remains in­clud­ing the pharyngeal bones of a parrotfish (Scarus spp. or Sparisoma spp.). Other marine animals included blue land crab (Cardisoma guanhumi) claw fragments and West Indian top shell (Cittarium pica) operculums, both species that are still exploited as food resources by Statians today (Gilmore and Goodrich 1998). Domestic species included chicken, cow, goat, and pig. Similar faunal evidence was recovered from a privy pit at the Duinkerk House, although there was a significantly higher proportion of fish here than any other animal (Gilmore 2004). Further comparisons among social classes are difficult to make, because the privy contained artifacts related to both the enslaved and slaveholder residents at this urban site. Ongoing organic lipid analy­sis of food residue recovered from an Afro-­Caribbean ware vessel may provide some additional insights into slave diet at this site. Surprisingly, no conch (Strombus gigas) shells were found at either location, although this marine species is heavily exploited today. Combining information from documents and the 1781 P. F. Martin map where provision grounds and sugarcane fields are clearly marked, it is possible to calculate the proportion of land dedicated to slave kostgronden (provision grounds in the Dutch Caribbean) for a dozen plantations. This shows that an average of 22% of each plantation was dedicated to provisioning (Gilmore 2004). At the Zorg en Rust plantation, not shown on the map, the quantities of land used for sugarcane, yams, and “negro grounds” are specifically identified in a deed from 1791 (DNA OASE 24, 91 Du Sart. ­FrPrisatie Gedaan maaken Vaste Goederen en Toebehooren [for Jacobus Seys, Senior], 1791, pp. 627–642). Eighteen percent of this plantation was dedicated to provisions. This suggests that the proportions of provision grounds calculated from the Martin map are reasonably accurate. The hectare of

40 / Gilmore

provision-­grounds-­to-­enslaved ratio was also calculated for each of the 12 plantations. On average, enslaved workers collectively had a total 4.3 ha of provision grounds or about 0.44 ha (1.08 acres) per slave to work on each plantation. In addition to growing sugarcane, the enslaved cultivated a wide variety of fruits and vegetables. Similar crops were grown on other islands, but on Statia the types of plants and the scale of their production is notably different. For example, beginning in the seventeenth century, citrus was grown extensively on the island for the provisioning of ships, and cotton, along with a variety of food crops, was exported until the mid-­eighteenth century (Goslinga 1990:375). Eighteenth-­and nineteenth-­century visitors commented on the vast quantities of yams produced on the island, which had a reputation for being the best in the Caribbean (Teenstra 1836:350).

Congo Free Black Village The Congo Free Black Village site is located on the east­ern outskirts of Oranjestad, St. Eustatius’s only town (see Figure 2.1). About 200 m to the north of the site is a burial ground traditionally known as the Congo Burial Ground. As depicted on the Martin map (1781), house lots run along a road frontage between the Congo Burial Ground and the Old Churchyard Cemetery that was the site of the seventeenth-­century Dutch Reformed Church (Knappert 1932; Martin 1781). The property owner requested that the St. Eustatius Center for Archaeological Research (SECAR) undertake an investigation on his property after noticing artifact concentrations on the surface. During excavations (2006–2008), a great deal was learned regarding vernacular architecture and creolization processes at work on the island. Excavations focused on three home sites identified during the initial phase of the investigation through surface finds and open-­area excavations. More than 17,000 artifacts were recovered from this site. The area was once divided into a series of seven house plots each covering approximately 0.01 ha (Martin 1781). Investigation concentrated on one of the middle plots (Figure 2.2). Colonial artifacts were scattered on the surface in several places. Excavation of 300 m2 revealed three domestic structures, a few animal burials, and several boundary walls and ditches (Gilmore, in press). The remains of a dry-­laid boundary wall were found running north to south, dividing this lot into two areas. Abutting this feature were colonial artifact concentrations related to the site’s eighteenth-­century domestic occupation, in­clud­ing glass, nails, ceramics, items of personal adornment, charcoal, and brick fragments. A hard-­packed “living surface” was encountered approximately 30–40 cm below current grade. Very few artifacts were

The Congo Free Black Village / 41

Figure 2.2. Site SE 132 plan showing sequence of structure occupation (1, 2, 4), outdoor cooking area (3), and earthen shrine (5). (R. Grant Gilmore III)

found on this surface except for several flint debitage concentrations and a few scattered unfinished gunflints, indicating that residents engaged in gunflint production at some point during their tenure on the site, as has been found by Ahlman and colleagues on neighboring St. Kitts (Ahlman et al. 2014). This is consistent with evidence that the enslaved and free engaged in a range of production activities to earn extra income across the Caribbean and the Americas (Berlin and Morgan 1991; Singleton 1985; Hauser

42 / Gilmore

2008; Heath and Bennett 2000; Heath 1999; Wilkie 1999, 2000). This enabled the enslaved to improve their lives and in some cases contributed to the purchase of their free­dom.

Structures Structure 1, the oldest domestic structural evidence at the site, was a wattle-­ and-­daub building and associated scatter of domestic artifacts (Figure 2.2). Structure 1 probably dates before circa 1760 and is comparable to a late eighteenth-­to early nineteenth-­century wattle-­and-­daub building excavated at the Coincoin plantation near Natchitoches, Louisiana (MacDonald and Morgan 2012). The dimensions of the extent structural remains were approximately 2.5 × 1.5 m. Structure 2 was a two-­chambered post-­in-­ground building approximately 8 × 3 m (see Figure 2.2). Corner postholes and post molds were 50–70 cm in diameter, with smaller postholes spaced at approximately 1 m intervals with even smaller post molds situated between these. The posthole ­pattern corresponds quite well with post-­in-­ground structures found across the colonial Atlantic World (e.g., Armstrong 1985, 2003; Carson et al. 1981; Deetz 1993; Gilmore et al. 2001; Kelly and Hardy 2011; MacDonald and Morgan 2012; Morgan and MacDonald 2004). Structure 2 was likely constructed entirely from wood with a thatch roof as indicated by the significant number of medium (3 to 5 cm) to large (5 cm+) wrought iron nails and the absence of clay daub. No smaller wrought nails were recovered, indicating the home was likely not shingled. Although few artifacts were recovered from the building, the preponderance of creamware indicates an occupation period spanning the 1760s to 1770s. Approximately 1 m to the southwest was a stone cobble-­ filled pit with an ash fill. The cobbles were topped by an Afro-­Caribbean yabba, or cooking pot fragments, suggesting that this was the cooking area for the compound residents. Structure 3 was located to the northeast along a north–south axis, opposite Structure 2, and measured approximately 8 × 6 m (see Figure 2.2). This structure differed from the other two excavated buildings in that it was not earth fast and instead was built on a low stone foundation made up of a mix of cobbles (dumped ballast from trading ships) brought from the beach along Oranjestad’s Lower Town and volcanic stone likely recovered from the immediate surroundings. Interestingly, the north­ern foundation wall was contiguous with this house lot’s east to west stone boundary wall. The structure was likely framed and sided with wood and had a cedar shingle roof as indicated by hundreds of wrought and machine-­cut nails of all sizes recovered in and around the structure. Lumber and cedar shingles were imported in large

The Congo Free Black Village / 43

Figure 2.3. Plan view of the earthen shrine at SE 132. (R. Grant Gilmore III)

quantities from the North Ameri­can colonies, especially South Carolina and North Carolina (Leonard and Pretel 2015; McCusker 1978; Morgan 1998). Spaces for wooden floor joists were also identified in the foundation walls. Recovered wrought iron thumb hinges indicate shuttered windows and/or door closures. Two fragments of a whiteware commemorative bowl, bearing the word “Freed” and the date “1863” were found within the building. The artifacts (in­clud­ing vari­ous pearlwares and some cut nails) suggest an occupation period starting late in the 1700s and stretching well into the 1800s.

Earthen Shrine Perhaps the most interesting feature identified from this site is an ovate earthen mound approximately 2 m in diameter, rising from the current grade approximately 30 cm. It was located about 1 m from the northwest corner of Structure 3 in a 3 m gap in the east–west boundary wall (Figure 2.3). Below this stone boundary wall was an earlier earth-­filled boundary ditch through which the earthen shrine intruded. Thus, the earthen mound was located at the intersection of boundaries along the cardinal N–S and E–W directions. The mound’s surface was covered with small, smooth cobbles, and a variety of red brick and terracotta floor tile fragments lay flat on its surface (Figure 2.3). On the mound’s surface were broken hollow-­form vessels, in­clud­ing Chinese porcelain cups, a hand-­blown glass case bottle and wine bottle kicks, and at least one Westerwald stoneware bottle base. Several larger rim and base sherds of glazed and unglazed Dutch coarse earthenwares were excavated from the shrine surface. One particularly unusual item was an intact pearlware ointment pot (5 cm rim diameter) with an un-

44 / Gilmore

identified maker’s mark consisting of the letters “——AW” above a “2” that is probably related to the volume of the container (the dashes represent illegible letters). All the vessels likely held liquid and dry offerings at the shrine. The feature fill extended only 10 cm into the subsoil resulting in a fill about 40 cm in total thickness. The feature fill was sterile with the exception of a disc-­shaped crumpled piece of gold weighing about 7 g recovered from the feature’s center. Initially, the mound was thought to be a burial mound. However, on sectioning the feature, the soil fill was found to be sterile with the exception of the gold offering. The lack of artifacts in the feature fill is highly anomalous on St. Eustatius archaeological sites as much of the surface of the island is littered with the remains of human occupation from the prehistoric period to the present. To my knowledge, no such mounds have been excavated in the Americas and, although documented in West Africa, none have been excavated in this region either. The evidence points toward only one interpretation: that this earthen mound was a shrine placed at a spiritually powerful location within the living landscape.

Site Interpretations The Congo Village occupation was continuous through three phases of vernacular architectural styles over the course of perhaps 150 years beginning in the mid-­1700s. Structure 1 represents the earliest site occupation and consisted of a wattle-­and-­daub building found commonly across the Caribbean during the colonial and postemancipation periods. The dwelling was made mostly of locally acquired materials. A liminal stage in the development of this landscape can be found with Structure 2, where evidence indicates an increase in purchasing power by occupants with greater access to manufactured goods such as wrought nails and imported ceramics. Finally, the construction materials used and artifacts associated with Structure 3 are equivalent to those of a middling family group on St. Eustatius (or the Atlantic World) during the latter 1700s and early 1800s. There was clearly access to significant financial resources for investment in improved building technology in­clud­ing wood siding and flooring and a shingled roof. This three-­step site evolution materially reflects a process directly correlated with Statia’s merchants’ rise to dominate the Atlantic trade by the end of the 1700s. Statia’s free blacks were clearly upwardly mobile, and their architecture reflects this over time (Gilmore 2009, 2013). The culmination of this diachronic process is manifest in the whiteware commemorative bowl fragments with the words “Freed” and the date “1863” that reflect progression from a relatively challenging personal economic situation to one of much better economic standing.

The Congo Free Black Village / 45

It is also clear that, at least in the St. Eustatius sociocultural environment, certain elements of West Af­ri­can spirituality/ritual and domestic space remained constant. Statian free blacks, as elsewhere in the Atlantic World, consciously maintained an affinity for some aspects of a culture that they were physically separated from for decades if not centuries. The earthen mound shrine has direct parallels with many cultural groups in West Africa. Among the vari­ous groups within the Black Volta region (forming the border between present-­day Burkina Faso and Ghana), earthen shrines were sacrificial sites erected by initial settlers usually under a large tree (Lentz and Sturm 2001). Lentz and Sturm go further to state that “sacrifices are made to the earth god held responsible for the fertility of women and the soil and for the peace of a particular area. Ritual control of the earth shrine simultaneously implies the right to distribute land to later settlers, sometime in exchange for quite substantial gifts, and the late-­comers must also request the earth priest’s permission to build houses and bury their dead” (2001:158). Similar spiritual significance is assigned to earthen shrines through­out West Africa, in­clud­ing the Dogon in Mali (Douny 2011), Yoruba in Nigeria (Probst 2011), and Igbo in modern Nigeria (Samford 2007). Samford (2007) also hypothesizes that subfloor pits (of­ten located on the interior of homes) served a similar role among the enslaved in colonial Virginia as covert expressions of connectedness to Africa. No similar earthen mound shrines have been excavated thus far in the Caribbean and to my knowledge in the Americas. It appears that, at least among free blacks, overt spiritual practices clearly tied to West Af­ri­can traditions were permissible on St. Eustatius during the slavery period. Behaviors of­ten characterized as subversive or otherwise overtly threatening to the status quo in other slave societies were of­ten relegated to hidden or concealed environments obscured in living landscapes (Ruppel et al. 2003). Whereas on St. Eustatius, it appears agency was not subverted by powerful actors to the extent that it was in other Caribbean societies. Here overt expressions of his­tori­cal cultural affinities were more openly practiced—with an earthen shrine built in plain sight at a village. Hollow-­form vessels indicate that traditional libational offerings were practiced at this shrine—a common West Af­ri­can practice (Thompson 1993). The location of the shrine along a property boundary also correlates with many West Af­ri­can traditions (Lentz 2003). The offering of gold indicates the power of this shrine in the literal and spiritual landscape for site occupants. At this time, it is impossible to say precisely which of the dozens of West Af­ri­can societies influenced the construction of this earthen shrine over 200 years ago. The absence of artifacts in the liminal spaces between buildings and boundary walls also indicates long and strong ties to West Af­ri­can traditions regarding domestic space. The space outside of structures was just

46 / Gilmore

as ­significant—­or even more so—as the sheltered interior space of a home (Battle-­Baptiste and Franklin 2011). Domestic landscapes included yards swept clean of grass and trash (Heath and Bennett 2000; Morgan 1998; West­macott 1992; Wilkie and Farnsworth 2005). This is different from the value ascribed to excavated European and creole settlers’ yards on St. Eustatius and elsewhere that are of­ten cluttered with artifacts (Barka 1988, 1990; Gilmore and Goodrich 1998; Gilmore and Miller 2011). For people of Af­ri­ can heritage, yards were extensions of domestic space—a portion of which may be given over to planting greenery, but only vegetation that was edible in the form of fruits and vegetables. The yard space around the domestic structures in part of the Congo Free Black Village was swept clean with artifact concentrations built up along the peripheral walls. Unlike West Af­ri­can spiritual and domestic practices, the free blacks living on St. Eustatius were clearly focused on ascending the socioeconomic ladder characteristic of the aspirational European cultural and economic values recognized by other island residents. As income increased, house construction technologies evolved, using increasingly more expensive (and imported) materials. The diversity and quantity of other material cultural remains also increased through time. This process perhaps culminated in the acquisition of a commemorative plate celebrating either emancipation in the Dutch Antilles or more likely Abraham Lincoln’s Emancipation Proclamation, both occurring in 1863. Artifacts from the earthen shrine suggest that its construction and use are associated with the occupants of Structure 2, who had achieved sufficient wealth that they could afford to place an offering of several grams of gold within the shrine. It may be that the offering was a celebration of thanksgiving for successfully earning free­dom as so many other enslaved people on St. Eustatius had done (Gilmore 2004, 2006a, 2009). The juxtaposition of what was likely a home indistinguishable from many middle-­class families on the island that overlooked a shrine completely West Af­ri­can in origin perfectly illustrates the dichotomies found in creole cultures in the Caribbean and in much of the world.

Conclusions The data recovered during an intensive archaeological investigation of a free black village site on the Dutch Caribbean island of St. Eustatius provide evidence for a multicomponent occupation over 100 years. The archaeological data, when supplemented by documentary sources, indicate that the lives of the unfree and those of free blacks were in many ways much different from those in other Caribbean islands and the Ameri­can South. In gaining free­

The Congo Free Black Village / 47

dom, the formerly enslaved on St. Eustatius were actively engaged in improving their material well-­being as well as maintaining a spiritual life tied to West Af­ri­can traditions. Slavery was in fact a liminal state in the transition from West Af­ri­can to a creole life. The gradual evolution of purchasing capacity, the ability to build and occupy ever more permanent structures, and the agency available for spiritual expression are all reflected at the Congo Free Black Village. Ceramic types, building materials, and building locations all indicate a freer existence for the formerly unfree. Until now, very little research has been reported for free black archaeological sites in the Caribbean. Robust documentary, comparative his­tori­ cal and archaeological evidence have been combined to form a clearer understanding of the socioeconomic and religious lives of free blacks on St. Eusta­tius. Population dynamics reflect changing geopo­liti­cal and economic transformations that affected both free and unfree actors across the region. Documentary evidence from Dutch archives provides a window into the unique life circumstances of St. Eustatius free blacks when compared to those living elsewhere in the Caribbean. Archaeological evidence indicates an evolving purchasing power that peaked just prior to emancipation in 1863. Perhaps the most intriguing discovery was the earthen mound located in an auspicious intersection on the Free Black Village site. The mere existence of this feature in a Caribbean context further underscores a continued tie to West Af­ri­can religious traditions that could not be extinguished through the Middle Passage experience. Social, economic, and religious agency were all manifested in vari­ous overt forms at this unique archaeological site.

Acknowledgments I would like to acknowledge my Statian brother Joshua Spanner, the landowner who first noted artifacts on the surface at the site and provided us with both moral and logistical support through­out our excavations. I also would like to thank the St. Eustatius Center for Archaeological Research Board of Trustees and the dozens of SECAR volunteers who worked on this site for many months. I would especially like to acknowledge Derek R. Miller, Brian Mok, and Stafford Smith.

3 Jamestown, Nevis, and Urban Resilience in the Early English Caribbean Carter L. Hudgins, Eric Klingelhofer, and Roger H. Leech

During four short field seasons between 2003 and 2006, a team of archaeologists and historians associated with the Department of Archaeology at the University of Southampton and the Nevis His­tori­cal and Conservation Society investigated the traditional site of Jamestown, Nevis, the sec­ond-­ largest urban center on the island during the seventeenth century (Figure 3.1). The goal of the Nevis Heritage Project was “to investigate the prehistoric and historic physical and social landscapes of Nevis as a sample of human history in the Lesser Antilles of the East­ern Caribbean prior to the destruction of that history by natural and human agencies” (Morris et al. 2003:7). At Jamestown, the team employed a regimen of geophysical testing and limited excavation to determine the size and character of the site, and to construct a chronology for the site that placed the town in the settlement history of Nevis and the development of towns in the colonial British Caribbean. The archaeological investigation of Jamestown addressed a third purpose: since the nineteenth century a romantic myth persisted that a tsunami destroyed the town and washed its residents out to sea in 1690 (Iles 1871:8; Hubbard 1996). Was there more to Jamestown’s story than this myth? Instead of abandonment, excavations at Jamestown found evidence of fluctuations in intensity of occupation between the town’s formation in the late seventeenth century and its ultimate demise in the mid-­nineteenth century. Evidence of resilience and persistence rather than destruction and abandonment led the project to (1) compare the urban character of the site with the island’s dominant plantation landscape, (2) compare patterns of occupation at the site with fluctuations in the island’s population, and (3) compare Jamestown’s rise and fall with broader economic patterns in the east­ern Caribbean. Wracked by the tsunami of 1690, Jamestown was not abandoned but persisted for more than a century, assailed by war, drought, and economic competition until it ebbed away in the postemancipation era. The history of this small seventeenth-­century town on the leeward shore of Nevis is entangled with the tsunami of 1690. An eyewitness wrote that

Figure 3.1. Leeward Islands and site of Jamestown on west coast of Nevis. (Carter L. Hudgins)

50 / Hudgins, Klingelhofer, and Leech

the disaster unfolded on the afternoon of April 5, “about five of the Clock.” The eyewitness reported, “We heard a rumbling Noise, like that of distant Thunder, . . . from the Bowels of the great Mountain.” Seismic events had rattled colo­nists on Nevis and other Caribbean islands before this, but this event was different: “So strong was the Motion that . . . [a] few Moments after the Noise began, . . . a[n] . . . Earth Quake, which shook the whole island to that degree, that all the Houses in Charles Town that were built of Brick or Stone, dropt of a sudden down from the Top to the Bottom.” As the earth shook, the sea suddenly went flat. More ominous, “The sea itself for a time forfeit the Shoar for about three quarters of a Mile together, and left a great Number of Fish of a large size to lye gaping upon the Sand, till it returned again.” The eyewitness reported that the “violent Motion of the Water happen’d diverse times” (Smith 1690). We can imagine that when the water “returned again” its “violent motion” lifted whatever was loose, throwing it together in mats of debris, slamming it into walls, houses, fields, trees, or anything else that lay in its path. It is here, however, at the end of the eyewitness account, that his­tori­cal fact, his­tori­cal imagination, and his­ tori­cal memory became tangled. Jamestown, then the sec­ond-­largest town on Nevis, was one of the places slammed by the “violent Motion of the Water.” The legend of Jamestown— a town tumbled down by the rising sea, its residents swept away, all lost save a few who ran to higher ground—first appeared in an amateur history of Nevis published in 1871 (Iles 1871:8). Like other fanciful Victorian tales, the story of Jamestown’s destruction appears to be a concoction of great inventiveness and imagination, spun, in part, from fact. John Alexander Burke Iles, a postemancipation justice of the peace and colonial secretary for Nevis, took the truth and gilded it in a romanticized history of the island he called An Account Descriptive of the Island of Nevis, West Indies. Iles incorrectly dated the earthquake to 1680 and wrote that “the capital Jamestown situate on the north west margin of the sea coast was submerged by an earthquake, carrying with it, its population and wealth” (Iles 1871:8). An earlier history of the island published in 1745 provided a less dramatic account. Writing in 1720, three decades after the earthquake struck, William Smith, then minister of St. Thomas Parish, recounted in his Natural History of Nevis and the Rest of English Leeward Charibee Islands in America only that Nevis had experienced dozens of earthquakes. But, Smith wrote, “none of them did us any farther harm, then frighting us, and cracking the walls of a few Boiling Houses and Cisterns.” He further wrote, “About twenty years before my arrival, there was so violent a Shock, that the Sea retired a good Furlong from Charles Town, and in two minutes, or a little more, came back again to its usual bounds.” Smith said nothing about lost towns. He did anticipate

Jamestown, Nevis, and Urban Resilience / 51

some future catastrophe, predicting that “Charles Town will one time or other sink down into the sea (by shock of a more violent one) as Port Royal in Jamaica did” (Smith 1745:62–65; O’Loughlin and Lander 2003:58–59; Mulcahy 2006). Charlestown never sank as Smith predicted, but it was perhaps in his conjecture that the myth of Jamestown, a place that did indeed disappear, was born. Facile, superficial, and wrong, Iles’s version of history won, gaining official recognition in 1920 in the form of a new Ordinance Survey Map of Nevis (issued by United Kingdom’s Department of War) that marked a site on the island’s northwest coast with this label: “Site of Jamestown, destroyed 1680.” Which narrative is right? Was Jamestown really one of the “relics of this terrible visitation,” drowned beneath the sea like much of Port Royal, Jamaica, when it was struck by an earthquake in 1692? Iles likely knew about that infamous place and its earthquake. Smith certainly did. Where Smith speculated, however, Iles perhaps conflated, drawing a conclusion from what he knew of the more famous of the two places, exaggerating in the process the effects of the 1690 Nevis earthquake.

Sugar, Slaves, War, and Commerce: The His­tori­cal Context Nevis is a small island, roughly 11 km (7 mi.) long and 8 km (5 mi.) wide. Neighboring St. Kitts joined Nevis in forming the independent Federation of St. Kitts and Nevis in 1983. Like the other Leeward Islands, Nevis is a volcanic peak that rises to a height of 985 m (3,232 ft.). Nevis was covered by thick vegetation and trees when Columbus first visited in 1493; now, most of the sugarcane fields that later climbed the shoulders of Mount Nevis have reverted to forest. The traditional and accepted site of Jamestown is located on the leeward side of the island and is today bisected by the island’s coastal shore road (Figure 3.2). Ruins of stone walls pierce the ground surface on both sides of the road. The terrain here slopes downward gently from south to north toward a ghut that slices along the north­ern edge of the site. The site also slopes gently from east to west, falling away more abruptly on the west side of the coast road where a modern commercial coconut grove shades stone foundations, one of them the ruins of a small square structure being slowly enveloped by a banyan tree. Other stone foundations occupy a wedge between the coconut grove and the shore road right-­of-­way. The ghut empties here into a brackish lagoon that forms the north­ern boundary for both the coconut grove and the site. A heavily restored ruin now called Fort Ashby, a masonry demilune once part of the island’s eighteenth-­century coastal defenses that faced Morton’s Bay and the Caribbean Sea, occupies a low site on the north­ern side of the ghut (Machling 2012:249–257).

52 / Hudgins, Klingelhofer, and Leech

Figure 3.2. Plan of excavations and ruins, Jamestown, Nevis. JT = Jamestown feature. (Carter L. Hudgins)

A settlement along the shore of Morton’s Bay had coalesced by 1672 (Morris et al. 2003:31–38). An inventory made of forts on the island in that year contains the first mention of Morton’s Bay where a cannon “platform,” or small coastal fortification, was located. Morton’s Bay was made a lawful port in that same year (Wheeler, Sir Charles, 1672 TNA CO 1/29/161:​ 14/12/1672; TNA CO 154/1/114:1672). Governor William Staple­ton reported to the Board of Trade and Plantations in 1676 that Morton’s Bay was then the sec­ond-­largest town on the island. Even so, it was still a small place: “In Nevis . . . [of] five places for trade, there are but two considerable. First Charles Town where there are good dwellings and store houses . . . [and] Morton’s Bay, the other place of trade where there are the like houses, but few in number because the ships doe ride at Charles Town and doe send their long boats to Morton’s Bay to fetch their lading from thence” (Stapleton,

Jamestown, Nevis, and Urban Resilience / 53

William, 1676, Letter to Board of Trade and Plantations, 22 No­vem­ber 1676 TNA CO 1/38/162:22/11/1676). Few in number, the houses at Morton’s Bay were roofed with thatch. Nevisian officials, seeking to reduce the risk of fire, soon banned this practice with an act “for the suppression of thatcht houses” in Morton’s Bay and Charles Town. Passage of similar acts in 1680 (“against the making of dangerous fires in Charles Town and Morton’s Bay”) and a requirement in 1682 that chimneys be built of brick or stone attested to both the town’s persistence and its residents’ inclination to use local, if combustible, roofing material (Nevis Acts 1671, 1681, 1682, 1683). These acts suggest that housing density in Morton’s Bay had reached a point that a fire could spread from house to house and threaten the entire settlement. Jamestown as a place name first appears in late 1684 when the town was enumerated as one of the island’s “lawful places receiving goods.” From that year on, the name Morton’s Bay all but disappears from pub­lic rec­ ords, replaced by Jamestown. Why the change in name? The application of this new name for a relatively new place perhaps reflected royalist sentiment as the reign of Charles II drew to a close and England anticipated a new monarch, James II. While seventeenth-­century maps give the small settlement the same name as the shallow bay where it coalesced, Morton’s Bay, later maps drawn by French cartographers label the town “Petitbourg” or “Jacqueville” (Figure 3.3). Maps drawn in the early nineteenth century, however, no longer marked Jamestown. In place of the older name there is on an 1818 map a new label, “Pleasure House,” a designation that signals both a change in function and perhaps a severing of his­tori­cal memory between seventeenth-­century commerce and early nineteenth-­century recreation (Langley 1818 in Machling 2012:122). Following English settlement in 1628, the island’s social and economic trajectory was much like Barbados, where mixed agricultural cultivation (spices, tobacco) gave way to the single-­minded cultivation of sugar. Nevisian colo­nists made the transition to sugar quickly. Historian David Gaspar has observed that Nevis was the first of the Leeward Islands to make a “decisive move” from tobacco to sugar (Gaspar 1985:66–68). By 1652, the island’s population reached 5,000, and sugar had overtaken tobacco, cotton, and indigo as the island’s preferred crop. Nevis was, it was reported, “esteemed the best island for sugar: it makes little of any other commodities, only some tobacco to windward, which is valued more than any of the English plantations” (Sheridan 1974:162). Even though Nevis was by the 1670s, one observer wrote, still “not half planted for want of Negroes,” it was said to be the most advanced of the Leeward Islands (Bridenbaugh and Bridenbaugh 1972:13). By the end of the seventeenth century the island’s population had reached 6,000, one-­half of it enslaved. Cane fields soon covered

54 / Hudgins, Klingelhofer, and Leech

Figure 3.3. “Carte de L’Isle de Nevis” from Geographical Description of the Antilles Possessed by the English (Paris: M. Bellin, 1758). The site of Petit Bourg (Jamestown) appears along northwest coast.

the lower slopes of Mount Nevis, and terracing pushed cultivation above 300 m (1,000 ft.). Sugar production rose to 6,000 and then 7,000 hogsheads annually. Forts constructed on the leeward side of the island protected sandy coves and bays that provided easy commercial access from attack (Mach­ling 2012; Leech 2008:127–138). Landings to facilitate trade were strategically placed around the island. One of them, a small aggregation of warehouses

Jamestown, Nevis, and Urban Resilience / 55

and domestic buildings at Morton’s Bay, later called Jamestown, emerged as colo­nists created the commercial infrastructure necessary to ship sugar to English merchants in Lon­don, Bristol, and Plymouth. European competition for advantage and profits of­ten pulled Nevis into war. Soon after initial settlement of the island, colo­nists beat off Spanish attempts to reestablish its claims to the island. Conflicts between France and England periodically roiled Nevis from 1660 to 1713 (see González-­Tennant and González-­Tennant [chapter 9 herein] for additional discussion of the military history of Nevis). Militia forces on Nevis beat off a French assault in 1666–1667. Shore batteries drove off a Dutch fleet in 1673. During King William’s War (1688–1697), Nevis was the base from which English forces attacked other Caribbean islands. Nevis’s luck ran out, however, during Queen Anne’s War (1701–1713). Under Pierre Le Moyne D’Iberville, French forces, seeking retribution for an English attack on Guadeloupe in 1702, successfully landed on the island’s southeast coast in 1706 and defeated the island’s militia. The French stayed long enough to “damage enemy property rather than annex it” (Dunn 1972:23). As they withdrew, the French, in an act of early modern industrial sabotage, burned cane fields, sugar works, and windmills, and dismantled and removed mills and coppers. It was later reported that the French also took 3,187 slaves, nearly all of the island’s slave population (Dunn 1972:23). When the Treaty of Utrecht restored prewar po­liti­cal boundaries in 1713, the Caribbean became a less volatile region. Peace did not restore Nevis’s former prosperity or its privileged position. In the years that followed French withdrawal, Nevis ceded its once dominant economic position to other islands. In 1720 William Matthew, governor of the Leeward Islands, observed that “but now and for years past the chief Trade . . . is at Antigua, next and very near to St. Christophers” (Matthew 1734). Matthew concluded his assessment that “Nevis has quite lost its trade, and is a desert island to what it was thirty years ago.” As early as 1708, Antigua had passed Nevis in both population and sugar production. One colonial official reported that Antigua had expanded “both in extent and trade and by consequence more business is done there” (Gaspar 1985:68). In 1734 Thomas Butler, agent for Nevis, wrote that “the Island, from being the seat of trade of the four, is no longer so” (Butler 1734). There were then, Butler complained, only 400 men on the island capable of bearing arms. He attributed the decrease in the number of white men “to the great sufferings and losses they underwent in the late warr and from the great decay the sugar trade hath of late fallen into and to the weak and defenceless condition in which the island is at present and the great danger it is in of being lost.” Butler observed that “the great decay in the sugar trade” had convinced “many of the inhabitants . . . to retire to North America.” Nevis

56 / Hudgins, Klingelhofer, and Leech

was then not attracting new colo­nists, “no new settlers having come either in their place or to supply the great decrease in their white inhabitants by the death of the old settlers.” What would it take to restore confidence and draw new investment in people and capital? The memory of the damage inflicted by the French in 1706 was, nearly three decades later, still fresh. Butler concluded that preparing for some future French attack was the best way to reverse the island’s population decline. He concluded that “unless a speedy supply of ordnances and small arms with a sufficient strength of ships of warr for their defence be sent them, it is to be feared many of the inhabitants who still remain on the Island will retire from the same on account of the great dread and apprehension of a French invasion” (Butler 1734). Other British officials shared Butler’s views. Governor Lord Howe reminded the Council of Trade and Plantations in the same year that the Leeward Islands “lye much exposed by reason of the many easy landing places therein.” There was, he warned, great danger of some future attack and probability of French success. “The number of their inhabitants, who have never been very numerous,” Howe wrote, “hath decreased one half since the Peace of Utrecht” (Howe 1734). Fluctuations in the island’s population, a cause for official alarm, mirrored the ebb and flow of the island’s sugar-­based economy through the eighteenth century. Archaeological and his­tori­cal investigations that followed ours, particularly those conducted by the Digital Archaeological Archive of Comparative Slavery at Jessups and New River plantations, confirmed sugar’s persistent ability to shape the island (Galle et al. 2009; Meniketti 2006:​ 53–80). As the movement toward emancipation gathered momentum in the first decades of the nineteenth century, some of the island’s planters increased their operations, expanding their labor forces and investing heavily in industrial capacities to crush cane and boil sugar. Construction of new sugar works was one reflection of a wave of consolidation of estates that put more cane fields under the management of fewer planters just as opposition to slavery gained po­liti­cal momentum at home. Emancipation ultimately dashed expectations for increased sugar production. The demise of sugar precipitated a new round of official reports and private complaints that echoed eighteenth-­century assessments that consistently portrayed Nevis as a place intent on capturing better times.

Archaeology of a Small Urban Place: The Evidence of Disaster and Adaption Modern archaeological exploration of Jamestown began when British Chan­nel 4’s Time Team archaeologists and historians conducted excavations

Jamestown, Nevis, and Urban Resilience / 57

at James­town in 1998. Three years later, improvements to the shore road prompted additional investigations. Archaeologists working with the Nevis Heritage Project observed and recorded four stone foundations visible at the ground surface and, like the earlier Time Team effort, conducted geophysical exploration of the site. Both projects drew additional attention to the site that both tradition and island myth identify as Jamestown (Meni­ketti 2006:60–61). Four additional field seasons (2003–2006) under the aegis of the Nevis Heritage Project followed (see Figure 3.2). The first, in 2003, focused on ascertaining the boundaries of the site, mapping surface artifact distribution, and completing a geophysical survey of the portion of the site that occupied a coconut grove to the west of the shore road (designated Area II). Geophysical exploration of this portion of the site provided disappointing results. Soil resistivity readings fluctuated wildly in the damp, salty sands of the coconut grove, not only pulled, analy­ sis later determined, by high conductivity of the sandy soil but also confused by deep crab burrows and numerous large buried stones. Produced by eruptions of the island’s volcanoes, Nevisian stones are magnetically charged, a geological fact that was an additional source of electronic “noise” in collected geophysical data. Resistivity was unable to distinguish crab burrows from buried stones or buried stones from buried walls. Land crabs did, however, provide valuable information. Crab burrows, many dozens of them, disturbed the west­ern portion of the site. The tailings of these burrows were littered with artifacts that delineated the west­ern boundary for the site and marked an edge to occupation that ran northwest to southeast from a brackish pool fed by the ghut that drained the slopes above the site. These finds also appeared to reveal the seventeenth-­century shoreline of Morton’s Bay. Beyond the artifact boundary is deep sand, evidence, we surmised, of the filling of what was once a broader and deeper bay. Surface collection of artifacts established a south­ern boundary for the site, indicating that the settlement stretched some 500 m south from the ghut along what had been the seventeenth-­century shoreline. Two short field seasons investigated stone foundations that broke the ground surface as well as buried walls and foundations in the two occupation clusters that defined Areas I and II. Both areas contained deeply buried, complicated patterns of walls and foundations. Excavation of small, exploratory units established stratigraphic contexts and a building sequence for the walls and foundations. Larger area excavations conducted in 2006 established chronological relationships between stone structures identified across the site and clarified the timing, pace, and character of the settlement. In both Area I and Area II, repeated repair, rebuilding, and replacement of walls provided evidence for two phases of occupation, an initial phase of

58 / Hudgins, Klingelhofer, and Leech

activity that extended from about 1680 and ended about 1720 followed by a sec­ond phase that began about 1790 and ended about 1840. Metropolitan slipware, Borderware milk pans, lead glass baluster-­type wine glass stems and pipe bowls, among them tightly datable pipe bowls, confirmed European occupation in the third quarter of the seventeenth century. Absence of white salt-­glazed stoneware tended to confirm earlier observations that occupation here ended before 1720. There were no artifacts in the nearly 1 m of soil that accumulated over the earliest walls in Area I, clear evidence of a hiatus of activity. At both building clusters, superimposed walls and artifacts associated with both their construction and destruction signaled a later reoccupation of the site. This sec­ond phase of occupation began after 1790 and extended until about 1840. Fragments of creamware, transfer-­printed pearlware, whiteware, and case bottles dominated the artifact assemblages associated with this sec­ond phase of occupation. Exploration of Area I discovered thick accumulations of soil-­covered walls erected in the third quarter of the seventeenth century that were rebuilt and then abandoned (Figure 3.4). North Devon pans and pipe bowls mixed into the soil matrix used in the construction of the first phase of this wall also confirmed initial active occupation at this portion of the site to the third quarter of the seventeenth century. In 2006, the team opened a 2 by 6 m trench excavated entirely by hand and a 9 by 8 m area in which the overburden (approximately 0.40 m thick) was removed mechanically. The intention of opening these larger units was to reveal the entire ground plan of the building discovered during initial survey two years earlier. A more complicated picture, however, emerged. Running through both units were the walls of not one but six buildings, each built successively closer to the ghut in a pattern of serial rebuilding. Every wall was truncated at the north and east. Modern utility trenches struck the walls of three of these structures (Structures 9, 11, and 12) (Figure 3.5). Structures 9 and 10, whose walls were, respectively, 0.80 and 1.15 m wide, were, stratigraphically, the deepest and oldest encountered in Area I. The remnants of Structure 9 consisted of three courses of irregularly coursed stones laid in a mud mortar wall that survived to a height of half a meter. The south face of this wall consisted of irregular unworked stones. Stones on the north­ern face were mortared in place, the mortar in the joints between stones struck flat. This sparing use of lime-­based mortar in a wall made up otherwise of stones laid in mud mortar distinguished Structure 9 from Structure 10, whose walls were entirely laid in mud mortar. Three hand-­wrought nails remained in the north­ern face of Structure 9. This line of nails, each 50 cm below the top of this section of surviving wall, had been inserted into the mortar. A portion of the shank of each nail protruded beyond the fin-

Figure 3.4. Overview of Area I excavation showing superimposed foundations. (Carter L. Hudgins)

Figure 3.5. Plan of excavation Area I showing superimposed foundations of seventeenth-­century (Structures 9 and 10) and nineteenth-­century (Structures 11, 12, 13, and 14) buildings. (Carter L. Hudgins)

60 / Hudgins, Klingelhofer, and Leech

ished mortar surface, evidence perhaps that these nails had once fastened a board, possibly a form, applied against this side of the wall during construction. The mortared side of the wall is thus interpreted as its exterior face. Structure 10 abutted the east­ern end of Structure 9 and was of similar character but set slightly askew to the alignment of its neighbor. A fragment of a North Devon sgraffito milk pan and the neck of a green glass case bottle were embedded in the wall of Structure 9. Both artifacts date from the third quarter of the seventeenth century. Above the erosional soils that had buried Structures 9 and 10 were truncated foundations of four buildings, remnants of a later generation of construction that overlay the two seventeenth-­century stubs of walls. Destruction debris associated with the terminal occupation of this complex contained artifacts dating from the sec­ond quarter of the nineteenth century. The surviving walls of Structure 11 were 0.75 to 0.95 m wide and consisted of round stones, irregular in size laid with large facing stones on the interior and exterior faces of the wall and a core of small stones and cobbles mortared in place. A similar method was employed to raise the wall of Structure 12 (0.70 m wide), but this wall (its north-­side orientation paralleling that of Structure 11) consisted of dressed stones laid in mortar with a mortared core of small stones and stone fragments. Structure 13 (0.55 m in width) overlays Structure 12 and exhibited a similar character of material and construction. Structure 14 abuts the truncated end of the surviving section of Structure 13. The slightest of these walls, Structure 14, was 29 cm wide and consisted of parallel lines of regularly coursed dressed stones laid in lime mortar. Ceramic and glass wares that circulated widely in Atlantic markets from the close of the eighteenth century to circa 1840 (creamware and transfer-­ printed pearlware) were discovered in the destruction debris associated with these four wall segments. The absence of white salt glazed stoneware was particularly telling in defining two periods of occupation separated by an interval of little to no activity. White salt glazed stoneware is a finely potted stoneware whose forms include a wide range of table wares in­clud­ing capuchin cups, plates, and drinking jugs. It circulated widely through­out the British Atlantic, appearing initially in 1720. The absence of white salt glazed stoneware indicates that the initial period of occupation at Jamestown that began in the third quarter of the seventeenth century ended about 1720, that is, before the appearance of this ceramic ware. The absence of this ceramic marker from the thick alluvial strata that covered the site suggests an interval of little to no activity on the site from about 1720 to about 1790 when the presence of transfer-­printed pearlware indicates the resumption of activity. Reoccupied toward the end of the eighteenth century, this portion of the site appears to have been abandoned a sec­ond time about 1840.

Jamestown, Nevis, and Urban Resilience / 61

Figure 3.6. Excavation Area II showing superimposed stone foundations of seventeenth-­ century Structure 6 under nineteenth-­century Structure 7. (Carter L. Hudgins)

The ruins and buried walling investigated in Area II reflected equally complicated phases of repair and rebuilding (Figure 3.6). Stone walls visible at the surface proved to be part of the sec­ond nineteenth-­century occupation of the site. Stones cut to generally uniform size and shape laid with lime mortar in regular courses characterized this sec­ond phase of construction. These nineteenth-­century buildings also bore evidence of successive rebuilding. On the higher portion of Area II adjacent to the shore road, Structure 3 replaced an earlier building (Structure 2), occupying a nearly identical footprint and using stones robbed from it. Transfer-­printed and annular pearlwares, whitewares, and green glass case bottles dated the end of occupation here to about 1840. Excavation revealed a similar building sequence lower on the site and closer to the shore where the gable end of a large rectangular building (Structure 7) appears to have been entirely rebuilt (Figure 3.6). The ridge of the roof of this building was oriented on an east–west axis perpendicular to the shoreline. The foundations and walls of this structure were identical to the materials used in the construction of Structures 2 and 3. Stones cut and shaped into a uniform size and whose exterior faces were dressed were laid in regular courses. Voids between facing stones were filled with a matrix of small stones, cobbles, and other materials mixed into thick dumps of mortar to form the core of the walls of this nineteenth-­century building. A shallow stoop butted against the north­ern wall of Structure 7 indicates the position

62 / Hudgins, Klingelhofer, and Leech

of a doorway that for a time opened onto a yard that lay between the structure and the ghut that separated Jamestown from Fort Ashby. This feature may have been added during reconstruction of the building’s west gable. During both of its construction phases, Structure 7 was 9 m wide. Here too sherds of polychrome transfer-­printed pearlware plates and serving dishes and fragments of molded green glass case bottles were found in the robbers’ trenches associated with the destruction of Structure 7. This evidence suggests a date of destruction for this building about 1840. An earlier building, Structure 6, lay under and askew to nineteenth-­ century Structure 7. Both the walls and a partially surviving mortar floor in Structure 6 showed signs of significant displacement. Squared, dressed stones formed what appeared to be the jamb of a doorway that led into what was interpreted as a lobby entrance at the southeast corner of the structure. Ramped stones just inside this feature suggested narrow stairs that once led to an upper level of the building. Beyond this possible lobby entrance and along the building’s west wall survived remnants of a mortar floor that continued into the balk and unexcavated portions of the building. Deeply cracked, the floor surface on the west­ern edge of the building tilted down and west into the lobby. The floor surface at the north­ern edge of the excavations appeared, in a stratigraphic section along the balk, to undulate and to pitch up from north to south. The warping, undulating condition of the mortar floor in Structure 6, the oldest of the ruined buildings in Area II, was distinct to that building. The weight of accumulated sand pressing against the surviving walls of Structure 6 had also skewed its walls. The upper courses of stones in the structure’s south­ern wall had tumbled above its lower course, cracked and displaced by as much as 10 cm. Stones in the south­ern face of Structure 6, especially those related to the door jamb, were pitched some 17 degrees out of plumb. Subsidence, wracking, and rippling of the mortar floor and walls of Structure 6 may be indications of seismic activity. The surviving stone walls associated with Structure 7, although diminished by robbing of some dressed stone, had not subsided or buckled. A conclusion can be drawn that the two generations of buildings had experienced different geological forces. Excavations in Area II discovered other buried walls. A 1.5 × 1 m extension of Area II revealed, just at the east­ern edge of the unit, a 60 cm wide wall oriented north to south in a line that paralleled the gable ends of Structure 6. More intriguing, a 2 × 2 m unit found additional evidence of early occupation at the site under the corner of a large and very solidly constructed nineteenth-­century building (Structure 8). Carefully dressed stones formed the outer face of this structure whose foundation extended slightly more than a meter below the modern ground surface. The mortar used in

Jamestown, Nevis, and Urban Resilience / 63

this structure was very hard, an indication of a higher lime content than the softer, sandier mortars used in Structures 6 and 7. The size, purpose, plan, and date of this structure remain undetermined. More tantalizing was that this solid stone structure overlay an earlier phase of occupation, the earliest and most intriguing encountered on the site. Just at the water table, more than a meter below the surface and under this nineteenth-­century founda­ tion, ceramic sherds and small white kaolin clay pipes dating to the sec­ond quarter of the seventeenth century were mixed in a charcoal layer that contained iron fragments, clinkers, slag, links of iron chain, and the priming pan of a snaphaunce (musket or pistol), evidence, perhaps, of the first phases of the island’s settlement.

Conclusions: Jamestown in the Shadow of Nevis Nevis was reputed to be the most prosperous of the Leeward Islands in 1678. The single-­minded pursuit of sugar made it so. John Oldmixon reported in 1700 that little if any tobacco had been planted on Nevis for the previous 30 to 40 years. Oldmixon wrote, “sugar only taken care of; of which great Quantities have been made and 50 or 60 ships loaded in a year from this Island to Europe” (Oldmixon 1741:2:238). Nevis claimed early primacy among the Leewards. The island claimed four of the ten churches built in the Leewards in the seventeenth century and had at Charles Town the region’s only well-­built port. The royal governor of the Leewards resided there (Sheridan 1974:161). That reputation was soon lost. The island suffered heavy population losses in 1689 and 1690 when Governor Christopher Codrington estimated that 1,500 men died of fever. As a result, Codrington reported that “the strongest of the Leeward islands has become the weakest” (quoted in Pares 1968:23). This acute population loss was part of a longer debilitating trend traced by historian Richard B. Sheridan (Sheridan 1974:164). In the fourth quarter of the seventeenth century, the white population on Nevis began what Sheridan has described as an “irregular” downward trend that continued through the eighteenth century. By 1775 only 1,000 white men, women, and children resided on the island, less than a third of its 1678 population. This decline, attributable in part to disease, was also propelled by shifting strategies in sugar production. Large planters consolidated small landholdings, a process at work well before the end of the seventeenth century that mirrored a shift already completed on Barbados. Historian William Pares has estimated that in 1719 there were perhaps as many as 100 plantations on Nevis. A century later there were perhaps three dozen (Pares 1968:23). The demographic trajectory of the island’s enslaved population hewed

64 / Hudgins, Klingelhofer, and Leech

a different path (Table 3.1). Between 1672 and 1678, the enslaved population on Nevis leaped from 1,739 to 3,860. Reports by colonial officials indicate that despite the French invasion, the slave population remained level, expanding in the 1720s to more than 5,000 and growing to 11,000 by 1775. Fluctuations in the slave population marked the decades that followed. At emancipation, Nevis was home to 10,722 slaves and 700 free people of color. These two distinct demographic paths, a long decline by the island’s white population contrasted with the general pattern of growth by the enslaved, reflect planter determination to continue sugar production. The pattern for the island’s white population maps more comfortably with the architectural and archaeological patterns of the emergence, decline, reoccupation, and final abandonment discovered at Jamestown. When the French attacked Nevis in 1706, some of the island’s best days were already in its past. French depredations shook an island already in decline. There were additional challenges to recovery. In the fall of 1725 rumors of a slave insurrection gripped the island amid reports that some slaves plotted to “cut off the whites, and take the island for themselves” (Gaspar 1985:68; Zacek 2009). Threats of a possible slave rebellion amplified other bad news. In that same year Governor John Hart informed colonial authorities that Nevis was “in the most deplorable Condition from the Dry weather” (Hart, 1726, Report to Board of Trade, Janu­ary 1726, TNA CO 152/15/1 [Board of Trade: Original Correspondence, Leeward Islands, 1725–1727]). In aggregate, reports from colonial administrators in the decades that followed the French invasion of 1706 emphasized population loss, slackening trade, and vulnerability to attack. In short, Nevis experienced a period of economic and demographic retraction following the French invasion. The tsunami of 1690 did not destroy Jamestown. The archaeological foot­print at Jamestown describes cycles of urban growth and failure or resilience and revival that lifted and dropped Jamestown twice. The town’s first era of dynamic growth lasted from the 1660s until about 1720. The lull in activity at Jamestown during the middle decades of the eighteenth century (1720 to 1780), was, rather, a result of the shock of the French invasion and the economic and demographic chaos that followed, all exacerbated by drought, the threat of slave rebellion, and rising competition from neighboring islands. The evidence of stone walls, their repair, rebuilding, and replacement, and artifacts associated with them narrate the arc of the sugar economy from initial European settlement at Morton’s Bay to prosperous late seventeenth-­century town, to 1690 tsunami, to ruin and desolation compounded by French depredations, to general abandonment. There was, after this gap in significant activity, a late eighteenth-­century resurgence of investment that is mirrored elsewhere on the island. Indeed, much of the

Jamestown, Nevis, and Urban Resilience / 65

Table 3.1. Nevis Population, 1672–1842  Year

White

 Slaves

1672 1678 1708 1720 1729 1734 1745 1756 1775 1788 1834 1836 1842

Not Reported 3,521 1,104 1,343 1,296 Not Reported   847 1,118 1,000 1,514 Not Reported Not Reported   700

 1,739  3,860  3,676  5,689  5,646 Not Reported  6,511  8,380 11,000  8,420  8,915  9,250* 10,722**

Total  7,381  4,780  7,032  6,942  6,330  7,368  9,498 12,000 10,254 11,422

* = Number of apprentices ** = Number of free blacks in the post emancipation period Sources: for 1729, http://www.archive.org/stream/1953colonialrecordsc41gre auoft/1953colonialrecordsc41greauoft_djvu.txt, accessed July 10, 2009; all other years, Sheridan 1960, 150. Governor Matthew reported that the total population of Nevis in 1734 was 6,330 (Governor William Matthew to Board of Trade, Commissioners for Trade and Plantations August 31, 1734, A State of H. M. Leeward Caribbae Islands in America); for 1842, R. W. Allen, The World in Miniature (London, 1842); Robert Montgomery Martin, Statistics of the Colonies of the British Empire (London, 1839)

surviving evidence of the island’s plantation landscape is the result of early nineteenth-­century building and most of it associated with the construction of new sugar works and a movement toward the application of steam power to cane crushing. This construction reflects a kind of “doubling down” in the face of growing abolitionist sentiment, a bet against what, even to the most tone deaf West Indian planters, must have looked like the inevitable end to enslaved labor (Galle et al. 2009; Meniketti 2006). Jamestown revived as the island’s sugar economy recovered in the fourth quarter of the eighteenth century. New clusters of buildings appeared in both Area I and Area II, superimposed on older foundations (see Figures 3.4 and 3.5). Some of the building and rebuilding evident at both clusters of buildings may be evidence of a sec­ond devastating French attack in 1794. And some of this sec­ond generation of buildings relate to what an 1818 map labels a “Pleasure House.” Evidence of domestic use, high quantities of glass and ceramic objects, among them large serving vessels such as platters and

66 / Hudgins, Klingelhofer, and Leech

tureens, and straight-­sided, square mold-­blown green glass bottles popu­larly called case bottles, soon accumulated around them, obscuring evidence of earlier activity even more. The ruins justice of the peace John Alexander Burke Iles saw at Jamestown were for him “relics of this terrible visitation” and became grist for his fanciful 1871 history of the island and proof of Jamestown’s demise. Nevis was in 1871 not as wealthy as it once was, and it may be that Iles, seeking some four decades after emancipation a culprit for the island’s diminished position, found it easy to blame the earthquake for the island’s decline. A dramatic explanation was, if nothing else, easier for readers to grasp than slow economic decline. Missing in this explanation were the shifting fortunes of the sugar trade, the rise of competitors in Antigua, depredations by the French in 1706, and demographic decline. The fates of Jamestown and the more infamous Port Royal reveal factors that alternately supported or weakened urbanism in the early English Caribbean. Both towns were largely destroyed (in 1690 and 1692, respectively) by earthquake and tsunami. In Jamaica, even though Kingston became that island’s new population and commercial center, the site of Port Royal was too important militarily to be abandoned (Mayes 1972). Jamestown, in comparison, could not recover, tied as it was to a sugar economy and plantation sys­tem that saw landholdings on Nevis concentrated into fewer and fewer large estates that had less need of subsidiary ports, especially a port whose harbor was silted by earthquake and then erosion. Jamestown-­ on-­Nevis, then, is the story of a flourishing “frontier” settlement that, after coalescing to facilitate shipping sugar, was subsequently injured mortally by natural calamity and war, fading away between about 1720 and 1780 when it reemerged as a satellite of the colonial capital as a suburban “Pleasure House” for “Caribbee” aristocrats. With emancipation and the collapse of the plantation system, Jamestown faded away a sec­ond time, its roles in the island’s history recalled only in romantic legend.

Acknowledgments Our debts over the course of four field seasons are many, and our thanks go first to the members of the field crews, especially Eric Deetz and Captain David Kohler (USN Ret.) and our students from Mercer University and the University of Mary Wash­ing­ton. Both universities and the University of Southampton provided material and logistical support on which we depended as did the Nevis His­tori­cal and Conservation Society. To John Guilbert who owned the equestrian center that now occupies much of the site of ancient Jamestown we owe much for his interest in our work, his patience

Jamestown, Nevis, and Urban Resilience / 67

with us, and his support. Our chapter is all the better thanks to comments from several anonymous reviewers and for encouragement and advice we received from editors Todd Ahlman and Gerald Schroedl. Our benefactor Vincent K. Hubbard passed away before this chapter was published, but we dedicate it to him for his friendship, his enthusiasm for Nevisian history, and his support of our excavations at Jamestown.

4 Inter-­and Intraisland Trade of Afro-­Caribbean Ware in the Lesser Antilles Todd M. Ahlman, Gerald F. Schroedl, Barbara J. Heath, R. Grant Gilmore III, and Jeffrey R. Ferguson Afro-­Caribbean ware is a low-­fired, earthenware pottery that was made and traded by enslaved and freed Af­ri­cans in the Caribbean. The pottery was first produced in the seventeenth century and production continued through the postemancipation period. On some islands pottery production has continued well into the twenty-­first century. Over the past 50 years, the analy­ sis and interpretation of Afro-­Caribbean ware have become criti­cal to understanding the lifeways of freed and enslaved Af­ri­cans. Because the pottery constitutes a direct material linkage between freed and enslaved Af­ri­cans, these ceramics are essential to studies of his­tori­cal cultural transformations in the Caribbean. In this chapter we use the all-­encompassing term “Afro-­ Caribbean” ware rather than island-­specific terms like Afro-­Kittitian ware or Afro-­Nevisian ware. It is unclear whether island-­specific terms apply to the users or makers of the pottery, and given the extent of interisland trade documented here, we prefer the all-­encompassing term. Afro-­Caribbean-­ware studies have generally focused on the role that the pottery played in food preparation, ceramic production, internal markets, and cultural identity. These studies have employed a variety of macroscopic (Ahlman et al. 2009; Heath 1988; Young 2002), petrographic (Hauser 2008; Hauser and Kelly 2011; Heath 1988; Gilmore 2004), and chemical analyses (Ahlman et al. 2008, 2009; Hauser 2011b) and generally have taken an intraisland approach focusing on internal production and markets (Hauser 2011c). Interisland trade has been alluded to in these studies but generally has not been the main focus of these investigations. Recent analyses focusing on Afro-­Caribbean ware production and trade across the Caribbean have shown the utility of employing instrumental neutron activation analy­sis (INAA) in studying clay sources, production, and trade of the pottery. With this in mind, we expand here on previous studies (Ahlman et al. 2008, 2009; Hauser et al. 2008; Kelly et al. 2008) by analyz­ ing Afro-­Caribbean ware and clay samples from Antigua, Montserrat, and Nevis (see Hudgins et al., chapter 3 herein); St. Croix and St. Eustatius (Statia) (see Gilmore, chapter 2 herein); and St. Lucia and St. Kitts (Table 4.1).

Trade of Afro-Caribbean Ware / 69

Our goals were (1) to identify clay and production sources, (2) explore intraand interisland trade networks, and (3) learn about past behaviors of free and enslaved Af­ri­cans. First we look at past studies of Afro-­Caribbean ware and island geography, then we present the results of our INAA study, and finally we present a discussion of the results relative to our research questions.

Previous Macroscopic and Microscopic Research The study of Afro-­Caribbean ware began in the 1960s with Jerome Hand­ler’s studies (1963, 1964) of coarse earthenwares and pottery making from Antigua and Barbados. Studies through­out the 1960s, 1970s, and early 1980s of coarse earthenwares from Jamaica (Mathewson 1972a, 1972b; Ebanks 1984), Nevis (Grigsby 1962; Platzer 1979), and St. Croix (Gartley 1979) brought the significance of Afro-­Caribbean ware to light within his­tori­cal archaeological circles. Barbara Heath’s 1988 doctoral thesis on Afro-­Caribbean ware from Statia was the first in-­depth analy­sis of Afro-­Caribbean ware. Heath’s goal was to study the range of forms and attributes of the non-­European coarse earthenwares found in Af­ri­can site contexts on Statia and to determine whether the earthenwares were a product of an Afro-­Caribbean potting tradition (Heath 1988). Heath examined over 1,200 sherds (Heath 1988:152) from the Honen Dalim Synagogue site (also designated SE217) to develop a paste and inclusion typology that was then compared to three other sites (SE1, SE19, and SE2) to assess its validity. Using macroscopic paste characteristics and inclusions, Heath discerned six types from the Honen Dalim Synagogue assemblage. When Heath compared the type classifications with the other sites she identified two additional types. The macroscopic types were verified by thin-­section petrography of 49 sherds by Christopher Bell (Heath 1988:Appendix 1). Bell identified eight microscopic types (Types A–H) with Type F divided into two subgroups. All but one of Bell’s types correlated with Heath’s macroscopic types. Grant Gilmore (2004) followed Heath’s research by examining 41 Statian Afro-­Caribbean ware sherds via thin-­section petrography as well as looking at pottery from Nevis (n = 2), St. Lucia (n = 7), Antigua (n = 5), and St. Croix (n = 5). Gilmore explored interisland ceramic variation among Afro-­ Caribbean sherds from three Statian sites to see whether any sherds exhibited similar characteristics to Afro-­Caribbean ware from the other islands. The 41 Statian sherds came from three sites: Pleasures Estate, the Duinkerk House, and Battery St. Louis. The petrographic study led to the identification of four groups within the Statian material with most sites containing more than one group. In their studies, Heath and Gilmore found a variety of paste recipes that

Table 4.1. Sites with Number of Sherd Samples in Afro-Caribbean Ware Study by Island Island

Site

Antigua

Indian Ridge Fort George on Monk’s Hill

Site Type

Time Period

Number of Sherds   1

Military

  1

Unknown

  2

Montserrat

Carr Estate

Sugar plantation

Mid-seventeenth century to nineteenth century

 14

St. Croix

Estate Prosperity

Sugar plantation

Mid-seventeenth century to nineteenth century

  5

St. Lucia

AN02

Nevis

Jessups

Sugar plantation

Mid-seventeenth century to nineteenth century

 45

New River

Sugar plantation

Mid-seventeenth century to nineteenth century

 51

Fenton Hill

Sugar plantation

Mid-seventeenth century to nineteenth century

 10

Mountravers

Sugar plantation

Mid-seventeenth century to nineteenth century

 10

Upper Rawlins

Sugar plantation

Mid-seventeenth century to nineteenth century

 10

Charlestown Waterfront

Urban site

Mid-seventeenth century to nineteenth century

 10

Brimstone Hill Fortress

Slave and British officer contexts at British fort

1780s–1830s

 47

Earl Romney Slave Village

Sugar plantation slave village

1800–1820

  7

Southeast Sugar Peninsula Site 1 plantation slave village

1750–1800

 10

St. Kitts

  7

Trade of Afro-Caribbean Ware / 71

Table 4.1. Continued Island

Site

Site Type

Time Period

Number of Sherds

Southeast Sugar/cotton Peninsula Site 2 plantation slave village

Late eighteenth to nineteenth century

 10

The Spring

1800–1800

  6

Late seventeenth century to early nineteenth century

  1

Sugar plantation slave village

St. Eustatius Battery St. Louis Defensive position Duinkerk House Urban site

1740–1800

 16

Honen Dalim

1738–1795

 18

Synagogue

Pleasures Estate Sugar planta- Ca. 1750–1830s tion converted to cattle/goat farm

 21

Simon Doncker Urban site House

Mid-seventeenth century to nineteenth century

  1

Warehouse 1

Commercial site

Mid-seventeenth century to nineteenth century

  1

Warehouse 2

Commercial site

Mid-seventeenth century to nineteenth century

  3

Total

307

they concluded likely represented local as well as nonlocal clay sources. Gilmore found similarities in all the materials, which he concluded was a function of the similar geological base characteristic of the islands where the sherds were found. Heath concluded that most of sherds she examined had pastes and inclusions consistent with the volcanic geology of Statia as well as the neighboring islands. But some inclusions, like the presence of mica, led Heath to hypothesize that some of the sherds had inclusions from nonvolcanic islands. In addition to Heath’s and Gilmore’s studies, Derek Miller (2008) ex-

72 / Ahlman, et al.

amined the commercial production of earthenware sugar molds made by enslaved Af­ri­cans on Statia. Miller concluded that the commercial manufacture of these sugar molds was congruent with Statia’s role as a primary trading depot in the Caribbean. Because sugar production was minimal on the island, their presence represented the illicit trade of sugar products through the islands as planters attempted to hide some of their products from the heavy tariffs from the English, French, and Spanish governments. Antiguan pottery traditions and Afro-­Caribbean ware have been the subject of study for over 50 years (Handler 1964; Nicholson 1984, 1985, 1990; Heath 1988; Petersen et al. 1999; Hauser and Handler 2009). Handler (1964); Nicholson (1984, 1985, 1990); and Heath (1988) examined the contemporaneous (to their studies) Antiguan pottery tradition, relating it to the his­tori­ cal period. Through time, it became apparent that the Antiguan pottery tradition had become focused on tourist and local markets, and by 2009 only one woman on the island was making pottery (Hauser and Handler 2009). Petersen and colleagues (1999) examined paste characteristics and pottery inclusions from Antigua as well as smaller samples from Barbuda, Montserrat, and Anguilla. They identified two general types in their study: Category 1 that has volcanic tuff inclusions typical of “tuff-­rich” geological environments that occur on islands like Antigua; Category 2, which has black mineral inclusions derived from volcanic materials most likely from the islands of St. Eustatius, St. Kitts, Nevis, and Montserrat. They found Category 1 and 2 sherds in the Antigua, Anguilla, and Montserrat assemblages but only Category 1 sherds in the Barbuda assemblage. The findings suggested to them that there was a trade network among the enslaved and freed Af­ri­cans living on these islands. Jean Howson (1995) briefly mentions Afro-­Caribbean ware from Galways and Delvin plantations on Montserrat in her doctoral thesis but conducted no chemical or petrographic analyses of the sherds. Michael Young (2002) conducted an analy­sis of pottery from the Galways plantation simi­ lar to Heath’s analy­sis (1988) of the Statia Afro-­Caribbean ware and used her types as the basis for his types. Young macroscopically examined the 724 sherds from the Galways plantation by first sorting the sherds according to paste attributes (smooth to coarse), grouping the sherds into four types, and then describing the paste and inclusion characteristics. All of Young’s types were dominated by white angular to subrounded minerals that he thought were “variations of calcite” (Young 2002:61) with clear and black minerals also present. The black minerals were interpreted as mica in his Type 1 and augite in Types 2, 3, and 4. Young also subjected 14 sherds to X-­ray radiography and found that they exhibited a random particle distribution, suggesting to him that they were molded or paddled rather than coiled.

Trade of Afro-Caribbean Ware / 73

Research into Afro-­Caribbean ware from Nevis, St. Lucia, and St. Croix has been limited. Gartley’s identification (1979) of Afro-­Cruzan wares was one of the first studies of slave-­made pottery, but there have not been extensive studies of pottery on St. Croix since then. St. Lucia pottery studies have generally focused on the modern pottery tradition on the island. The study of pottery made by enslaved laborers on Nevis began recently with work at the Jessups and New River sites (Galle et al. 2009). Their study concluded that the decline of Afro-­Caribbean ware production through time was due to improved material conditions that enslaved people experienced in the early nineteenth century. INAA has been used over the past 10 years to study Afro-­Caribbean ware, and some of the first studies are presented in a 2008 special publication of the Journal of Caribbean Archaeology (Descantes et al. 2008). The volume includes articles on Precolumbian pottery (Conrad et al. 2008; Crock et al. 2008; Descantes et al. 2008; Fitzpatrick et al. 2008; Isendoorn et al. 2008; Siegel et al. 2008) and historic period Afro-­Caribbean ware from St. Kitts (Ahlman et al. 2008), Jamaica (Hauser et al. 2008), and Mar­ti­nique and Guadeloupe (Kelly et al. 2008). The Afro-­Caribbean ware studies mainly focused on intraisland trade (Ahlman et al. 2008 and Hauser et al. 2008), although Kelly and colleagues (2008) examined pottery trade in the French West Indies and with neighboring British islands. The Afro-­Caribbean ware from archaeological investigations at the Brimstone Hill Fortress National Park has been the subject of two published studies (Ahlman et al. 2008, 2009). Ahlman and colleagues (2009) macroscopi­ cally examined paste and inclusions of 665 Afro-­Caribbean ware sherds from Brimstone Hill. Eight types were drawn from differences in primary and sec­ondary inclusion classes. Black and opaque mineral inclusions were found in five of the types with differences in size, amount, and sec­ondary inclusion classes marking the differences among these types. A small num­ber of types were marked by black sand inclusions, mica and schist, and quartz. INAA was conducted on 40 sherds from Brimstone Hill and five St. Kitts clay sources (Ahlman et al. 2008, 2009). The INAA analy­sis led to 38 sherds and four clay samples being assigned to four groups. Two sherds and one clay sample were not assigned to any of the groups. There was no apparent correlation between the macroscopic types and INAA groups with at least two of the macroscopic types being assigned to each INAA group. Three clay samples were assigned to Group 1 and one was assigned to Group 2, which led Ahlman and colleagues (2008, 2009) to conclude that the bulk of the pottery was manufactured with St. Kitts clay. The production and trade of pottery by enslaved and freed Af­ri­cans on St. Kitts were interpreted as a risk management strategy (Ahlman et al. 2009).

74 / Ahlman, et al.

Under this scenario, potters made and traded the pottery to create social networks via trading partners that allowed them to limit their economic, social, and physical risks, such as food shortages, brutality, and death, inherent to enslavement. For Ahlman and colleagues (2009) “pottery ­production . . . was just one of many mechanisms implemented by enslaved Af­ri­cans to cope with variation in nutrition and the severities of slavery” (39).

Island Geology and Soils Parent geology typically influences the chemical composition of clays, but how this translates to the study of the chemical composition of pottery has not been widely studied. That being stated, we briefly explore the geology of the Lesser Antilles here to gain a better understanding of how to possibly identify clay sources, production centers, and trade networks. The Lesser Antilles arc is a chain of small islands 800 km long that runs from the Virgin Islands in the north to Grenada in the south (Figure 4.1) (Macdonald et al. 2000:6). The islands have formed along the subduction of the Atlantic floor under the Ameri­can plate (Macdonald et al. 2000; Wadge 1984), that continues to result in the formation of volcanic islands along this arc. To the north of Mar­ti­nique the arc is split: the east­ern, older islands are known as the Limestone Caribbees, and the west­ern islands are active and more recently formed, known as the Volcanic Caribees (Macdonald et al. 2000; Wadge 1984). The Limestone Caribees have been active since the Eocene to Mid-­Oligocene epochs while the Volcanic Caribees have been active from the early Miocene to present (Macdonald et al. 2000:6). The more recent volcanic islands (Pleistocene to present) occur in relatively narrow zones (less than 10 km wide) that occur within three zones: Saba to Montserrat, Guadeloupe to Mar­ti­nique, and St. Lucia to Grenada. Geologically, the Lesser Antilles has been divided into north­ern, central, and south­ern zones (Brown et al. 1977; Macdonald et al. 2000; Toothill et al. 2007). The north­ern islands (Saba to Montserrat) are composed of a tholeiitic suite of rocks; the central islands (Guadeloupe to St. Vincent) are calc-­ alkaline rocks; and the south­ern islands (Grenada and the Grenadines) are alkaline, silica-­undersaturated, highly magnesian rock varieties (Brown et al. 1977:798; Macdonald et al. 2000:5). In general, the rocks of the north­ ern islands are andesites and dacites with rare rhyolites on St. Kitts and Statia. The central island rocks include andesites with basalts, dacites, and rare rhyo­lites. The south­ern islands rocks are basalts and basaltic andesites. The geochemical composition of the rocks on the islands is a product of the localized magma and the surface materials from the subduction layer.

Trade of Afro-Caribbean Ware / 75

Figure 4.1. Location of islands in study. (Todd M. Ahlman)

Toothill et al. (2007) note the presence of multiple magma domes at each island and that variation within island geochemistry is present. In the broad sense there is an unsurprising overlap in the geochemical composition of the rocks of the different island groups (north­ern, central, and south­ern) as well as localized variation in elements, especially the rare earth elements. Macdonald and colleagues (2000) note that in a broad sense there is a gradation from low-­potassium tholeiitics in the north­ern islands to low-­and medium-­ potassium calc-­alkalines in the central group, to medium-­potassium C-­and M-­Series basalts in the south­ern islands (see also Brown et al. 1977). In addition, there are high-­calcium group rocks found within the C-­Series Grenada rocks, Statia, St. Kitts, Redonda, Guadeloupe, Domi­nica, and Bequia

76 / Ahlman, et al.

and low-­calcium rocks on Saba, Montserrat, Mar­ti­nique, St. Vincent, and the Grenada M-­Series rocks. When combined, the potassium and calcium composition of the rocks shows that the north­ern islands (Saba to Montserrat) are low-­potassium and high-­calcium types; Guadeloupe to St. Lucia are medium-­potassium and high-­calcium types with minor low-­potassium on Guadeloupe; St. Vincent is low-­potassium and low-­calcium rocks; Bequia is low-­potassium and transitional between low and high-­calcium; and Grenada and the Grenadines are mixed. In regard to trace elements and rare earth elements, Macdonald and colleagues (2000:21–22) note that the C-­and M-­Series basalts found on Grenada and the Grenadines tend to have higher amounts of barium, niobium, rubidium, strontium, thorium, zirconium, and light rare earth elements as well as higher radiogenic strontium and less radiogenic neodymium than the low-­potassium and medium-­potassium rocks of the north­ern and central islands. There is, however, overlap within these groups except for the C-­ Series basalts, and Macdonald and colleagues (2000:64) believe that they are derived from very similar parent magma materials and conditions. Toothill and colleagues (2007) found variations in the trace elements from St. Kitts and hypothesized that contaminants from the subduction slab as well as the presence of at least two magma domes cause these differences. The presence of contaminants and multiple magma domes across the Lesser Antilles has resulted in localized variation that is likely normalized in a pan­regional approach; therefore, the geochemistry of the parent materials across the Lesser Antilles may have greater intraisland variation than seen within the north­ ern, central, and south­ern island groupings discussed here. There is little research about the geochemistry of Caribbean clays, and, in fact, the most extensive research probably relates to the INAA of clay samples for Afro-­Caribbean ware studies. The generalized studies of the recent rocks and volcanic materials on places like St. Kitts do not allow for direct comparisons to the clay samples as the clays tend to occur near the older volcanic craters where there has not been chemical analy­sis conducted on the parent material. There probably are similarities in soils within the three broad zones just mentioned that mirror the zones seen in the vol­canic materials, yet finding these similarities is beyond the scope of this study.

Sherd Samples in This Study To meet the goals of identifying clay sources and production centers as well as the interisland trade networks, 307 Afro-­Caribbean ware sherds from 24 sites on seven islands and 23 clay samples from five islands were submitted to the University of Missouri Research Reactor (MURR) for analy­sis by

Trade of Afro-Caribbean Ware / 77

INAA (see Table 4.1). The 307 sherds include samples from Antigua, Nevis, Montserrat, St. Croix, St. Lucia, St. Kitts, and Statia. MURR also com pared the results with data from six sites on Nevis and one site on St. Kitts in the MURR database (Ferguson and Glascock 2010a, 2010b; Ferguson 2011), and these data are included in the INAA interpretations presented here. In addition, 11 clay samples from Statia, St. Kitts, St. Lucia, and Antigua and 12 samples in the MURR database from Nevis are included in the analy­sis. The sherd samples came from a variety of sites that date from the early eighteenth century to the late nineteenth century (see Table 4.1). The 80 sherds from St. Kitts come from the Brimstone Hill Fortress (n = 47), the Earl Romney plantation slave village (n = 7), two plantation slave village sites on the island’s southeast peninsula (n = 20), and the Spring plantation (n = 6) (see Table 4.1). The seven clay samples include one from the Conaree Hills and six from four different locations on the southeast peninsula. Forty of the sherds were reported by Ahlman and colleagues (2008, 2009). During the eighteenth and nineteenth centuries, Statia was one of the biggest and most active ports in the world. At one point, over 600 warehouses lined Oranjestad’s waterfront serving the trade of sugar as well as the trade goods that travel into and out of the Caribbean (see Gilmore, chapter 2 herein) (Gilmore 2004). The 61 Afro-­Caribbean ware sherds from Statia come from Pleasures Estate plantation (n = 21), Battery St. Louis (n = 1), the Duinkerk House (n = 16), Honen Dalim Synagogue (n = 18), Simon Doncker House (n = 1), and two warehouses in lower Oranjestad (n = 4) (see Table 4.1). Two clay samples from Statia were also analyzed (Table 4.2). Twenty-­three of these sherds were those subjected to petrographic study in Heath’s doctoral thesis (1988) and 38 were petrographically studied by Gilmore (2004) in his doctoral thesis. In his doctoral thesis, Gilmore also petrographically examined small sherd samples from three sites on Antigua (n = 4), Estate Prosperity on St. Croix (n = 5), the New River site on Nevis (n = 2), and Site AN02 on St. Lucia (n = 7), and these are included in this INAA study (see Tables 4.1 and 4.2). Clay samples from Sea View farm on Antigua and Morne Sion on St. Lucia collected by Heath were also included in the INAA study. The MURR INAA database also includes data from the New River (n = 49) (Ferguson and Glascock 2010b), Jessups (n = 45) (Ferguson and Glascock 2010b), Fenton Hill (n = 10) (Ferguson 2011), Mountravers (n = 10) (Ferguson 2011), Upper Rawlins (n = 10) (Ferguson 2011), and Charlestown Waterfront (n = 10) (Ferguson 2011) sites on Nevis and The Spring (n = 6) on St. Kitts (Ferguson and Glascock 2010b) (see Table 4.1). In addition, there are clay samples from the New River roadside, Beaumont, Clay Ghaut, Hichmans, Potworks, and Jessups locations on Nevis (see Table 4.2).

78 / Ahlman, et al.

Table 4.2. Location and Number of Clay Samples in the Study Number of Samples

Island

Location

Antigua

Seaview Farm

 1

St. Lucia

Morne Sion

 1

Nevis

Jessups New River Beaumont Clay Ghaut Hichmans New River roadside Potworks

 5  2  1  1  1  1  1

St. Kitts

Southeast Peninsula Site 2 Southeast Peninsula clay survey Conaree Hills

 1  5  1

St. Eustatius

Venus Bay Unknown

 1  1

Total

23

Instrumental Neutron Activation Analysis The sherd and clay samples were analyzed by INAA following standard MURR procedures that can be found on the website of the University of Missouri Research Reactor (MURR 2017) and in Glascock (1992). Interpretation of the INAA results also follows standard MURR techniques with initial analy­sis by principal components analy­sis to identify patterns in the chemical composition among the sherds. Mahalanobis distance was also calculated for each sherd to assess group membership based on similarities in the chemical composition of the paste recipes.

INAA Results The 307 Afro-­Caribbean ware sherds submitted to MURR for INAA analy­ sis and interpretation led to 272 sherds being assigned to 11 compositional groups (Table 4.3, Figure 4.2). Thirty-­five sherds, 13%, were not assigned to a compositional group. Nine clay samples have a greater than 0.5% probability of belonging to one of the groups. Compositional Groups 1 (n = 181) and 2 (n = 37) include the majority

Trade of Afro-Caribbean Ware / 79

of the sherds (see Table 4.3) and are very similar in chemical composition (Ferguson and Glascock 2010a and 2010b; Ferguson 2011). The Group 1 samples are higher in thorium, lanthanum, and arsenic than Group 2 sherds, which are lower in lutetium, ytterbium, scandium, and europium. The most obvious difference is Group 1 sherds generally have higher thorium content than Group 2 sherds (see Figure 4.2). Compositional Group 1 consists of sherds from 15 sites on four islands (Ferguson and Glascock 2010b; Ferguson 2011). The group includes 34 sherds from four St. Kitts sites, 13 sherds from three Statia sites, two sherds from two Antigua sites, and 132 sherds from six Nevis sites. In fact, all the sherds from Nevis except for one assigned to Group 3 and three that were unassigned to a group belong to Group 1. Of the 181 sherds assigned to Group 1, a number of sherds (n = 30) have abnormal concentrations of one or two elements and statistically are different from the rest of the group but carry so many similarities that they are included with Group 1 as two subgroups. Ferguson and Glascock (2010b) have termed the first Group 1b (n = 27) and sherds from St. Kitts (Brimstone Hill), Statia (Duinkerk and Honen Dalim), Antigua, and Nevis (Jessups, New River, Fenton Hill, Upper Rawlins, and Charlestown Waterfront) have been assigned to this group. Ferguson (2011) further identified Group 1c (n = 3) based on higher zinc and lower thorium content than the remainder of Group 1. Only sherds from the Mountravers site on Nevis fall into this group. Two clay samples from St. Kitts’s southeast peninsula and seven from Nevis have a higher than 0.5% probability of being within Group 1. The clay sample from the Beaumont area on Nevis falls within Group 1b, suggesting that this subgroup was made with Nevis clay. Two samples from St. Kitts’s southeast peninsula have the highest percent probability (12.87 and 5.94) of being within Group 1. Compositional Group 2 consists of 37 sherds with 31 from four St. Kitts sites (Brimstone Hill, Earl Romney, South­east Peninsula Site 1, and The Spring) and six from two Statia sites (Battery St. Louis and Pleasures Estate). The bulk of the sherds come from Brimstone Hill (n = 23). One clay sample from St. Kitts’s southeast peninsula plots with Group 2, but based on the Mahalanobis distance analy­sis is not statistically within this group. It is probably fair to state that sherds from Groups 1 and 2 were made with locally available clays that most likely came from Nevis and possibly from St. Kitts. The paste and inclusions in the sherds are very similar, and the chemical analy­sis also bears this out. The very similar geological history for both islands has led to similar clays and inclusions, making macroscopic paste differences difficult to discern as well as making it difficult to discern a source based on chemical analy­sis of the paste.

Table 4.3. Allocation of Samples to Instrumental Neutron Activation Analysis (INAA) Groups Island

Site

Antigua

Indian Ridge Fort George on Monk’s Hill Unknown

 1

1b

2

3

4

5

6

7

8

9

10

12

 1  1

Montserrat

Carr Estate

St. Croix

Estate Prosperity

St. Lucia

AN02

Nevis

Jessups New River Fenton Hill Mountravers Upper Rawlins Charlestown Waterfront

 39  45   7   7   5   7

Brimstone Hill Fortress Earl Romney Slave Village Southeast Peninsula Site 1 Southeast Peninsula Site 2 The Spring

 14  1   3   6  10

St. Kitts

1c

13

5  4  4  3

 1

Unassigned

Total

 1  1

  1   1   2

 1

 14

 5

  5

 2

  7

 2  1

 50  53  10  10  10  10

 6  2  1

 47   7  10  10   6

3  5  3 23  2  2  4

2

1 1

 2

St. Eustatius Battery St. Louis Duinkerk Honen Dalim Pleasures Estate Simon Doncker House Warehouse 1 Warehouse 2 Total

 1   1  4   4  1   3

 1  5  4

 5

1 4

2 5

1

1 2

 7  3  1

4

4

35

1 1 3 151 27

3

37

14

10

6

4

2

5

5

  1  16  18  21   1   1   3 330

Figure 4.2. Bivariate plot of thorium (Th) and cesium (Cs) showing all groups. Ellipses represent a 90 percent confidence level for membership in the group. (Todd M. Ahlman)

Trade of Afro-Caribbean Ware / 83

Group 3 includes 13 sherds from the Carr Estate on Montserrat and one from Nevis. Group 3 is generally similar to Groups 1 and 2 but has higher thorium concentrations and lower sodium concentrations. The Montserrat samples primarily have black angular inclusions, while the Group 1 and 2 samples generally have black, clear, and opaque inclusions. The remaining groups are composed of small sample sizes (generally fewer than 10 sherds). Group 4 contains 10 sherds from three sites on Statia and is characterized by low europium. Group 5 includes six sherds from three Statian sites and has high cerium levels. Group 6 includes the four sherds from the two warehouse sites on Statia. These samples are high in scandium, rubidium, and chromium. Group 7 includes two sherds from Brim­ stone Hill on St. Kitts that have the highest thorium levels of all the sherds in the study. Group 8 consists of five sherds from Pleasures Estate on Statia that have relatively higher levels of thorium and tantalum and are low in scandium. Group 9 consists of 5 sherds from St. Lucia. Group 10 includes a sherd from Brimstone Hill on St. Kitts and sherds from two Statia sites that are enriched in thorium, scandium, and tantalum. Group 12 includes one sherd from St. Kitts and sherds from two sites on Statia that are relatively high in tantalum and low in chromium. None of these groups are similar to any of the clay samples. The St. Croix sherds were not assigned to any of the groups. For the most part, the unassigned sherds (n = 35) were similar to the identified groups but when plotted fall outside the assigned bivariate plots (Figure 4.3). When compared statistically with any of the groups, but especially to Groups 1 and 2, they are not assigned to any group. Nine of the 17 sites have sherds from only one group. For the most part, this may be more a reflection that all of these sites have a sample size of fewer than 10 sherds than any actual differences in ceramic production and acquisition. Only one group was identified for the sherds from Antigua, Montserrat, and St. Lucia; however, these are small sample sizes and undoubtedly this is unrepresentative of the range of variation in paste recipes from these sites or on these islands. Five sites (Brimstone Hill and South­east Peninsula Site 1 on St. Kitts and Duinkerk, Honen Dalim, and Pleasures Estate on Statia) have sherds from three or more compositional groups. All five of these sites have sherds assigned to compositional Group 1 with vari­ous sherds assigned to seven other compositional groups. Groups 2 and 4 have the next highest frequency of analyzed sherds at these five sites. Three sites (Brimstone Hill, Duinkerk, Honen Dalim) have sherds from four different groups and Pleasures Estate on Statia has sherds assigned to six groups. On St. Kitts and Statia there appear to be numerous paste recipes in circulation among the vari­ous studied

Figure 4.3. Bivariate plot of thorium and cesium showing the relative similarity between the unassigned samples and the established compositional groups. Ellipses represent a 90 percent confidence level for membership in the different groups. (Todd M. Ahlman)

Trade of Afro-Caribbean Ware / 85

sites that indicate more than one place of pottery production and long-­ distance interisland Afro-­Caribbean ware exchange among enslaved and free Af­ri­cans.

Discussion In our study of Afro-­Caribbean ware we set out to identify clay sources and production centers, examine intra-­and interisland trade to start connecting the pathways within islands and between islands, and try to understand more about trade networks among free and enslaved Af­ri­cans. Our results show trends that shed light on the breadth of pottery production and trade in the historic Caribbean. These data show that across the Caribbean there were numerous paste recipes being used in pottery manufacture by enslaved and freed Af­ri­cans and that there was likely more than one pottery manufacturing center on each island. The internal markets on these islands varied greatly based on the regulations that controlled the ability of enslaved and freed Af­ri­cans to move about and sell or trade their wares, but in general it seems that people had access to markets that offered pottery from more than one source whether the ultimate source was local or nonlocal. Group 1 is the best example for identifying possible clay sources and production centers. The INAA results indicate that Group 1 was most likely made with Nevis clays and quite possibly St. Kitts clays. The similar vol­ canic geology of the two islands suggests that the clays on both islands may have similar chemical composition. One clay sample from St. Kitts plots within Group 2, but there are enough isotopic differences between it and the larger group that statistically it does not group with the sherds. Group 3 is primarily found on Montserrat and was probably made with Montserratian clay. Group 3 is also the only group other than Group 1 and its subgroups found on Nevis. The Groups 1 and 2 sherds show that there likely is greater intraisland variation in the chemical composition of clay sources than between island variation. The chemical composition of the sherds may be complicated by clay mixing from within island sources or between islands, but it is apparent that Groups 1 and 2 were likely made with Nevis and/or St. Kitts clays. There is strong evidence for interisland trade of Afro-­Caribbean ware by enslaved and free Af­ri­cans in the INAA results between St. Kitts, Nevis, Statia, and Antigua based on the presence of Group 1 sherds on all four islands. Group 1 includes the most sherds in the analy­sis and Group 1 sherds are found at 15 sites on Antigua (n = 2), Nevis (n = 6), St. Kitts (n = 4), and Statia (n = 3). Sherds from the other groups occur on no more than two islands with five of the 11 groups occurring on only one island. St. Kitts and

86 / Ahlman, et al.

Statia share the most groups with four groups, 1, 2, 10, and 12, appearing on both islands. Groups 4, 5, and 8 only occur on Statia, but do not have similarities to either of the clay samples from Statia, so the likely clay source is unknown. Group 5 only occurs within the St. Lucia sherds and does not share similarities with the one St. Lucian clay source in the study, but sourcing for the St. Lucian pottery is hampered by the small sample of sherds and clays. Group 3 was likely made with Montserratian clay, and one sherd from Nevis is assigned to Group 3. None of the Antigua or Statia clay samples grouped with pottery from those islands, but we are hampered by small sample sizes of sherds and clays from Antigua and small clay sample size from Statia (see Tables 4.2 and 4.3). St. Kitts and Nevis have a long history of being trading partners because of the his­tori­cal ties between the two islands, and as Ahlman and colleagues (2009) note, there is a long history of Nevisians selling pottery on St. Kitts well into the twentieth century. Heath (1988:107–109) found his­ tori­cal and ethnographic evidence for Nevisian pottery being sent to neighboring (St. Kitts, Anguilla, St. Maarten, St. Eustatius) and distant (St. Croix, St. Thomas, Domi­nica) islands during the mid-­to late nineteenth century and well into the mid-­twentieth century. Merrill (1958) discusses the mid-­ twentieth-­century pottery industry on Nevis and notes that Nevis pottery is found on St. Kitts and Nevis, and Crist (1949) has a photo of Nevisian women selling pottery on the St. Kitts beach. In the mid-­twentieth century, records also point to Antiguan potters exporting pottery to St. Kitts, Domi­ nica, Montserrat, and Nevis (Heath 1988:123). Heath (1988) could find no evidence of pottery exportation from St. Lucia. While external trade was an important component of pottery trade, Heath’s his­tori­cal and ethnographic evidence points to internal markets being the most important market for Caribbean potters. The selling of other hard commodities, such as housewares, goods, and liquor, was of­ten regulated by island governments; however, the exchange of produce and locally manufactured goods (e.g., pottery) was allowed. His­tori­cal sources cited by Heath (1988) indicate that pottery was a primary exported commodity from Antigua and Nevis in the nineteenth and twentieth centuries. Pottery may also have played a role as a vessel to hold and transport other commodities like fruits, vegetables, grains, and cooked foods that were sold in vari­ous markets; however, most period paintings and twentieth-­century photos do not show people using pottery as vessels to market or transport other products. Handler (1963) indicates that during the mid-­twentieth century people only took pottery to markets on Barbados. Our study finds that pottery made with Nevis clays occurs on multiple islands, suggesting that Nevis was an important production center. On Nevis we see that the majority of the pottery in the study was likely made with

Trade of Afro-Caribbean Ware / 87

Nevis clay, and only one sherd from Nevis is tied to another group and that is Group 3, which is the Montserratian group. This does not mean that there were not multiple pottery makers on Nevis, only that pottery from other islands was not widespread in the Nevis market(s). In fact, it appears that there was a robust pottery-­making sys­tem on Nevis where there were many pottery makers, and the Group 1 subgroups likely represent these different pottery-­making centers. Conversely, it appears on Statia that there may not have been a robust local pottery-­making industry, but that pottery from across the north­ern Caribbean entered the Statia port, like many other goods, and reached the different households on the island. This likely demonstrates the effects of being close to a port and the diversity of goods available to the people near the port. We can further look into this effect on St. Kitts. Brimstone Hill, where many people came and went, is near the port town of Sandy Point. There is a greater diversity of paste recipes at that site than at the southeast peninsula sites that were relatively isolated. As the study of Afro-­Caribbean ware expands, it is likely that we will see differences in access to pottery from other islands based on the relatively closeness to island ports. The study here shows that not all enslaved and freed Af­ri­cans in the Caribbean had or strove to have access to different pottery sources. We see on Nevis a virtual lack of pottery from other islands at the sites we examined, which means a robust internal pottery-­making economy but may also mean that a network of trade existed where pottery left the island and other goods, but not pottery, returned. The robustness of the internal market may have meant that potters from other islands did not see a viable market on Nevis, where they could have sold their wares effectively. On St. Kitts we see differences in market access. On the southeast peninsula it appears that people made pottery with local clays or traded exclusively with Nevisian potters. At Brimstone Hill, which is near a port, there are many different paste ­recipes, and some are likely from other islands. It is interesting, however, that at the Earl Romney slave village, which is at the base of Brimstone Hill, only two groups (1 and 2) are present at the site. These are likely local paste recipes and show that not everyone was tied into the interisland market, even those close to ports. What can we learn about past behavior when examining the results of our study? First, the production and trade of Afro-­Caribbean ware by enslaved and freed Af­ri­cans was extensive and widespread in the eighteenth and early nineteenth centuries. Although there is strong evidence for interisland trade in the Lesser Antilles, at least in the north­ern Lesser Antilles, it appears intraisland trade was more prominent than interisland trade. The evidence shows that on St. Kitts and Nevis internal markets dominated the

88 / Ahlman, et al.

trade networks, and on Statia markets included pottery from a multitude of other islands. While the data are not extensive, it seems that for sites near markets and ports—where interisland trade of a wide variety of goods occurred regularly—we have evidence for interisland pottery trade, as seen in the Statia and Brimstone Hill data. We can also surmise that there may have been potters who made and supplied wares at many of the plantations so that acquisition of pottery from other islands was unnecessary. At locations like Brimstone Hill, a military fortress, or the urban sites in Statia the production of pottery was probably not feasible, and slaves and freedman acquired pottery through trade. Why pottery from a single island, like Nevis, was selected over other sources is unknown at this time. Afro-­Caribbean ware from the islands in the study is virtually indistinguishable, so it does not appear that aesthetics played a specific role in consumer choices in pottery. There may be other products sold with the pottery that drove sales from specific islands; however, we have no proof that pottery makers and traders packaged and sold other commodities with the pottery. The driving force may have had more to do with the ability of the pottery makers from certain islands, Nevis in particular, or plantations to get their wares to market rather than a consumer choice. Pottery makers may have held special privileged positions on plantations and were allowed to engage in commodity trade that others were not allowed to undertake. For the makers and consumers of pottery, internal and external trade demonstrates an interconnectedness among people who sought to create social and economic networks that would minimize the risks inherent to enslaved and freed Af­ri­can populations in the Caribbean. Some purchasers may have sought out certain producers because of their social networks and elite positions in order to develop or build their own networks and social position. These social networks, however, were unequal either within or between islands. Planters and government officials saw local markets as destabilizing to the internal power structure and po­liti­cal economy of each island, and they sought to reduce movement by slaves and products sold in markets. Whether unequal access was a result of government and planter regulations or availability of pottery both internally or externally is not known at this time; however, we see two levels of exchange. First, on most islands it appears that the internal markets dominated pottery trade among enslaved and free Af­ri­cans in the Lesser Antilles. Internal markets likely included commodities from numerous producers providing consumers multiple options for products as well as for developing social networks. Second, inter­island trade was common but it appears to have occurred among producers, sellers, and occupants in urban areas and near ports. These producers and sellers held

Trade of Afro-Caribbean Ware / 89

special status in the plantations where they lived and worked, having abilities that others did not have to engage in long-­distance trade. Consumers may have chosen these sellers over local producers because of their social networks rather than their specific wares. Pottery production and trade are labor intensive and risky. The reasons posited for engaging in pottery production and trade range from making cash to acquire costly European goods (Galle 2011) to creating networks for risk reduction (Ahlman et al. 2009). The reason probably rests somewhere at the conjunction of these two concepts. For some potters, the former was likely the impetus for manufacture and trade as they meant to signal their relative wealth or other advantages. The latter was important for potters and consumers who lived in marginal areas, like St. Kitts’s southeast peninsula, where pottery trade may have helped them acquire necessary goods, like food, that may have been seasonally scarce. For most, however, pottery production was a way to acquire more expensive goods that were not provided by planters and to create support networks to address shortfalls in times of need.

5 A Danish Colonial Merchant’s Residence in Charlotte Amalie, St. Thomas Material Colonialism and the Intersection of Local and Global Trade at the Bankhus Christian Williamson and Douglas V. Armstrong The development of Charlotte Amalie and the island of St. Thomas in the Danish West Indies is directly connected to the growing importance of its harbor and the commodities, capital, and people that pass through it. The free haven history of St. Thomas has long been correlated to its prosperity gained by providing all ships a neutral, concealed, and prime trading location at the juncture of several intercontinental trading routes. Like other European colonies in the Caribbean that lacked military might (particularly those of the Dutch and Swedish), the Danish had to establish niche markets that could attract merchants, traders, and settlers from other Caribbean islands and from those European nations lacking their own colonies (primarily Germany and the Mediterranean countries) to their island through a free port status and maintaining neutrality in the vari­ous European conflicts that spread to their Caribbean colonies (Sonesson 2004). The Danish lured settlers from through­out the Caribbean and Europe in­clud­ing Dutch, English, French, and Jewish, as planters, traders, merchants, sailors, and craftsmen (Moolenaar 2005). Through the eighteenth and nineteenth centuries, St. Thomas became a center for Caribbean and Central Ameri­ can trade with Europe and the United States, especially through communication links to Europe. According to Gøbel (1994:158), Charlotte Amalie was at the center of a vast web of steamship lines, with four smaller ships in port ready to travel to destinations through­out the Caribbean and Central America. Archaeology at a Danish colonial merchant’s residence in Charlotte Amalie, St. Thomas, illustrates the complex yet distinct array of consumer goods available in a nineteenth-­century Danish Caribbean port town. A walled compound known as the Bankhus housed a series of nineteenth-­and early twentieth-­century merchants/bankers and their household servants. This study explores the intersection of micro-­and macrohistory as it assesses the

A Danish Colonial Merchant’s Residence / 91

material and documentary record of the site. The abundant material record derives from a pattern of storm-­related damage and the means to rebuild and replenish the buildings. The study of this complex walled and terraced house compound illustrates a combination of broad-­based global trade, specific Danish trade patterns and interactions, and the incorporation of local goods and trade networks. The Bankhus is the sec­ond house compound on Government Hill in the Kongens Quarter that has been investigated by researchers from Syracuse University’s Maxwell School. Findings from the adjacent Magens site are available in several publications (A. Armstrong 2011; D. Armstrong and Williamson 2011; D. Armstrong, Williamson, and A. Armstrong 2009, 2013; Williamson and D. Armstrong 2009, 2011). Exploration of the Bankhus pro­vides the basis for comparisons of walled urban house compounds and the role of Danish colonialism in the Caribbean (D. Armstrong et al. 2013; D. Arm­strong et al. 2015). In addition, this research adds to the growing ar­chaeo­logi­cal literature on historic port towns in the Caribbean like Port Royal, Jamaica (Hamilton 2006), and Bridgetown, Barbados (Smith and Wat­ son 2009; Welch 2003) as well as parallel temporal contexts like the ur­ban housing studies in Charleston, South Carolina (Herman 1999; Zier­den 1999, 2005; Zierden and Calhoun 1986). The walled Bankhus compound encompasses a complex that includes a main residential house, occupied by the families of a series of port town merchants, with outbuildings, occupied by household servants and their fami­ lies. A combination of census and tax records allows us to document many of the people whose lives intersect with the material record deposited on the property. This chapter draws on the economic and demographic details abstracted from the his­tori­cal record in comparison with the material rec­ ord recovered from well-­defined and spatially segregated midden areas on the property. Artifacts recovered reflect international trade and access to a wide range of goods of all cost ranges and from many regions. Most of these goods can be linked to the property owners who were merchants engaged in international trade through St. Thomas’s bustling port. Some goods, such as Royal Copenhagen fluted lace pattern porcelain, have direct ties to the Danish presence; others, like numerous overglaze porcelains, show affinity with unique Danish trading partners in the Far East. Yet the site also contains evidence of specific patterns of regional and interisland trade and social interactions. Moravian wares provide material testimony to the presence and influence of the Moravian mission in the Danish West Indies. Bone buttons, produced in smaller numbers than in the neighboring Magens site, are represented by

92 / Williamson and Armstrong

single-­hole bone button blanks and likely point to a cottage industry, as has been documented elsewhere in the Caribbean (D. Armstrong and Williamson 2011; Klippel and Schroedl 1999).

The Bankhus’s Port Town Setting: An Intersection of Local and Global By the beginning of the nineteenth century Charlotte Amalie had become a crowded port with a dense row of waterfront warehouses, backed by steep hillsides filled with multistory residences (Westergaard 1917:31; Zabriskie 1918; Gjessing and Maclean 1987; Dookhan 1994:37; Highfield 1997). Merchants, like the owners of the Bankhus, were involved in trading networks on several scales, between islands, across the Atlantic and beyond, with St. Thomas’s warehouses serving as transshipment centers for incoming and outgoing cargoes to other islands in the Caribbean and to centers of commerce and trade in Europe, the Americas, Africa, and the Far East (Gøbel 1997). Danish colonial neutrality and a laissez-­faire approach, so long as port taxes were paid, provided a basis for continuity in trade and a venue for commerce that circumvented the restrictive covenants of mercantilism (Gjessing and Maclean 1987; Gøbel 1990; Moolenaar 2005; Ragster 1986; Sonesson 2004; Green-­Pedersen 1971; Taylor 1888). The significant growth of St. Thomas in the late eighteenth and early nineteenth centuries can be tied to the sacking of its chief economic rival St. Eustatius by the British in 1781 and its ideal location in the network of trade routes from Europe to the Caribbean and South America (Larsen 1954; Tyson 1986, 1991; Gøbel 1990; Hurst 1996; Sonesson 2004). The Bankhus property is situated in an important geographical position close to the port and the merchant warehouses of the harbor while still being uphill enough to enjoy the favorable, cooling trade winds. Socially, po­ liti­cally, and economically, the Bankhus occupies an important position in Danish colonial influence on St. Thomas. Directly to the west is the Magens compound, home first to a prominent Danish colonial official and his family and then to a Scottish merchant whose firm traded around the globe. Along the north­ern border is the Crown House, former home of Governor General Peter von Scholten and residence of the St. Thomas harbormaster. Directly to the east is the Lutheran parsonage with the Lutheran church itself across the street to the south. The building directly east of the Lutheran parsonage on Kongens Gade is the Governor’s House. It is no surprise then that the Bankhus was owned by a series of merchants who no doubt enjoyed the property’s unparalleled location to the power structure of St. Thomas,

A Danish Colonial Merchant’s Residence / 93

whether it was the Danish colonial government, the Danish national religion, or some of the other most prominent merchants and traders on the island. The Bankhus compound is divided into 11 walled terraces that demarcate the current grounds. These terraces allowed for the leveling of yard space and for access between structures and outbuildings. From an ar­chaeo­logi­ cal standpoint, the terraces served as catchment areas in which the material residue of daily life and events affecting the households were captured. The stratigraphic deposits recovered from the Bankhus reflect both gradual accumulation of materials from daily life and the punctuated and traumatic impacts of a series of damaging hurricanes and tropical storms that swept through the island in the nineteenth century. The combination of a thriving port town, the means to acquire and replace goods, and the recurrence of damaging storms resulted in dense cultural deposits filled with household items. After each succeeding storm, damaged household goods were used to fill what became a series of leveled walled gardens, after which the main house and outbuildings were refurbished and reoutfitted. As recently as 1995, Hurricane Marilyn did major damage to the front porch and facade of the house. However, we know from photos of the cleanup and rebuilding process that the materials from that traumatic event were removed from the site and deposited in the island’s landfill. This spatial and material study explores the intersection of micro-­and macrohistory as it assesses the material and documentary record of the site. The fact that Bankhus was recognized as part of the Danish West Indian cultural identity is reflected by the fact that, when the Danish government began the process of divorcing itself from direct colonial imperialism in the early twentieth century, the government produced a photographic documentary of the house and its furnishings (D. Armstrong et al. 2015). Photographs of the house provide a unique, commemorative view of Danish colonialism that we are able to examine and critique along with the ar­chaeo­logi­cal record (D. Armstrong et al. 2015).

The Bankhus: Demography and Social Relations The array of people who lived at the Bankhus reflects a cosmopolitan ensemble of people with diverse points of origin and backgrounds. Denmark’s open colonial structures encouraged immigration from both within and outside the region (Sonesson 2004; Moolenaar 2005). The property owners and occupants of the main house included merchants born in St. Thomas, Hamburg (Germany), and Bayonne (France) as well as a banker from Denmark. Most acquired the property when they were in their early thirties and sold

94 / Williamson and Armstrong

the property within a decade (Table 5.1). They were white, black, and probably of mixed racial backgrounds, although that option is not provided in the census records. Their wives were from St. Thomas with the exception of a black Jewish Frenchman whose wife was from St. Domingue (Haiti). Most of their children were born in St. Thomas. The owners were identified as being Jewish, Toraelite (defined as Black Hebrews), and Lutheran, with the religious affiliations of their wives not always matching their own. Among the wives, we find all of the former religious practices along with an affiliation with the Dutch Reformed Church. The household of the owners and occupants included not only their children but also a brother and a mother-­in-­law. Most servants were from other islands, and their places of birth included St. Thomas, St. John, St. Eustatius, St. Croix, St. Domingue, Tortola, Saba, and St. Martin. Through the nineteenth century the support staff in­cluded servants as well as a nurse, shoemaker, seamstress, chambermaid, and a cook. The support staff practiced an equally diverse array of faiths in­clud­ ing Protestant, Lutheran, Dutch Reform, Episcopal, Methodist, and Roman Catho­lic. One enslaved in­di­vidual is associated with this property. In 1846, a 15-­year-­old “house-­boy” named Chesterfield is listed for the property.1 Through the first half of the nineteenth century slavery continued to play a dominant and direct role in agricultural production in the Danish West Indies. However, on St. Thomas, whose population was dominated by the burgeoning port town, the majority of people of Af­ri­can descent were free by 1803 (Green-­Pedersen 1971; Harrigan and Varlack 1977; Knight and Prime 1999). The constant flow of ships and people as well as the need for skilled labor made Charlotte Amalie a magnet for free blacks and those seeking to escape slavery. The maritime setting of the free port provided an opportunity to leave the Danish West Indies by signing aboard a departing vessel (Hall 1985, 1992, 1997a, 1997b). However, even in this setting, slavery was present until emancipation in the Danish West Indies in 1848. The property owners reflect diverse European points of origin, while the laborers were more likely to have been “down-­islanders” than to have been from St. Thomas. The demography of the Bankhus is consistent with the town as a whole, which has been defined as a “virtual polyglot” of nationalities, in­clud­ing settlers from through­out the Caribbean and Europe.

Bankhus Archaeology: Spatial and Material Analysis The social complexity present at the Bankhus is similar to that reported for the neighboring Magens property (Armstrong et al. 2013). However, unlike the Magens House, Bankhus had no doors or gates linking it to other neigh-

Table 5.1. Census Records for 24, 25 Kongens Gade and Quarter, 1846–1911 Year

Name

1846 1846 1846 1846 1846 1846 1846 1846 1846 1846 1846 1846 1850 1850 1850 1850 1850 1855 1855 1855 1855 1855

Isaac Pretto Rebecca Pretto David Pretto Dorothea Jacobs Elizabeth Newton Elizabeth Petersen Sarah Johnson Susannah Caron Dr Pretto Deborah Pretto Rebecca Pretto Appleonia Isaac Pretto Rebecca Pretto David Pretto Hans Rasmussen Sarah Monsanto Carl Geo Heise Amory Heise A. E. Reed George Amory Heise Anna Louise Heise

Continued on the next page

Own/Occ

Sex

Birth Place

Age

Occupation

Religion

Both Occ Occ Occ Occ Occ Occ Occ Occ Occ Occ Occ Both Occ Occ Occ Occ Both Occ Occ Occ Occ

M F M F F F F F M F F F M F M M F M F F M F

St. Thomas St. Thomas St. Thomas St. John St. Eustatius St. Croix St. Thomas St. Thomas St. Thomas St. Thomas St. Thomas St. Thomas St. Thomas St. Thomas St. Thomas Toulstrop, Denmark Saba Hamburg St. Thomas St. Croix St. Thomas St. Thomas

33 26 4 24 40 55 22 24 40 40 12 39 37 30 8 30 45 38 21 40 4 2

Merchant

Servant Merchant

“Jew” “Jew” “Jew” Protestant Protestant Protestant Protestant Protestant “Jew” “Jew” “Jew” Lutheran “Graclete”?

Y Y — N N N N N Y Y — N Y Y

Shoemaker Servant Merchant His wife Her mother Their son Their daughter

Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran

N N Y Y Widow

Servant Servant Servant Servant Servant Doctor

Marry

Table 5.1. Continued Year

Name

1855 1855 1855 1855 1855 1855 1855 1855 1857 1857 1857 1857 1857 1857 1857 1857 1857 1857 1857 1857 1860 1860 1860 1860

Frederick Fischer Sarah Sampson Hannah Ragiers Eliza Frernifs Judith Prytz Anthony Benjamin Carmelin Philipps Christian Friederich Carl Geo Heise Amory Heise George Amory Heise Anna Louise Heise Caroline Henrietta Heise Ann Eliza Reed Friederick Fischer Eliza Warner Sarah Simpson Emilia Smith Antony Benjamin Christine Freicines(?) Carl Geo Heise Amory Heise George Amory Heise Anna Louise Heise

Own/Occ

Sex

Birth Place

Age

Occupation

Religion

Occ Occ Occ Occ Occ Occ Occ Occ Both Occ Occ Occ Occ Occ Occ Occ Occ Occ Occ Occ Both Occ Occ Occ

M F F F F M F F M F M F F F F F F F M F M F M F

Detmara St. Croix St. Croix St. Croix St. Thomas St. Thomas St. Thomas St. Thomas Hamburg St. Thomas St. Thomas St. Thomas St Thomas St. Croix Delmaed St. Thomas St. Croix Tortola St. Thomas St. Thomas Hamburg St. Thomas St. Thomas St. Thomas

21 60 36 58 20 12 10 12 39 23 6 4 2 43 23 60 66 22 14 14 43 27 10 8

Servant Servant Servant Servant Seamstress Servant Servant Servant Merchant His wife Their son Daughter Daughter Mother-in-law Servant Servant Servant Servant Servant Servant Merchant His wife Their son Daughter

Lutheran Lutheran Lutheran Dutch Ref Lutheran Moravian Lutheran Moravian Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Dut Reform Lutheran Episcopal Moravian Moravian Lutheran Lutheran Lutheran Lutheran

Marry N N N N N N N N Y Y N N N Widow N N N N N N Y Y

1860 1860 1860 1860 1860 1860 1870 1870 1870 1870 1870 1870 1870 1880 1880 1880 1880 1880 1880 1880 1880 1901 1901 1901 1901

Caroline Henrietta Heise Francis Julius Heise Anny McDonald Eliza Warner Margaret Williams Friederick Fischer Jens Cecitinsethrgust Hingelberg Caroline Rogiers Hingelberg Amory Helene Hingelberg George Robert Hingelberg Cecil Edward Hingelberg Sophia Simpson Mathilde Alick Edward Willink,Esq. Felise Nunes Alice Nunes Albert Nunes Emile Nunes Clemence Nunes Felicia Grant Therese Hachet John B. M. Monsanto Mrs. M. Monsanto Herman M Monsanto Desire M Monsanto

Continued on the next page

Occ Occ Occ Occ Occ Occ Both Occ Occ

Own Occ Occ Occ Occ Occ Occ Occ Both Occ Occ Occ

F M F F F M M F F M M F F

St. Thomas St. Thomas St. Croix St. Croix St. John Detmara Copenhagen St. Thomas St. Thomas St. Thomas St. Thomas St. Thomas St. Thomas

5 2 32 70 65 26 48 39 13 9 4 74 28

Daughter Son Servant Servant Servant Servant Merchant His wife Children Children Children Nurse Chambermaid

Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran Lutheran R. Catholic

M F M M F F F M F M M

Bayonne St. Dominique St. Thomas St. Thomas St. Thomas St. Croix St. Croix St. Thomas St. Thomas St. Thomas St. Thomas

43 27 5 4 2 26 19 65 31 32 9

Merchant Wife Daughter Daughter Son Servant Servant Broker Wife Son Son

Toraelite Toraelite Toraelite Toraelite Toraelite Episcopal Episcopal Lutheran R. D. Church R. D. Church R. D. Church

N N N N Y Y N N N N N Y Y

N N Y Y Widow

Table 5.1. Continued Year

Name

1901 1901 1901 1911 1911 1911 1911 1911

Jenny M Monsanto “Girl” Mrs. Vilmus National Bank for the DWI Johee Hulsi Ida Elsa Hulsi Stelge Hulsi Jenn’ Emily James

Own/Occ Occ Occ Occ Own Occ Occ Occ Occ

Sex

Birth Place

Age

Occupation

Religion

Marry

F F F

St. Thomas St. Thomas St, Martin

4 1aar 42

Daughter Daughter Nurse

R. D. Church Methodist

N

M F M F

Denmark St. Croix St. Thomas St. Croix

31 23 1 20

Sub Director Wife Son Cook

Lutheran Lutheran Lutheran Moravian

Y Y

N Sources: Vestindiske Reviderde Regnskaber Matrikler for St. Thomas, St. John, and St. Croix (1835–1915), copy on file at the Caribbean Genealogy Library, St. Thomas, USVI, on microfilm; Virgin Islands Census Records for Charlotte Amalie, St. Thomas, copy on file at the Caribbean Genealogy Library, St. Thomas, USVI, on microfilm; Virgin Islands Title and Deed Books, Recorder of Deeds), Lt. Governor’s Office, St. Thomas, USVI.

A Danish Colonial Merchant’s Residence / 99

boring properties. The lack of connections at the Bankhus property can be attributed to the fact that, unlike the Magens family, the owners and residents of the Bankhus shared no familial ties with those living in the adjacent properties, providing no impetus for free movement between lots. The Bankhus had one formal “Welcoming Arms” entry and a service gate, accessible only from the main street of Kongens Gade. The property at 24, 25 Kongens Gade was not formally recorded until it was separated from 23 Kongens Gade (the Lutheran parsonage) and sold by the Guardians of the Lutheran Church to merchant Creighton Whitmore on July 22, 1843 (Recorder of Deeds Office [ROD], Book GG 110:4; ROD, Book HH 10–11:8). The 1844 matrikal (tax record) showed a decrease in building area slightly from 503 Alen in 1843 to 456 Alen in 1844, which might have represented the initial phases of Whitmore’s remodeling of the property (Caribbean Genealogy Library [CGL], St. Thomas Matrikal, 1843 and 1844). Two men, three women, and one girl were living on the property in 1844 (CGL, St. Thomas Matrikal, 1844). The port town of Charlotte Amalie by this time had seen considerable damage by fire and storm and passed laws that had an impact on the design of the new construction. Not only was the Bankhus built primarily of stone, but it was also divided from neighboring structures by a drainage and passageway, designed to provide access between buildings and to serve as a fire break. In addition, each structure was mandated to have fire buckets and fire rods (to pull down burning structures). Two origi­nal fire buckets bearing the name C. Whitmore and the property address have remained in the main house, having been passed on through a series of owners. A map by Christian Hingelberg (dated March 1837) shows three buildings on the 24, 25 Kongens Gade parcel (Figure 5.1). This map predates a massive hurricane that struck the Danish West Indies on August 2, 1837, and likely significantly affected this property as well as most of those around the harbor and on Government Hill. An 1897 insurance map by Goad shows essentially the same building layout found on the site today, minus a few outbuildings (Figure 5.2). A site map shows the layout of the site and the spatial divisions used in our analy­sis (Figure 5.3). The divisions are as follows: Zone A (Terraces 4–5): The garden terraces located beside the main houses gallery. Zone B (Terraces 6–10): Upper terraces associated with outbuildings and gardens. Zone C (Terrace 11): A paved terrace located on the east side of the main house. The stratigraphic detail and quantity of artifacts found in Zone A provide details on site formation and social processes at the site (Figure 5.4). Unit 16 (and other units in Zone A) yielded an abundance of material remains and provide material evidence of a series of traumatic events. At the base of Unit 16 (depth 200–220 cm) are a path and steps that corre-

Figure 5.1. Map of 23, 24, 25 Kongens Gade, 1837, by Christian Peter Hingelberg.

Figure 5.2. Insurance map of 24, 25 Kongens Gade, 1897, Charles E. Goad, Ltd.

Figure 5.3. Map of Bankhus illustrating zones, terraces, and units. (Douglas V. Armstrong)

102 / Williamson and Armstrong

Figure 5.4. Ceramic column, Unit 16 (left), and stratigraphic profile for Units 16 and 24 (Zone 1, Terrace 4) (right). (Douglas V. Armstrong)

late with late eighteenth-­and early nineteenth-­century use of the property. Above these deposits are two stratified layers of rubble suggesting one or more hurricanes or severe storms. Several fires caused severe damage to the city in the decades of the later eighteenth to the early nineteenth century, but none of them are reported to have affected the upslope side of Kongens Gade (Svensson 1964). The ceramics recovered from Unit 16 show that the bulk of the materials recovered date to the first half of the nineteenth century with massive deposits in the lower half of the profile followed by gradual accumulations of materials. The mean ceramic dates (MCD) for the lower levels are earlier than the documented construction and destruction sequence. The MCDs range from 1806 to 1824 for deposits that are probably associated with the 1836–1844 hurricanes, destruction, and construction events. Many of the

A Danish Colonial Merchant’s Residence / 103

household goods were damaged by the storms and had to be sequentially replaced. In this context the owners had the means to acquire replacement wares. The 2-­m-­deep deposits in Zone A and all levels below the yellow brick pavers that capped units in Zone C are from an earlier occupation associated with the Lutheran parsonage. Later deposits in Zone A and all data from Zone B correspond with merchant owners and their support staff. Ceramic decoration shows a gradual change over time but does not suggest any significant division between owner and laborers on the property. However, higher proportions of chamberware are found in the upper garden terraces. Glassware shows a more significant distribution, with large quantities of relatively expensive stemware found primarily in terraces associated with the main house. Small finds from the main house middens include brass furniture hardware and shell inlay used in fine furnishings along with chandelier crystals. These middens also show greater affluence. Most of these items came from dense layers associated with hurricane damage to the main house. The combined data suggest dramatic damage to the property in the mid-­ 1830s with a reconstruction effort in 1836, probably corresponding to the buildings shown on the 1837 Hingelberg map. These structures and their material contents were likely damaged by the August 2, 1837, hurricane. Rebuilding apparently did not occur until the property was acquired by Whitmore in 1843. This construction is commemorated with a stone marker bear­ ing the date “1844” that remains on the building’s facade. The stratigraphic profile of Unit 16 supports this sequence of events where a layer of buried blue walkway stones reflect a series of garden paths, meaning that very little soil was added to the surfaces of the formal gardens during the late nineteenth and twentieth centuries. A shift from domestic to mixed commercial and domestic use of the property occurred with the purchase of the house and the opening of a Danish West Indian branch office by the Danish National Bank in 1904 (Hoist 2009). This action coincided with the creation of a Danish West Indian currency to replace Spanish gold coins (Danmark National Bank 2011; Hoist 2009). The Danish West Indian office of the Danish National Bank remained a primary banking institution on the island until it and its properties were transferred to the Virgin Islands National Bank in May 1935. Ownership of the property has since passed through several individuals, in­clud­ing the law firm of Grunert and Stout, before being purchased by its current owner and incorporated into a historic homes tour that includes Blackbeard’s Castle and Government Hill. On Janu­ary 1, 2013, a new museum and map shop dedicated to Camille Pissarro, the globally influential painter who grew up on St. Thomas, was opened within the Bankhus property.

104 / Williamson and Armstrong

Summary The Bankhus compound provides a wide range of information related to colonialism in the Danish West Indies during the nineteenth and early twentieth centuries. The Bankhus was built and occupied by Danish colonial merchants then sold to the Danish National Bank, which owned the house at the time of transfer. The materials present reflect both the broader mercantile systems of the nineteenth century and the specific permutations of trade and interaction on St. Thomas. We conclude with five observations based on the data. First, the material record reflects the relative affluence of the owners but lacks a strong material footprint for servants and support staff. This contrasts with the adjacent Magens House compound, where greater separation covaried with a distinct pattern of lower-­cost items associated with laborers, in­clud­ing servants and house staff and clerks employed in the property owner’s warehouse businesses (D. Armstrong and Williamson 2011). The spatial differences arise from the size of the lots and also from how the lots were used by their residents. The Magens House compound contained four parcels with distinct residential areas separated by gates and terraces as well as a zone on the upper terrace used for small livestock and produce. Merchants, clerks, and servants occupied different spaces within the compound, resulting in a greater variety of material goods distributed across the site. In contrast, census records show that the Bankhus was primarily occupied by more established merchants, with fewer servants and renters represented on the property. The ceramic decoration styles from the Bankhus indicate a higher affluence but overall less diversity in wares when compared with the adjacent Magens House. Second, the primary role of trade and colonial commerce is clearly defined in the material record. The range of materials present reflects access to a wide range of goods from global markets in the Far East and Europe, from which a wide range of refined ceramics, glassware, and furniture (hardware) were observed. This diversity, which is not limited to only Danish colonial trade and shipping routes, speaks to the importance of St. Thomas as a transshipment hub through­out the eighteenth and nineteenth centuries. Regional trade is also indicated but primarily in the form of course earthenware. The presence of Moravian missionaries is seen in the ubiquitous distribution of Moravian pottery across the site. Moravian religious affiliation is recorded in the deeds for several servants (all from St. Thomas or St. Jan [John]), but at lower proportions than for people of Af­ri­can descent in the Danish West Indies as a whole. Interestingly although Moravian Church membership can be correlated to people of Af­ri­can descent on the

A Danish Colonial Merchant’s Residence / 105

island, use of goods produced by the Moravians suggests a generalized acceptance and use by the broader St. Thomian population. We still do not know whether these wares were made locally or in Herrnhut, Salem, North Carolina, or Bethlehem, Pennsylvania, but potters were among the Moravian missionaries on St. Thomas. Third, despite the diverse backgrounds of the site’s occupants and the open nature of the free port at St. Thomas, the material culture at the Bank­ hus is surprisingly uniform and representative of popu­lar styles of the time period. As merchants, the property owners had access to a wide range of goods in the warehouses of St. Thomas as the Danish colonial policy allowed ships and goods from all nations to find an open market in the Danish West Indies. It is surprising then that although the merchants living at the Bankhus had access to virtually any material goods they desired, the bulk of the wares reflects the broader trends of early nineteenth-­century trade. The theme of variety and diversity is consistent not only with access to goods but also in the population itself. Merchants and sailors from every nationality were welcome in the Danish West Indies where commerce was king. The household at the Bankhus was made up of people from different origins and diverse backgrounds, yet again we surprisingly see very little differentiation stratigraphically in material use during periods in which the owner was from Denmark, France, or St. Thomas. While a distinct or separate material record was not found for the servant and support staff, social stratification is clearly indicated. Class was at the core of the Danish colonial system, and it is well defined in the differential size of dwelling structures and in the daily tasks of the residents. The inclusion of course earthenware and Moravian wares probably correlates to the presence of a support staff. The location of the kitchen within the main house rather than in one of the upper outbuildings means that the cook would have likely used their preferred wares for cooking, blurring spatial separation in a compound by depositing both merchant and staff material remains in Terraces 4 and 5 adjacent to the main residence. Fourth, Danish colonialism did not play by all of the generally assumed “rules” of colonialism and mercantilism. Danish colonialism centered on alternatives, repacking shipments, and resorting and redistributing goods. Business thrived on the Danish colonial middleman role as the Bankhus compound was situated in a free port with relatively unregulated trade. This is seen in the household composition of lots like the Magens and Bankhus and to some degree with the range of material goods and their spatial distribution. Finally, even as the Danish West Indies provided a neutral haven, encouraged immigrants from many places and faiths, and provided employment

106 / Williamson and Armstrong

to people escaping slavery, or at least looking for gainful employment in the warehouses and ships of trade, the Danish West Indies was entrenched in colonialism. The basic imperial and financial infrastructure and class-­based differentiation of the colony resulted in a sys­tem bound to social inequality. The underlying role of Danish colonialism is clearly defined through the commemorative photo documentary of the Bankhus (D. Armstrong et al. 2015). We note here the major themes of this colonial vignette while also acknowledging a future potential for a more in-­depth microhis­tori­cal analy­sis of the photographs. The exterior views project thick and imposing exterior walls separating the outsider from the house’s interior spaces. The walls provided the occupants a safe haven in a busy urban environment. Moreover, the interior photographs reveal a completely different aesthetic. The rooms were filled with fine furnishings and antiques in homage to colonial life, and walls were filled with symbolic representations of power and authority. The display of colonial wealth and authority represented in the interior spaces far exceeds expectations. The selection of this household to represent the Danish colonial period requires careful consideration and tempers notions that might otherwise be generated relating to diversity and openness within the Danish colonial system. Using the Bankhus as a backdrop, we conclude with the observation that as acutely hypercolonial are the Bankhus interiors, as captured in the commemorative “self-­portraits,” colonialism did not stop in the Virgin Islands with the transfer to the United States. The initial military government the United States imposed had virtually no understanding of West Indian lifeways or comprehension of the how a house compound like the Bankhus could have functioned within such a setting of cultural diversity. An examination of the impact of Ameri­can imperialism on the former Danish West Indies both ar­chaeo­logi­cally and his­tori­cally with house compounds like Bank­hus would be an excellent direction for future research on the island.

Note 1. In the 1846 Census of Unfree Living in 24 and 25 on King Street belonging to D. Pretto, there was Chesterfield, a houseboy, 15 years of age, who was unmarried and of “indifferent” moral character. Data from Vestindiske Reviderde Regnskaber Matrikler for St. Thomas, St. John, and St. Croix (1835–1915), and Virgin Islands Census Records for Charlotte Amalie, St. Thomas. Copies on file at the Caribbean Genealogy Library, St. Thomas, United States Virgin Islands, on microfilm.

6 The Investigation of Daily Practice of Enslaved Laborer and Sharecropper Households on an Eighteenth-­to Nineteenth-­Century French-­Caribbean Plantation Diane Wallman and Kenneth G. Kelly Investigations of slavery in the Caribbean reveal that despite differences in colonial regime, economic development, and geographical variation, the lives of enslaved laborers on many islands shared general structural similari­ ties. Yet, enslaved laborers on each island and each plantation developed unique and particular adaptive strategies to navigate the structures of power and control fundamental to the plantation system. This chapter pre­sents the results of ar­chaeo­logi­cal excavations and his­tori­cal research conducted on Habitation Crève Cœur, an eighteenth-­and nineteenth-­century sugar plantation in Mar­ti­nique. Specifically, we integrate his­tori­cal research with ceramic and faunal data recovered from the platforms of five houses inhabited by enslaved laborers and later free sharecroppers who lived and worked on the plantation. With these data, we evaluate the strategies the community developed to negotiate the severe constraints of their social and natural landscape. In the Caribbean, ar­chaeo­logi­cal investigations of slavery have focused primarily on Anglophone islands, in­clud­ing Barbados, the Bahamas, Nevis, St. Kitts, Montserrat, Antigua, and Jamaica (e.g., Ahlman et al. 2009; Armstrong 1990, 2003; Armstrong and Fleischman 2002; Armstrong and Kelly 2000; Bardoe, chapter 7 herein; Bates et al. 2016; Delle 2000; Handler and Lange 1978; Hauser 2008; Hauser and Armstrong 2012; Hauser and D ­ eCorse 2003; Farnsworth 1996, 1999, 2001; Hicks 2007; Higman 1998; Loftfield 2001; Mathewson 1972a, 1972b; Petersen et al. 1999; Pulsipher 1990; Pulsipher and Goodwin 1999; Reeves 2011; Wilkie 2000, Wilkie and Farnsworth 2005). Research has expanded to the Netherlands Antilles (Haviser, ed. 1999; Haviser and DeCorse 1991; Heath 1999) and the Danish West Indies (Armstrong 2003; Hauser and Armstrong 2012; Lenik 2009a), but studies of slavery in the French West Indies remain rare (Brunache 2011;

108 / Wallman and Kellly

Delpuech 2001; Gibson 2009; Kelly 2004, 2008, 2009, 2011; Kelly and Wall­man 2014; Lenik 2012b; Wallman 2014). The archaeology of slavery in the French Caribbean has only recently developed (Delpuech 2001; Kelly 2004, 2009). Delpuech (2001) and Kelly (2004) suggest that this trend was largely due to a lack of governmental oversight and infrastructure capable of managing cultural resources in Mar­ti­nique and Guadeloupe. In the past 20 years, however, with the establishment of governmental offices to supervise archaeology on the islands, plantation archaeology has expanded and now includes research on plantation sites in Mar­ti­nique, Guadeloupe, St. Martin, and French Guiana (Bain et al. 2011; Bonnissent et al. 2012, 2013; Brunache 2011; Croteau 2004; Garros et al. 2013; Gibson 2009; Lenik 2012b; Kelly 2004, 2008, 2009, 2011; Kelly and Wallman 2014; Tomadini et al. 2014; Wallman 2014). The excavation of Habitation Crève Cœur was the first major investigation of an enslaved laborer village on the island of Mar­ ti­nique (Kelly 2009). With few to no primary accounts from enslaved laborers themselves, scholars of Caribbean enslaved laborer life must examine European documentary sources for most ethnohistoric data. Unfortunately, his­tori­cal accounts of the sociocultural and domestic lives of enslaved laborers are limited and of­ten fragmentary. While European accounts of the po­liti­cal and economic aspects of plantation slavery are more common, there is a dearth of his­tori­ cal information on the everyday lives of enslaved individuals in the Caribbean. The sociocultural data contained in those sources that do exist reflect Eurocentric prejudices and biases and do not provide detailed explorations of enslaved laborer life (DeCorse and Chouin 2003:9; Handler and Lange 1978:1–4; Handler and Wallman 2014; Kelly 1997; Lightfoot 1995:201). This study integrates the documentary record associated with slavery in the French West Indies with localized material evidence of the consequences of and responses to plantation slavery. We first detail the his­tori­cal setting of the plantation and then present ar­chaeo­logi­cal data to evaluate any variation in daily practice among enslaved households within the estate, and we also examine social and economic organization over time, into the post­ eman­cipation (1848) period. By assessing the production and consumption of both local and imported pottery and foodstuffs, we deconstruct the local material expressions of larger social and economic systems.

Mar­ti­nique, Sugar, and Slavery Mar­ti­nique remained unsettled by Europeans during the sixteenth century until French colonial policy began to focus on the development of Caribbean colonies. During the early seventeenth century, the island was used

Enslaved Laborer and Sharecropper Households / 109

as a way station, with the Dutch establishing several small ephemeral agricultural settlements, as European traders visited Mar­ti­nique for tortoise shell, tobacco, hides, and provisions (wood and water) (Kimber 1988:107). Largely due to its position as a windward island, vessels frequently arrived first at Mar­ti­nique after their Atlantic crossing. This island became a major focus of French attention and soon emerged as the seat of administration for all the French Caribbean colonies (Munford 1991:2:360; Watts 1990:171). The French sugar industry grew steadily from 1664 to 1717, and by 1700, the French islands were producing 9,000 tons of sugar annually (Higman 2012:104; Stein 1988:9). Large sugar planters dominated the economy in Mar­ti­nique. In 1770, for example, 100 large, lucrative, self-­sufficient sugar plantations had 12,000 enslaved laborers (averaging 120 each), accounting for nearly 17% of the total 71,142 population on the island at this time (Stein 1988:44). The average sugar plantation in the eighteenth century encompassed 153 acres, whereas the average tobacco plantation on Mar­ti­ nique was about 48 acres (Munford 1991:2:417). Most whites, aside from the very poor, relied to some extent on enslaved laborers, and a hierarchical class sys­tem quickly developed in the French West Indies, with divisions between the grands blancs or békés (rich planters) and the petits blancs (poor whites) (Stein 1988:34). Overall, accounts suggest that a hierarchy within the sys­tem of enslaved labor on French plantations was in place by the eighteenth century. Enslaved laborers with the highest standings were generally men, and included the econome (overseer), commandeur (slave driver), and gereur (plantation manager). Specialist or artisan enslaved individuals (typically male), were also considered to hold a more elite status, as were some domestics, such as the housekeeper, cook, infirmary worker, seamstress, and nurse (Debien 1964:92; Moitt 2001:78). The field workers, les negres or les negresses, were at the bottom of this stratified system. In fact, mixed-­race men and women and specialists had higher estimated values than field laborers and those of solely Af­ri­can descent (Debien 1964:92). Mixed-­race creole children, whether free or enslaved, were a higher class, known as the gens de coleur, and were of­ten given more elite positions as artisans or domestics on plantations and were more likely to gain their free­dom (Robin 1807:41). After the first enslaved Af­ri­cans arrived on Martiniquan soil in the mid-­ seventeenth century, the number of enslaved individuals increased steadily in the eighteenth century, peaking at over 85,000 around 1830 (Moreau de Jonnès 1842:17). Throughout most of the plantation era, enslaved laborers outnumbered European settlers nearly eight to one. After emancipation, chattel slavery was eliminated, but equality was far from achieved. In 1848, a sys­tem of association established contracts between former enslaved laborers

110 / Wallman and Kellly

and their owners, allotting portions of lands to the now free individuals, use of cabins, weekends off, and a third of the sugar produced before expenses (Schloss 2009:211). This sys­tem shifted to sharecropping by 1849, through government-­organized associations of free laborers, as sugar harvests were failing and prices decreasing (Schloss 2009:211; Tomich 2004:184). Sugar estates continued to dominate the Martiniquan landscape after emancipation and almost to the turn of the twentieth century, despite the falling prices and diminished harvests. While nearly two-­thirds of former enslaved laborers continued to work on plantations as sharecroppers, many moved to towns to pursue other employment or established farms on abandoned estates (Tomich 2004:177). Government reports from the time immediately after emancipation indicate that laborers were attached to houses and provision grounds, which was a principal factor in keeping previously enslaved in­di­vidual resident on plantations (Tomich 2004:179).

Subsistence of Enslaved Laborers in the French West Indies The early diet of enslaved laborers in the colonies was similar to that of the Europeans, both the colo­nists and the engagés, and included generally salted beef, cod, or turtle and manioc or other tubers and root vegetables (Debien 1974:152). In the seventeenth century, Du Tertre (1667–1671:3:515–516) observed that many plantations relied significantly on rationing systems during the early colonial period, but that many planters were beginning to apply what he refers to as the “Brazilian” model, whereby the enslaved laborers were permitted to grow and raise some of their own foodstuffs on household and provision gardens on free Saturdays. The enslaved communities were provided land or “provision grounds” to cultivate and also raised crops in yard gardens adjacent to their cabins. In Mar­ti­nique, “provision grounds,” known as places à nègres, were plots of land on the plantation margins, of­ ten in the hills above the cane fields (Kimber 1988; Tomich 1991). As elsewhere in the Caribbean, Martiniquan enslaved laborers were permitted to grow crops in their house plots, raise livestock and poultry, and harvest fish and shellfish on “free” Sundays and sometimes Saturdays (Du Tertre 1667– 1671:3:515–516 cf. Tomich 1993). In 1685, French King Louis XIV issued an edict outlining the treatment of enslaved laborers in the French colonies, known as the Code Noir, or Black Code. The code provided guidelines for Christianization, encouraged marriage, discouraged the separation of families, provided detailed standards for food and clothing, and regulated manumission and punishment. Ostensibly to combat the “liberties” and malnourishment generated by the implementation of free Saturdays, the Code Noir of Louis XIV mandated a standard-

Enslaved Laborer and Sharecropper Households / 111

ized dietary minimum ration for enslaved laborers in all the French colonies. These weekly rations of cassava or manioc flour and fish or salt beef were known as the ordinaire (Debien 1964:176–186). In most of the French colonies, however, the planters overtly ignored the regulations (Dubois 2004:30). Throughout the eighteenth century, the French government passed vari­ ous ordinances to compel planters to provide rations, but in reality, these ordinances were not observed (Debien 1964:184; Tomich 1993:225). Finally, in 1786, the crown codified the allotment of provision grounds and gardens for self-­provisioning, requiring planters to distribute a portion of land to each enslaved in­di­vidual for their own profit, and to ensure that the gardens were in good state. The order mandated that if the gardens were not providing adequate subsistence, the planter must provide supplemental rations. While Mar­ti­nique planters allocated some imported rations, such as salt beef and salt cod, the provision grounds sys­tem (with house yard gardens) and free Saturdays and Sundays was the norm for most enslaved laborers in Mar­ti­nique during the nineteenth century and the primary sources of their subsistence (Schoelcher 1841; Tomich 1993:229). Enslaved laborers in Mar­ti­nique became so successful at gardening that by the early nineteenth century they were supplying the entire colony with crops from provision grounds and garden plots. Depictions of enslaved laborer settlements in Mar­ti­nique not only describe well-­cultivated house gardens and “vast” provision yards but also note the ubiquitous presence of pigs and poultry (chickens and guinea fowl), of­ ten in enclosures next to cabins or next to provision grounds (D’Orbigny 1841:26; Moreau de Jonnès 1817:65). In the early eighteenth century, Labat (1724:2:62) observed that enslaved laborers kept poultry and livestock in enclosed parks, noting that pigs were pervasive on plantations but were only slaughtered at festivals or holidays; otherwise, they were traded or sold at local markets. Poultry, and particularly eggs, were highly prized resources (Desalles 1786:131; Thibault de Chanvalon 1763:131). Further sources corroborate that enslaved laborers raised sheep, goats, and even some cattle but generally indicate that pigs were a more central focus (Ministère de la Marine 1844:332; Moreau de Jonnès 1817:65; Schoelcher 1841:13; Thibault de Chanvalon 1763:92). Despite the abundance of marine resources available, Moreau de Jonnès (1817:67) suggests that fresh fish furnished only a small portion of subsistence on the islands, and that cod and other salted fish from the metropole was necessary to provision the enslaved laborers in Mar­ti­nique. Thibault de Chanvalon (1763:108) observed that, as many planters provided few if any rations, enslaved laborers “sont obligés de la chercher d’ailleurs” (they are obliged to search elsewhere for their food), and that crustaceans, particularly

112 / Wallman and Kellly

crabs, offered “de grandes ressources” for subsistence (Thibault de Chanvalon 1763:108). The rare mention of fishing in his­tori­cal documents acknowledges the existence of a “few” enslaved fisherman but indicates that they shared their returns with their owners (Ministère de la Marine 1844:120, 33; Moreau de Jonnès 1817:67). Detailed descriptions of food processing and preparation are limited in written accounts related to plantation life in Mar­ti­nique. Some records, however, provide insight into food-­related activities. Labat (1724:1:56–57), for example, notes that some households of enslaved laborers constructed a small enclosure next to the house as a kitchen, but for the most part, they had fires within the cabin, and that most houses contained calabashes, canaris (cooking pots), tables, and a few utensils. Over a century later, Schoelcher (1841:2) observes that most cabins contained a bench or chair, a few pots of faience, or possibly just a single canari, a chest, and an earth floor, with a fireplace in the corner. The differentiation between “faience” and “canaris” in Schoelcher’s account suggests that the latter refers to locally produced ceramics typically used as cooking vessels, whereas faience seems to signify imported earthenwares. Archaeological and ethnohistoric research in Guadeloupe has determined that faience were used primarily as tablewares and not as serving vessels in the French West Indies (Arcangeli 2015:95). Refined wares are noted as “porcelains” in many French accounts and are only mentioned in passing as commodities purchased by enslaved laborers (e.g., Robin 1807:85). While the his­tori­cal record thus provides little indication of the ceramic foodways traditions that developed within enslaved communities on Martiniquan plantations, ar­chaeo­logi­cal data offer insight into local practices.

Habitation Crève Cœur Habitation Crève Cœur is situated on the south­ern Ste. Anne Peninsula of Mar­ti­nique, a karst landscape with a series of small hills composed of limestone and volcanic rock interspersed with small drainage systems (Figure 6.1) (Kimber 1988:17). The plantation was established approximately 1.5 km from the Marin Cul-­de-­Sac, an important economic and ecological harbor, with extensive shoreline mangrove swamps and coral reef formations. This region of Mar­ti­nique, in­clud­ing the communes of Marin, Ste. Anne, and Ste. Luce, among others, contained numerous large sugar plantations held by a small number of powerful and wealthy planter families. This area of the island drew settlers largely because of the Marin Cul-­de-­Sac, which offered an anchorage protected from winds and access to natural resources (Renouard 1822:207). The plantations in Ste. Anne were generally large and productive, focused almost exclusively on sugar production, and the plant-

Enslaved Laborer and Sharecropper Households / 113

Figure 6.1. Location of Mar­ti­nique and Habitation Crève Cœur (star). (Diane Wallman using ESRI topographic map. Sources: Esri, DeLorme, HERE, TomTom, Intermap, increment P Corp., GEBCO, USGS, FAO, NPS, NRCAN, GeoBase, IGN, Kadaster NL, Ordnance Survey, Esri Japan, METI, Esri China [Hong Kong], swisstopo, MapmyIndia, and the GIS User Community.)

ers in this region formed what was referred to as a “privileged aristocracy” (Ministère de la Marine 1844:96). Habitation Crève Cœur was part of this network of elite and prosperous planter families that controlled the south­ern portion of the island. Specifically, Crève Cœur was one of numerous sugar plantations owned by the Blondel family (which became Blondel La Rougery or Rougerie), an economically powerful family in Mar­ti­nique from the seventeenth through nine-

114 / Wallman and Kellly

Figure 6.2 Habitation Crève Cœur on 1770 map, marked as V. (Veuve) Blondel. (From René Moreau du Temple, 1770, modified by Diane Wallman.)

teenth centuries. Archival research has not yet established the exact founding date, but the plantation first appears on the Carte Moreau du Temple in 1770 as a sugar estate, with extensive industrial buildings, an animal mill, the maison de maître (the planter’s house/great house), and a row of at least 14 enslaved laborer houses (Figure 6.2) (Bousquet-­Bressolier et al. 1998). We have not been able to locate, as yet, detailed archival records of the makeup of the enslaved community at Crève Cœur. Based on the cartographic representations, the enslaved population likely averaged between 100 and 150 people, but at its peak in the early nineteenth century upwards of 200 or more enslaved people worked and lived on the plantation. Crève Cœur was an average-­sized plantation, approximately 200 ha (about 500 acres) at the turn of the eighteenth century (Blondel La Rougery 2009). The plantation produced solely sugar, as indicated in the 1817 agricultural census. By 1822, based on cartographic and financial records, the estate had nearly doubled in size, as the Blondel heirs purchased the adjacent plantation. On the eve of emancipation, however, demand for Caribbean sugar decreased and revenues diminished, leading to bankruptcies and the abandonment of plantations (Blondel La Rougery 2009:27). In 1844, the Blondel heirs began the sale of the plantation (which was to be a 10-­year process), possibly in anticipation of impending emancipation, to the former overseer, Dupperier Desportes, listed as a free man of color, finalizing the transfer

Enslaved Laborer and Sharecropper Households / 115

in 1854 at a total cost of 75,000 francs (Blondel La Rougery 2009:51). Despite the mass sale or abandonment of sugar plantations after the abolition of slavery in 1848, Desportes continued operating the plantation through sharecropping until the late nineteenth century. Today, the site contains standing ruins of industrial buildings, the maison de maître, outbuildings associated with sugar production and storage, and material deposits of enslaved laborer occupations. The site is situated on pub­lic lands used by the local government as a heritage site (Kelly 2008:396). Archaeological investigations targeted five terraced platforms on the slope of the hill above and beyond the planter’s residence that once accommodated the cabins of enslaved laborers. Excavation identified architectural features and household deposits within the occupation areas. House construction seems to follow the general pattern of slave cabins on Martiniquan plantations, as indicated by archival records. Houses were generally 16 to 20 ft. long by 10 or 12 ft. wide and had one or two rooms, depending on whether one or two families lived in the house (Labat 1724:1:56–57; Renouard 1822:108). Enslaved families on Mar­ti­nique, according to Re­nouard (1822:108), consisted of a married couple and their children. At Crève Cœur, the size of the residences, as well as the architectural features identified, such as post holes indicating interior walls, suggest a similar layout with one or two families residing in each cabin. Most of the trash deposits within the platforms were characterized as similar to sheet middens, where trash was deposited ad hoc through­out an area, of­ten downslope from the platform, as opposed to specific sec­ondary refuse areas, such as middens or prepared trash pits. Small house trash middens or pits with denser material deposits, however, were identified at Loci C, E, and F. Investigations recovered over 50,000 artifacts from Crève Cœur and focused on five occupation areas (Loci A, C, D, F, H) definitively associated with the enslaved and later sharecropping community, with limited testing around the planter’s residence and kitchen (Maison de Maître), and one locus determined to function as the plantation’s infirmary (Locus E) (Fig­ ure 6.3). Ceramics were the most abundant artifact class recovered, representing over 64% of the total assemblage. For each investigated locus, stratigraphy and features were analyzed to identify any relatively discrete occupations present within each plat­form. Mean ceramic dates (MCDs) were calculated for each deposit to provide a general chronology of occupation for the investigated loci (Binford 1987; South 1977). As this calculation provides a single date for deposits that span of­ten undefined periods of time, these dates are not intended to represent the exact date of occupation. Rather, the MCD provides an ap­proxi­ mate midpoint of an occupation and is useful in determining the rela­

Figure 6.3. Site map of Habitation Crève Cœur, Mar­ti­nique. (Diane Wallman and Kenneth G. Kelly)

Enslaved Laborer and Sharecropper Households / 117

Table 6.1. Mean Ceramic Dates for Distinct Deposits Locus

Context

E1 F1 C1 F2 C2 A1 F3 E2 D C3 A2 H

Infirmary Enslaved Laborer Occupation Enslaved Laborer Occupation Enslaved Laborer Occupation Enslaved Laborer Occupation Enslaved Laborer Occupation Enslaved Laborer Occupation Infirmary Slave/Sharecropper Occupation Slave/Sharecropper Occupation Slave/Sharecropper Occupation Sharecropper Occupation

Mean Ceramic Date 1766 1775 1787 1792 1808 1811 1820 1819 1830 1833 1834 1845

tive ­chronology of loci within a single site for intrasite comparison. Used alongside his­tori­cal records, these dates can provide rough approximations of the timing of particular deposits. After evaluating stratigraphy and features from each locus to refine chronology, the results reveal that the occupations span the period of operation of the plantation (Table 6.1). The chronology of the deposits allows for valuable comparison of intrasite patterns and variation over time, in­clud­ing a locus (H) likely occupied solely during the sharecropping period on the island.

Ceramics Overall, excavations recovered approximately 31,000 ceramics (excluding kaolin pipes). The assemblage includes a great diversity of ceramic types with a combination of imported European wares, local industrially produced ceramics, and local low-­fired coarse earthenware. Imported ceramics in­cluded mostly French-­produced commodities, in­clud­ing refined earthen­ wares (creamware, pearlware, whiteware, faience), European porcelain, utili­ ta­rian coarse earthenwares (Huvaeune, Vallauris, Saintonge), and both French and English stonewares. A low-­fired coarse earthenware, known colloquially as coco neg, was identified in the assemblage. This ware is coil-­or hand-­built, with thick and of­ ten smoothed walls. These ceramics have a highly variable paste and can be found slipped and/or burnished and were identified through surface collections at several plantations in Mar­ti­nique (Kelly et al. 2008). The production of this ceramic has been documented ethnographically in modern

118 / Wallman and Kellly

times (Verin 1963), and several forms were identified, but coco neg is specifically associated with hollowwares, primarily cooking pots. The pots had many forms and some had small applied handles, although analy­sis identified no clear decoration on the pots. Excavation also recovered evidence of locally produced industrial coarse earthenwares, referred to as terre cuite commune. These are locally mass-­produced coarse wheel-­thrown redwares found in Mar­ti­nique, produced through slave labor (England 1994; Kelly et al. 2008). These ceramics are characterized by a soft to medium red paste and for this study are separated into two general categories. Type 1 is an industrial ware, coarse, wheel-­thrown and thick walled, with orange-­red paste and large grog, limestone, and detrital inclusions and are generally found in the form of drip jars, sugar cones, and tiles (Kelly et al. 2008:89). Type 2 is also wheel-­thrown but is thin-­walled, has a coarse reddish-­brown paste with felspathic detrital and limestone inclusions. These ceramics are of­ten burnished and slipped and are found in the form of beverage vessels, such as carafes, but also as household utilitarian wares, such as cooking pots and serving vessels (Kelly et al. 2008:89). Local coarse earthenware (terre cuite commune), which includes mostly tiles and sugarwares (Type 1) but also food and beverage related vessels (Type 2), is the most common ceramic type identified, making up approximately 40% of the total assemblage. Imported refined wares are the next most frequent ceramic, representing approximately 24% of the assemblage. Coco neg represents about 24% of the total ceramic assemblage, with imported coarse earthenwares (utilitarian wares) (11%) and stonewares (1%) making up the remainder of the assemblage. To evaluate the importance of coco neg at the site, we examined the relative proportion of this ceramic type in relation to the frequency of imported ceramics of similar function, specifically those used in food preparation (imported coarse and refined earthenwares, and stonewares) by locus and chronology (Figure 6.4). Generally, coco neg was produced and used consistently by the enslaved community at Crève Cœur and into the postemancipation period. The data, however, indicate that coco neg decreases in importance in the early to mid-­nineteenth century within the enslaved households and is relatively less common during the sharecropping period (Locus H). Time, however, is not the only indicator of variation in the use of this ceramic within the site. Overall, Loci A, C, and D have high percentages of coco neg. For Locus A and Locus C, this trend is consistent through­out the occupation of the platform, although in the most recent deposits, the relative abundance of coco neg decreases. The samples from Locus F have a relatively low proportion of coco neg, and its importance decreases through time at this house platform. Coco neg is the least common at Locus E, which likely reflects that this platform is not directly associated with an enslaved household.

Enslaved Laborer and Sharecropper Households / 119

Figure 6.4. Relative abundance of imported ceramics and coco neg by locus and chronological context. MCD = mean ceramic date. (Diane Wallman)

Furthermore, imported ceramics increase over time, within particular loci and across the village in general. Specifically, refined earthenwares (creamware in earlier deposits and whiteware in later deposits) and European porcelain increase. Overall, the preferred ceramic types of enslaved laborer households shift from faience, Huveaune, and coco neg to whiteware and Vallauris. Vallauris is typically found in the form of cooking vessels and therefore is used as an alternative to coco neg. All deposits within the enslaved laborer village contained mostly hollowware forms of coco neg, with large assemblages of canaris (cooking pots), analogous to yabba pottery on Jamaica and other British islands. Approximately 42% (n = 2,438) of the coco neg sherds recovered were classified as cooking pot or crock, with bowls as the only other identified form.

Zooar­chaeo­logi­cal Analysis The methods used in faunal analy­sis involved standard zoo­ar­chaeo­logi­ cal techniques (Claassen 1998; Klein and Cruz-­Uribe 1984; Lyman 1994; O’Connor 2000; Reitz and Wing 2008; Wheeler and Jones 1989). Specimens were classified by skeletal part and symmetry, and the remains were identified to the lowest taxonomic category possible using a comparative col-

120 / Wallman and Kellly

lection and reference materials. Specimens without diagnostic features were assigned to animal size class categories (adapted from Thomas 1969). For this study, we used number of identified specimens (NISP) and minimum number of individuals (MNI) for mammal, fish, avian, reptile, and mollusk remains as proxy measures for relative contribution to the diet. For fish and shellfish, taxonomic richness, heterogeneity, and evenness were calculated to compare the diversity of taxa across deposits. Richness is equal to the total number of species identified within each assemblage (NTAXA) (Grayson 1984). Heterogeneity was measured using the Shannon index, ∑pi *ln pi, where pi is equal to the proportion of specimens in the assemblage identified as species (Grayson 1984; Lyman 2008:192). The number varies generally between 1.5 and 3.5, and the higher the number, the more heterogeneous the distribution of specimens across species (Lyman 2008:192; Magurran 1988:35). Evenness evaluates how individuals are distributed across taxa to establish if all taxa are represented by the same number of individuals (even) or not (Lyman 2008:194). A common measure uses the Shannon heterogeneity index divided by the log of NTAXA (e = HlnS ), with a value of 1 indicating an even distribution of fauna, and 0 indicating an uneven distribution. Analysis identified 11,508 specimens, in­clud­ing 8,848 specimens of fish, mammal, bird, amphibian, and reptile. Fish and mammals dominate the vertebrate assemblage, and birds, reptiles, and amphibians are less common (Figure 6.5). The invertebrate specimens identified include those from the phyla Mollusca (vari­ous bivalves and gastropods n = 2,399), Echinodermata (sea urchin, n = 214), and Arthropod (crab, n = 1). Old World domestic herbivores dominated the assemblage, a pattern commonly encountered on plantation sites, both in deposits associated with planter residences and in enslaved community contexts (Singleton 1996:​ 154). As indicated by his­tori­cal accounts, and based on NISP, pig was the most important mammal in the assemblage (Table 6.2). When evaluated by MNI, however, Caprines (sheep and goat) were the most common taxa, representing more than 30% of the mammal assemblage. This pattern is likely due to the high proportion of pig teeth in the assemblage and the fact that suids have more skeletal elements overall than bovids (Lyman 2008:228). Sheep and goat were represented approximately equally in the assemblage, suggesting that there was no preference for either particular species. Cattle accounted for approximately 10% of the total MNI, but as will be discussed, it is unlikely that the remains identified as Bos taurus represent whole animals. Both domestic dog and cat remains were identified within the assemblage. These animals were likely utilitarian or pets, imported to aid in the control of introduced rodents. Wild terrestrial mammals identified in the assemblage included opossum, mongoose, agouti, rat, and mouse. Of these, the

Enslaved Laborer and Sharecropper Households / 121

Figure 6.5. Assemblage composition by class and locus. NISP = number of identifiable specimens. (Diane Wallman)

agouti was the only animal endemic to the island, not introduced by human settlers. Known as the manicou in Mar­ti­nique, the opossum was likely introduced by Precolumbian migrants from South America (Masseti 2011). Finally, European rats and mice were identified within the assemblages. This trend is not surprising as reports (see Du Tertre 1667–1671; Labat 1724) indicate the presence of these exotic species on the island as early as the late seventeenth century. These animals would have found abundant sustenance in the crops and refuse associated with the plantation. It is unclear whether these rodents were consumed by the site occupants, as the remains show no indication of butchery or burning. Analysis of skeletal part representation of domestic mammals reveals the methods of acquisition and consumption of these fauna. If animals were raised on site, as opposed to imported in the form of barreled beef and pork, then remains should include skeletal elements from the entire carcass. Rations of salt beef or the purchase of meat from markets would be indicated by the presence of elements, such as ribs, vertebrae, and some long bones, associated with barreled and cured products, like beef, pork, and cod, imported from abroad (Klippel 2001). The cattle remains have more axial elements and fewer cranial and foot remains than expected, whereas the pig and sheep/goat samples have more cranial elements and fewer axial and foot specimens (Table 6.3). Results of

Table 6.2. Number of Identified Specimens and Minimum Number of Individuals of Identified Mammals from Habitation Crève Cœur Taxon Domestic Mammals Equus sp. Sus scrofa Bos taurus Ovis/Capra Ovis aries Capra hircus Medium Artiodactyl Canis familiaris Felis cattus Wild Mammals Didelphis marsupialus Herpestes sp. Dasyprocta leporina Rattus sp. Mus sp. Rodentia Large Mammal Medium Mammal Small/Medium Small Mammal Very Small Mammal Mammal Total

Common Name

NISP

% NISP

Weight

% Weight

MNI

% MNI

Mule/Donkey Pig Cattle Sheep/Goat Domestic sheep Domestic goat

3 213 106 93 31 26 3 21 6

 0.10  7.08  3.53  3.09  1.03  0.86  0.10  0.70  0.20

46.70 498.09 1,364.55 102.37 88.26 121.31 0.50 20.27 0.13

 1.58 16.81 46.06  3.46  2.98  4.09  0.02  0.68  0.00

  1  35  19  12  15  15   0  10   4

32 6 18 46 25 4 163 1,158 12 471 9 561

 1.06  0.20  0.60  1.53  0.83  0.13  5.42 38.51  0.40 15.66  0.30 18.66

12.08 11.70 6.71 1.85 0.55 0.06 213.85 376.74 4.58 45.21 0.12 47.08

 0.41  0.39  0.23  0.06  0.02  0.00  7.22 12.72  0.15  1.53  0.00  1.59

 14   2  10  16  10

 1 21 12  7  9  9  0  6  2  0  9  1  6 10  6

Dog Cat Opossum Mongoose Agouti Old World Rat Old World Mouse Cow, Horse size Pig, Sheep/Goat Dog Agouti, Opossum, Cat Rat, Mouse

3,007

MNI = mean number of individuals, NISP = number of identifiable species

2,962.71

163

Enslaved Laborer and Sharecropper Households / 123

Table 6.3. Expected Versus Observed Ratios of Skeletal Portion of Domestic Mammals Expected Bovid Observed (Cattle and Observed Ovis/Capra Expected Sheep/Goat) Bos (Cattle) (Sheep/Goat) Sus (Pig) Head Foot Axial Upper limb

36 62 83 14

5 5 69 27

100 9 19 22

36 94 61 14

Observed Sus (Pig) 142 12 31 28

χ2 analy­sis of expected frequencies of skeletal elements for each of the taxa suggest that for all species the observed frequency is significantly different from the expected (bovid = cattle and sheep/goat, Sus = pig) (Bos taurus χ2 = 55.92, df  = 3, p < 0.001; Ovis/Capra χ2 = 107, 58, df  = 3, p < 0.0001; Sus scrofa χ2 =140.91, 3, p < 0.0001). This pattern follows the expectation that pig, sheep, and goat were raised on site, whereas beef was imported. Axial and upper-­limb elements (vertebrae, ribs, scapulae, pelvises, humeri, ­radii, ulnae, femora, tibia, patellae) are of­ten considered high-­utility elements (Binford 1987; Klippel 2001:1193). Historic studies indicate that axial elements were included in barreled shipments but of­ten purport that marrowbones were not, as they would increase the likelihood of spoilage (Klippel 2001:1193). Using staple carbon isotope analy­sis on bovid remains, however, Klippel (2001) found long bone elements were in fact imported to the Caribbean as barreled, salted beef. This pattern is also evident at Crève Cœur. The relative low proportion of cranial and foot elements supports the supposition that cattle remains at Crève Cœur mostly represent imported, cured meats either distributed as rations or purchased by the enslaved laborers themselves. The sheep/goat and pig remains, on the other hand, both demonstrate a pattern consistent with what would be expected if these livestock were raised locally. The overrepresentation of cranial remains for both Caprines and Sus scrofa suggest that these fauna were not imported and were likely raised and slaughtered on or near the plantation. Cranial remains, in particular, are of­ten considered a “low-­utility” portion of the animal (despite containing ample calories from grease and fat), and the presence of these bones (the abundance of teeth, in particular) indicates that the primary butchery of these animals occurred at the site. If these animals are raised on site, the underrepresentation of axial, limb, and foot elements for sheep/goat and pig could reveal two patterns: variation in the taphonomic, depositional, or

124 / Wallman and Kellly

postdepositional processes affecting the underrepresented elements, or the marketing of the more “valuable” (high-­utility) elements by the enslaved community. The fragmentation and butchery patterns indicate that the former is more likely (Wallman 2014). Analysis identified 32 families of fish and 52 total separate taxa that were consumed by the inhabitants of Crève Cœur. All identified species are available in the mangrove swamps, inshore areas, and reef within the nearby bay (see Wallman and Grouard [2017] for further analy­sis of fishing practices). The most common families identified in the assemblage include Scaridae (parrotfishes), Serranidae (groupers), Haemulidae (grunts), and Lutjanidae (snappers). Carangidae (scads), Acanthuridae (surgeonfish), and Balistidae (triggerfish) are also important species in the assemblage. To evaluate differences between habitation loci, we can look at the overall percentages of MNI for the most common families represented across the site (Table 6.4) For this discussion, only those samples with sufficient sample sizes (NISP > 60) are included (Loci C, E, F, H). Richness, heterogeneity, and evenness indexes were calculated for each locus based on family. On evaluation, there are some definite patterns within the data. Locus C has a high proportion of Balistidae (triggerfish), Scaridae (parrotfish), Acanthuri­ dae (surgeonfish), and Carangidae (jacks and scads). Locus E is relatively evenly distributed (Evenness = 0.91). This locus has a very high percentage of Haemulidae (grunts), and other important taxa include Serranidae (grouper), Scaridae, Lutjanidae (snapper), and Carangidae. Locus F has higher proportions of Serranidae, Lutjanidae, Haemulidae, and Carangidae. Locus H is more even than C and F and has higher frequencies of Scaridae, Lutjanidae, and Haemulidae, and the data indicate a shift in procurement strategy and preferred species during the postemancipation period. Locus C and Locus F are the most diverse and also most uneven of these samples, which suggests a wider breadth of taxa that are procured more randomly. The variation among household deposits indicates that while certain taxa are common through­out the samples, household units had unique patterns of acquisition and consumption. When the contribution of different taxa (by MNI) is evaluated over time, other patterns emerge (Wallman 2014). Generally, Belonidae (needlefish) and Scaridae (parrotfish) increase in importance over time, whereas Carangidae, Clupeidae, Holocentridae, and Centropomidae generally decrease in relative abundance. Certain taxa vary slightly between deposits but are generally consistently represented through time (Lutjanidae, Serranidae, Haemulidae). Late eighteenth-­century deposits do exhibit less diverse and more even samples than later deposits, with Locus H exhibiting a unique pattern that suggests a shift in postemancipation foodways.

Enslaved Laborer and Sharecropper Households / 125

Table 6.4. Percentage Minimum Number of Individuals of Common Fish Taxa by Locus from Habitation Crève Cœur Taxon

Common Name

Locus C (N = 481)

Locus E (N = 65)

Clupeidae Belonidae Hemiramphidae Holocentridae Centropomidae Serranidae Malacanthidae Carangidae Lutjanidae Haemulidae Sparidae Labridae Scaridae Acanthuridae Sphyraenidae Balistidae

Herring/Sardines Needlefish Halfbeak, balaou Squirrelfish Snook Grouper Sand tilefish Jacks and Scads Snapper Grunts Porgy Hogfish Parrotfish Surgeonfish Barracuda Triggerfish

  3.50%   4.00%   1.00%   3.00%   2.50%   7.10%   2.00%   8.10%   5.60%   7.60%   3.50%   0.50% 12.10%   8.10%   5.60% 14.60%

  7.00%   7.00%   3.50%   5.30%   5.30%   8.80%   1.80%   8.80%   8.80% 17.50%   3.50%   5.30%   8.80%   0.00%   3.50%   5.30%

NTAXA (Families)*

Locus F Locus H (N = 641) (N = 165)   3.70%   4.20%   5.30%   1.70%   2.00% 19.10%   0.60%   9.30% 11.80%   9.80%   1.70%   2.20%   8.40%   5.10%   4.50%   5.60%

  6.00%   9.50%   1.20%   1.20%   2.40%   6.00%   3.60%   6.00% 10.70%   7.10%   1.20%   0.00% 19.00%   6.00%   2.40%   6.00%

26

15

26

20

Heterogeneity

2.55

2.47

2.57

2.70

Evenness

0.78

0.91

0.79

0.90

* Includes less common families not listed here. NTAXA = number of taxa identified

Burgot (Cittarium pica, West Indian topshell in English) is the most common shellfish species recovered from Crève Cœur. This gastropod represents over 50% of the total MNI of the shellfish sample. Lucines (10%) are the next most common, followed by mangrove oysters (Crassostrea rhizophorae, 9%) and arks (Anadara notabilis, 5.55%). The remaining taxa are uncommon, with some represented by a single in­di­vidual and make up just 25% of the total shellfish sample. Table 6.5 shows the frequency distribution of MNIs for only those taxa definitively categorized as food resources. Evaluating the samples across deposits, Loci H and F have the highest diversity of taxa (NTAXA = 13 and 12, respectively), and these households consumed relatively more oysters and clams when compared with the other enslaved laborer households. The sample from the maison de maître is the most even of the deposits, whereas C is very uneven. In contrast to the enslaved laborer occupations,

126 / Wallman and Kellly

Table 6.5. Relative Abundance of Food-Related Shellfish Taxa (% MNI) by Locus Taxon

A (%)

C (%)

D (%)

E (%)

F (%)

H (%) M (%)

Chitonidae Anadara sp. Dossinia concentrica Codakia orbicularis Lucina pectinata Crassostrea rhizophorae Strombidae Cittarium pica

 0  0  0  0  4  0 12 85

  1   8   1   1   4   3   3  79

 3  0  0  0  0  0  3 94

 3  3  0  5  8  5  3 70

  4   5   6   5  14  13   1  51

 6  5  1 10  4  8  2 64

  1  15   2  14   8  34   1  23

Total MNI

26

159

33

37

188

99

104

the sample from the planter’s complex has relatively more oysters and clams (Anadara, Codakia, Lucines).

Discussion Habitation Crève Cœur was established around 1760 during the peak of sugar monoculture in the French West Indies. Owned by a wealthy and pow­ erful béké family in Mar­ti­nique, this plantation was highly successful, with upwards of 200 or more enslaved laborers at its height in the early nineteenth century. The estate was typical of Ste. Anne, a region known for large and productive plantations run by a few elite planter families who constituted an aristocracy in the region. To investigate the local manifestations of the processes working on regional and global scales, ar­chaeo­logi­cal excavations were completed at this plantation, focusing on the recovery of materials associated with the enslaved individuals who lived and worked on the estate. The results of these investigations offer substantial insight into the lifeways of a particular community of enslaved laborers, working for the benefit of a wealthy white family. At Crève Cœur, archaeology reveals the development of extremely localized (at the household and community level) social and economic strategies by enslaved laborers to confront the constraints of plantation slavery. Ceramic production and consumption at the site suggest that enslaved ­laborers developed household and community traditions in how food was prepared and consumed. The use of different types of ceramics reveals changing access to commodities over time. The village was the site of extensive use and, based on abundance and ubiquity, possible fabrication of coco neg

Enslaved Laborer and Sharecropper Households / 127

(although production sites were not identified on the site). The area of Ste. Anne is known his­tori­cally as an area of ceramic production, and many enslaved households in this region likely made pottery. Immediately following emancipation, a government official in Ste. Anne observed that “the associated workers, whether on half or on third shares take all Friday to cultivate their gardens, to fish, to make pottery or other objects of personal interest” (Perrinon 1848 in Tomich 2004:181–182). The ruins of several large historic kilns were identified by survey in the south­ern region of the island, associated with pottery production estates (Habitations Poterie) (Kelly et al. 2008). These sites, however, focused primarily on the large-­scale production of wheel-­thrown domestic pottery (Type 2). Furthermore, contemporary observations of a local traditional potter who makes coco neg suggest that these ceramics are pit-­fired on a small scale, suggesting that the production of these wares was likely a community or domestic practice on plantations through­out the island. Enslaved households used coco neg primarily for cooking, as most of the forms identified were cooking pots, but bowls were also manufactured for food consumption. Overall, coco neg was present in all contexts, ranging in frequency among the houses. For all time periods, the occupants of Locus F used significantly less coco neg than contemporaneous households. The assemblage reveals a shift in the abundance of this ceramic over time, as later occupations at each locus demonstrate less reliance on coco neg in the mid-­ nineteenth century. In later occupations, and particularly postemancipation contexts, the production and/or use of coco neg decline, with a simultaneous increase in imported wares, such as Vallauris, which served the same function in food preparation. The changes in ceramic consumption over time and the variation among households indicate that economic free­dom and access to goods were higher in some households and that the community’s commodity consumption as a whole increased approaching emancipation. The production of coco neg represents a creative and resourceful strategy developed by the enslaved laborers, who had little capital and highly constrained lives. Over time, the enslaved laborers at Crève Cœur clearly gained increasing financial autonomy, allowing for the acquisition of previously inaccessible commodities, in­clud­ ing imported ceramics. This trend continued into the postemancipation period, as the sharecropper households consumed more imported materials. Archaeological evidence provides a much more refined and detailed evalua­ tion of the foodways developed by the enslaved individuals who lived and labored on a particular plantation. While the documentary record can reveal general patterns and evaluate how the colo­nists viewed the subsistence practices of enslaved laborers, ar­chaeo­logi­cal data offer substantial insight

128 / Wallman and Kellly

into how practice is manifested on the local level. The enslaved community at Crève Cœur developed strategies specific to its situation and the plantation’s particular labor regime and schedule, and based on its own his­tori­cal experiences. The ar­chaeo­logi­cal materials recovered present some indication of the food resources that constituted the subsistence regime, the fauna in particular, of the enslaved laborers at Crève Cœur. Yard gardens adjacent to enslaved households on the slopes of the hill likely provided fruits, vegetables, herbs, and medicinal plants. While the botanical evidence remains unanalyzed, the continued presence of tamarind trees on the site speaks to the planting of nonnative trees within the enslaved laborer village during the colonial era. The recovery of manioc graters from multiple platforms denotes the processing of manioc flour. Agricultural tools (e.g., hoes, machetes) might be tools of work in the cane fields and were of­ten used by enslaved laborers as personal tools. The ubiquity of these personal implements associated with enslaved laborer occupations indicates significant investment into the small-­scale cultivation of provisions (vari­ous tubers, root plants, and vegetables). The well-­preserved faunal materials provide the most telling evidence regarding the specific subsistence practices and traditions cultivated by the enslaved community at Crève Cœur. At Crève Cœur, enslaved laborers depended largely on fish and shellfish, particular domestic mammals, supple­ mental wild mammals, and the occasional bird or reptile. Sheep, goat, and pig, those animals mentioned in his­tori­cal accounts as rarely consumed (during celebrations, or festivals), but commonly raised and marketed by enslaved individuals, made up the majority of the mammal remains identified. Skeletal representation indicates that for medium mammals, the entire animal was used, but taphonomy reveals that these animals were butchered and fragmented into small pieces. As is common in the Caribbean and West Africa, the butchery techniques suggest preparation of stews or gravies and the retrieval of marrow in long bones (Armstrong 1990; DeCorse 2001; Wilkie and Farnsworth 2005). This pattern is also consistent with the preponderance of cooking pot forms ideal for stewing, and the high proportion of bowl forms identified in the slave village deposits. Locally raised animals were the primary mammal resources consumed, whereas beef, while consistent within the deposits, was less frequent and likely imported as salted or cured elements. According to documents, manioc, other provisions (yams, plantains, potatoes), and, in particular, salted cod (referred to as “le pain des nègres” or “the bread of the slaves”) formed the foundation of subsistence in the Antilles (Moreau de Jonnès 1817:4). Despite this implication in his­tori­cal accounts that both salt fish and salted beef remained important staples for enslaved

Enslaved Laborer and Sharecropper Households / 129

laborers in the French West Indies, the archaeology suggests that households consumed little to no salt fish and relatively negligible amounts of barreled beef. Other mammals supplemental in the diet include opossum, agouti, and mongoose. Bird shot recovered supports the evidence that small mammals were hunted, but they were also likely captured by hand or trapped. Birds were very rare within the assemblage, only represented by chicken, guinea fowl, and pigeon, all fauna observed in his­tori­cal accounts as common within enslaved laborer settlements. The high value of these resources suggest that the enslaved laborers at Crève Cœur saw more advantage in the profits acquired through the sale of these animals than in their dietary benefits. Poultry and eggs offered the opportunity for income with little investment of time and energy, and eggs also provided steady protein and calories for the enslaved laborers themselves (Handler and Wallman 2014; McDonald 1993:282). Other uncommon and supplemental food-­related taxa include sea turtle, frog, and possibly dog. The faunal assemblage revealed that fishing was an important practice for the enslaved laborers at Crève Cœur. The rare documentary account mentions that some wealthy plantations had specialized fisherman, and the his­tori­cal record implies that this practice played only a marginal role in enslaved laborer subsistence. Archaeological evidence presents a contrasting view of the importance of fishing. The identification of a great diversity of fish within the enslaved laborer occupation deposits suggests that fish were an important element of subsistence. Over time, the diversity of fish is relatively constant, although certain taxa became more important in later deposits. Generally, Belonidae (needlefish) and Scaridae (parrotfish) increase in importance over time, whereas Carangidae, Clupeidae, Holocentridae, and Centropomidae generally decrease in relative abundance. These patterns signify a shift in procurement location and species preference during the sharecropping period. Mollusks were also important to the diet of the enslaved laborers at Crève Cœur. Based on the ubiquity and density of shellfish remains through­out the deposits, mollusks provided a consistent and common resource for the enslaved community at the site. This pattern is in line with his­tori­cal accounts that suggest shellfish provided resources for enslaved communities. In particular, Cittarium pica, or burgot, was a significant element of subsistence on the plantation. This gastropod was identified in nearly all enslaved laborer household deposits in relatively high frequency. Other shellfish consumed included conch, vari­ous sea clams, and mangrove oysters. Like ceramics, the fauna recovered within the enslaved laborer village at Crève Cœur indicate generally shared community traditions that developed over years of daily life. The rue des negres (slave row), however, was also a

130 / Wallman and Kellly

space of conspicuous socioeconomic differentiation between enslaved households. Loci C and F provide important examples of this variation. Locus F, in particular, exhibited a pattern more similar to Locus E, in which beef was slightly more common. The inhabitants of Locus F also consumed more poultry (chicken and guinea fowl), exhibiting more economic flexibility, as these were highly valued resources. The households at both Locus F and Locus C consumed a high diversity of fish remains but preferred slightly different taxa, suggesting that each household provisioned for themselves. Additionally, Locus F had more diverse shellfish than Locus C and consumed more clams in a pattern similar to that of the planter’s residence. Finally, there are notable changes in species preference of fish and shellfish in mid-­to late nineteenth-­century deposits likely associated with sharecropping households.

Conclusion Confirming interpretations of the ethnohistoric record, ar­chaeo­logi­cal evidence indicates that certain households (represented here by Locus F), held privileged positions from the founding of the plantation. The hierarchical structure of plantation labor on the island was already well developed in Mar­ti­nique on the establishment of Crève Cœur in the mid-­eighteenth century (Debien 1964:92; Moitt 2001:78). The unique material culture associated with Locus F, a platform in close proximity to the planter’s complex that dates to the beginning of the plantation’s operation, demonstrates that the differentiation of enslaved laborers by status was instituted from the beginning of the operation of the plantation. This household sat just a few meters downslope of the plantation infirmary, which was directly adjacent to the planter’s residence and kitchen. The proximity to the planter’s complex alone may not indicate a household’s privileged status. When considered with the early dates of the material culture associated with the deposits at Locus F, however, the evidence suggests that individuals in this household already assumed privileged or skilled positions before the establishment of Crève Cœur. Other material culture (e.g., luxury commodities, sewing materials) recovered from this locus support this claim. While variation among households indicates differential access or privileged status on the plantation, overall, the patterns of consumption within the enslaved laborer village reveal the traditions particular to this community and an increase in access to commodities and an increase in self-­provisioning over time. The postemancipation deposits indicate some changes to foodways traditions, such as a shift in ceramic consumption and in the procurement of different fish and shellfish. This pattern suggests that the free labor-

Enslaved Laborer and Sharecropper Households / 131

ers from Crève Cœur had more time and different preferences and began to exploit alternative habitats after emancipation. Despite their situation, the enslaved laborers at Crève Cœur clearly developed skills and knowledge through their daily practices that allowed them to create and maintain certain lasting traditions and develop some measure of economic power. Their success in raising livestock and poultry allowed them to rely on these resources for provisioning and profit. The diversity of fish and abundant shellfish reveal the proficiency with which they procured marine resources, using traditions transferred from Africa and techniques developed within this new context to exploit these resources for their benefit. The enslaved community at Crève Cœur produced material commodities likely intended for marketing, in­clud­ing low-­fired coarse earthenware pots and jars, jewelry, and other crafts (see Ahlman et al., chapter 4 herein; Gilmore, chapter 2 herein). The cultivation of crops, the procurement and raising of animal resources, and the fabrication of marketable goods were creative adaptations developed by enslaved laborers at Crève Cœur to confront the severe constraints of the plantation landscape. These unique practices became a means of economic and social mobility, successfully setting the foundation for the transition to emancipation. Such “peasant-­like” characteristics (Mintz 1974:143), in­clud­ing small-­scale cultivation, access to land to cultivate, the production of some commodities for sale, and self-­provisioning, continued in Mar­ti­nique after the abolition of slavery.

7 From Slavery to Freedom Changes in Afro-­Antiguan Lifeways, 1790–1840 Samantha Rebovich Bardoe

Although slavery was abolished in the British Empire in 1834, through­out the Caribbean a sys­tem of apprenticeship was put in place in which formerly enslaved people would serve as apprentices under their former masters for four years before gaining full free­dom in 1838. In Antigua, how­ever, planters voted for full emancipation in 1834, setting the stage for a social experiment that would be watched through­out the Caribbean. In the case of postemancipation life in Antigua, planters and colo­nists were concerned with keeping the social relations of slavery by maintaining control over the newly freed labor force and ensuring that the profitability of sugar plantations was maintained. For formerly enslaved people, though, free­dom offered opportunities that were previously denied during the period of slavery. Archaeological excavations at Green Castle Estate, one of Antigua’s largest sugar plantations, offer insights into how Afro-­Antiguans lived during the periods of slavery and free­dom on the island. Established in the late seventeenth century and operated through­out the eighteenth and nineteenth centuries, Green Castle Estate underwent a series of renovations by its owners, the Martin family. Artifacts dating to the late eighteenth and mid-­nineteenth centuries demonstrate shifts in the lifestyles of laborers on the estate that point to broader social shifts as laborers became free. Interpreting the ar­ chaeo­logi­cal evidence from Green Castle through theories of labor, practice, and agency gives insights into the ways formerly enslaved people took advantage of the new opportunities available to them, such as greater access to markets, increased autonomy in their domestic lives, and more educational opportunities.

Building a Sugar Island: Antigua in the Seventeenth and Eighteenth Centuries In the early seventeenth century, as European powers clamored for ownership of the Caribbean islands, sugar had not yet made its mark on the region. The islands of the east­ern Caribbean are small in size, yet, the English, French, and Dutch recognized their potential for agricultural production

Changes in Afro-Antiguan Lifeways / 133

and fiercely competed to maintain their holdings in the region (Engerman 1996; Gaspar 1985; Gaspar and Geggus 1997; Goveia 1965; Mintz 1974). The British Caribbean was almost immediately transformed for the production of cash crops using the plantation as the center of production (Mintz 1985; Sheridan 1974). Although Antigua was Britain’s most profitable sugar island in the Leeward Islands, by the turn of the eighteenth century, the early settlement of the island in no way foreshadowed its future success. First spotted by Columbus during his sec­ond voyage in 1493 (Dyde 2000:8), Antigua was settled in 1632 by British settlers from St. Kitts (Dyde 2000:13). Agricultural production was the goal of settlement, yet the initial settlers were challenged by the task of clearing the land that was wooded up to the ­water’s edge (Dyde 2000:14–15). Drought also plagued Antigua through­ out its early settlement (Dunn 1972:117). Within nearly ten years, though, agricultural production was underway in Antigua, and tobacco was being grown on the island in 1643 (Dyde 2000:16). In addition to tobacco, other cash crops were planted on the island during its early settlement, in­clud­ing indigo, cotton, and ginger. By 1655, sugar had been introduced to the island (Dyde 2000:20; Gaspar 1985:66). From its initial settlement through the 1670s, the population of Antigua was largely transient. Throughout the sec­ond half of the seventeenth century, the popu­larity of sugar increased as a group of elite planters arrived and, seeking to mimic the success of sugar production in Barbados, be­gan to change the forested landscape of Antigua into a series of sugar plantations. In the 1670s, the Dutch invasion of Suriname spurred the arrival of several wealthy planters, in­clud­ing Major Samuel Martin and William Byam, who began to form a more permanent planter class on the island (Dunn 1972:130; Gaspar 1985:66). More land was granted to wealthy planters, and larger estates soon outnumbered the smaller farms that dotted the landscape in its initial settlement. This transformation marked a huge change to Antigua’s physical and social landscape. With the rise of large plantations came the shift from indentured, European labor to enslaved, Af­ri­can labor. The rise of plantations physically changed the landscape of Antigua in overtaking small farms and also drastically altered social relations on the island. Whereas the population of whites in Antigua only increased by a few hundred, from 2,308 in 1678 to 2,892 in 1708, the population of blacks increased by over 500%, from 2,172 people in 1678 to 12,960 in 1708 (Dunn 1972:141). No other island in the Leewards would match Antigua’s growth during this period (Dunn 1972:141). As is evident, within the 30 years from 1678 to 1708, the scale of Antigua’s sugar production dramatically changed. On the arrival of Christopher Codrington, the renowned Barbadian sugar planter, Antigua’s sugar industry reached new levels of production.

134 / Bardoe

Codrington established his central plantation, Betty’s Hope, in Antigua and also acquired the entire island of Barbuda off Antigua’s north coast as a provisioning plantation to support his sugar plantations (Gonzalez de Scollard 2009). The magnitude of Codrington’s enterprises catapulted the Antiguan sugar market. As early as 1708, Antigua had more than half of the slave population in the Leeward Islands and accounted for more than half of the Leewards’ sugar output (Dunn 1972:141; Gaspar 1985:68). While Antigua had made a mark for itself in the Caribbean sugar market in the late seventeenth century, its rapid rise to success did not secure its position as the most profitable and po­liti­cally important island in the Leeward Islands but rather only highlighted how much was at stake in maintaining this position. Such an increase in production led to incredibly quick and substantial increases in Antigua’s population, particularly its slave population, and a devastatingly rapid deforestation of the island and building of sugar plantations. Although the immediate outcome of such work might be a sudden surge in profitability for the island’s sugar industry, Antiguan planters would soon have to contend with challenges from Antigua’s semiarid environment and social situations that undermined the very foundation of the sugar industry. Despite these challenges to Antigua’s sugar industry, it continued to grow. By the third quarter of the eighteenth century, Antigua’s sugar production reached its peak. The slave population of the island reached its height at 37,808 people in 1774 (Sheridan 1960:127). In addition, the consolidation of estates continued, resulting in Antigua’s sugar industry being reliant on a number of wealthy planters with large estates. The tax records of St. Mary’s Parish, for instance, show that by 1767 all 65 taxpayers were slave owners, and a third of them (22) owned over 100 enslaved laborers. The number of total enslaved laborers in the parish also ballooned to 5,610 (from 1,150 in 1706), and the total number of acres under production was up to 12,350 (from 5,660 in 1706) (Dunn 1972:141; Oliver 1894:394–397). Throughout the eighteenth century, sugar society in Antigua thrived. At the turn of the nineteenth century, though, abolitionist sentiment rose in England until emancipation was decreed for August 1, 1834. The emancipation of enslaved laborers through­out the British Caribbean did not mean a shift away from sugar society in Antigua but rather led to increased attention in maintaining the social relations of power in a free society.

Shifting to Emancipation: The Antigua Model Despite the promise of free­dom that accompanied the shift to emancipation in 1834, there was great anxiety among Antiguan planters as to how Anti-

Changes in Afro-Antiguan Lifeways / 135

guan society would operate following emancipation. Therefore, in the years prior to emancipation, Antiguan planters endorsed a series of legislative acts restricting the rights of laborers by prohibiting Sunday markets, limiting how laborers could appeal for missed wages, and creating strict laws relating to social conduct (Antigua and Barbuda National Archives, the Laws of Antigua). These measures were passed to help ensure that the formerly enslaved would continue to work on sugar plantations following emancipation. In large part, the continued success of the sugar enterprise on Antigua has been attributed to the unique sys­tem of strong coercion immediately following emancipation (Bolland 1996, 1999). Although Antiguan planters advocated immediate emancipation, the Antiguan legislature developed a series of contracts that legally bound laborers to estates for specific periods of time (Bolland 1996:109). Antigua is also credited with introducing the wage/rent sys­tem in the British Caribbean in which planters served as both employers and landlords as formerly enslaved individuals continued to live in cabins inhabited during slavery and work provision grounds on estates (Bolland 1996:109). The Antiguan legislation was considered so harsh that colonial overseers in Lon­don overruled the Antiguan legislation and enacted a series of less harsh (although only slightly) labor laws (Bolland 1996:109). Under these acts, oral labor contracts were only valid for a year and could be terminated at a month’s notice (Bolland 1996:109). In addition, under these changes, a tenement was eligible as legal evidence for a tenancy contract between the laborer and his landlord (Bolland 1996:109). Limits were set on the length of workdays, too, with laborers working nine hours a day for a wage of nine pence a day, with one day off every week or two (Bolland 1996:109). These new acts were an improvement of labor and living conditions over those origi­nally passed by the Antiguan colonial legislature, but wage laborers were still impoverished under this system, and it has been argued that the material conditions of life for formerly enslaved people would not have changed much immediately following emancipation (Bolland 1996:109).

The Martin Family and Green Castle Estate The Martin family was the model of British imperial success. Coming from a prominent Anglo-­Irish family, Colonel George Martin was instrumental in the British colonization of Suriname and settled there with his family seeking economic success in the middle of the seventeenth century. The family was forced to leave Suriname in 1677 following the Dutch invasion of the ­ colony, and George’s son, Major Samuel Martin, moved his family to Antigua where he established his sugar plantation, Green Castle Estate, in St.

136 / Bardoe

Mary’s Parish (Gaspar 1985:66). The 1678 Antigua census documents only five individuals at Green Castle, in­clud­ing one white man and one white woman (presumably Samuel Martin and his wife) and two black men and one black woman (Oliver 1894:1x). The origi­nal property Martin purchased was approximately 550 acres of land, but at the height of productivity in the eighteenth century Green Castle consisted of 600 acres, and at least 400 were dedicated to raising sugar (Gaspar 1985:96; Sheridan 1960:132). Major Samuel Martin was instrumental in the colonial government of Antigua as he was a member of the Antiguan colonial Assembly by 1686 and was made speaker of the Assembly in 1689. In 1697, he was listed as a member of the Antiguan Council (Oliver 1894:1xvi). Martin’s position in the local government may have contributed to the rapid success of his plantation. Green Castle was one of the estates that greatly profited in the last decades of the seventeenth century, as Martin quickly added to his landholdings, increased his enslaved population, and profited handsomely off the sugar trade. By the end of the seventeenth century, the estate boasted two windmills, highlighting the scale of production underway. Green Castle quickly arose as one of the largest and most profitable sugar plantations on the island, but Martin’s management skills were called into question when he was killed by his enslaved laborers on De­cem­ber 25, 1701 (Gaspar 1985:185). According to accounts, the laborers rebelled, killing Martin with their field hoes while he was still sleeping because he did not grant them their Christmas holiday (Gaspar 1985:185). Other accounts contend that his refusal to recognize the holiday marked the culmination of a series of harsh treatments Martin committed against his enslaved laborers (Dyde 2000:44). Still, other documents reveal the surprise of members of Antigua’s elite planter class at the event, commenting that Martin must have gravely mistreated his workers somehow because they came from the most genteel of the Af­ri­can tribes (Dyde 2000:44). Following the death of Martin, vari­ous managers oversaw the estate, in­ clud­ing relatives and friends of the Martins, as Martin’s eldest son finished his education abroad (Sheridan 1960:128). The estate would rebound from the murder of its founder under the management of his son, Colonel Samuel Martin, through­out the sec­ond half of the eighteenth century, yet his management of the plantation offers insights into the extent of manipulation and coercion that had to be employed to maintain seeming stability within the plantation. By the middle of the eighteenth century nearly all the arable land on Antigua was used for sugar production. With further land expansion impossible, the only way to increase production was to intensify the output from in­di­ vidual plantations through improved agricultural techniques. Indeed, it is

Changes in Afro-Antiguan Lifeways / 137

during this time that Colonel Martin (1750) began his Essay upon Plantership, an agricultural treatise that became popu­lar in Antigua and through­ out the Leewards in the eighteenth century; seven editions were published and the final edition was reprinted twice following Martin’s death (Martin 1765, 1773, 1785). Not only does Martin’s Essay espouse agricultural advice, it also speaks strongly to the social relations of sugar production during the eighteenth century. Sugar production continued at Green Castle Estate following the death of Martin, and the estate was inherited by his son, Henry Martin, who continued the systems of improvement his father began on the estate through­out the early nineteenth century (Murray 1825:261). The estate continued to operate through the middle of the twentieth century.

Excavations at Green Castle Estate The history of Green Castle Estate provides a fascinating backdrop for a diachronic study of the lives of Afro-­Antiguans during the periods of slavery and free­dom because of the social struggles that occurred at the site. At Green Castle Estate, there are distinct areas of domestic refuse dating to the periods of slavery and emancipation; these allow for a comparative approach that examines how lifeways changed for Afro-­Antiguans from the period of slavery to free­dom (Figure 7.1). This is an innovation for plantation archaeology in Antigua because it provides evidence of the daily lives of enslaved laborers and wage laborers. Previous plantation research on Antigua has only examined the lives of wage laborers in the late nineteenth century (Gonzalez Scollard 2008) or focused on the planter’s house (Fox 2009). Archaeological fieldwork at Green Castle Estate was conducted between 2007 and 2009. An initial pedestrian survey of the site resulted in the designation of seven loci within the area of potential ar­chaeo­logi­cal interest. These were identified using historic representations of the plantation, local knowledge of the estate, and the presence of surface features within the site. Archaeological loci that had the potential for making comparisons between the enslaved and wage laborer populations of the estate were given priority for subsurface testing. Initially, a shovel testing program was used to determine the presence, nature, and extent of subsurface deposits at Green Castle Estate. Loci 2, 4, and 5 were targeted in this program, and shovel tests in all three loci revealed the presence of intact ar­chaeo­logi­cal deposits. The deposits at Loci 2 and 4 were relatively shallow (between 20 and 40 cm deep), whereas at Locus 5 shovel testing revealed deposits as deep as 50–70 cm. Unit excavations were conducted in Loci 2 and 4 with the goals of identifying the presence of distinct house features, developing a greater sense of

138 / Bardoe

Figure 7.1. Green Castle survey map highlighting areas of ar­chaeo­logi­cal deposits. ­(Samantha Rebovich Bardoe)

the spatial distribution of artifacts within each locus, and beginning to collect data relating to the daily practices of enslaved and wage laborers at the estate. Six 1-­by-­1-­m units were excavated in Locus 2, resulting in scant ar­ chaeo­logi­cal deposits dating to the nineteenth century. This area had been targeted for ar­chaeo­logi­cal investigation because it was designated as the site of the slave village in an 1804 watercolor painting of the estate by Nicholas Pocock. The ar­chaeo­logi­cal evidence does not support this interpretation because the artifact assemblage dates to the late nineteenth century, and therefore excavations in Locus 2 were halted. Oral tradition identified Locus 4 as a place of domestic activity among wage laborers in the sec­ond half of the nineteenth century. Twenty units were excavated, and the artifact assemblage from this locus consistently dates to the sec­ond half of the nineteenth century. The ar­chaeo­logi­cal evidence suggests this area of the estate was used as domestic quarters for wage laborers in the postemancipation period. Unit excavations at Locus 5 were conducted because of distinct terracing along the hillside, possible building remains, and the presence of cultural deposits, as evidenced by shovel testing. A total of 20 units were excavated

Changes in Afro-Antiguan Lifeways / 139

along two terraced platforms of Locus 5, resulting in the recovery of an artifact assemblage that dates to the late eighteenth century and the identification of a single building feature. Excavations in Locus 4 and Locus 5 focused on obtaining a representative sample of the ar­chaeo­logi­cal deposits by sampling across the area of each locus. In this discussion of the ar­chaeo­logi­cal assemblages from Green Castle Estate, I focus on the material recovered from unit excavations in Locus 4 and Locus 5. These excavations provide discrete ar­chaeo­logi­cal deposits that can be closely associated with the postemancipation lifeways and the en­ slaved laborer lifeways, respectively, on the plantation.

Daily Practice at Green Castle Estate A great strength of ar­chaeo­logi­cal research is the ability to examine the seem­ ingly mundane, to find meaning in the quotidian, and to highlight the importance of daily practice (Dietler 1998:289; Silliman 2001:384; Stahl 2001:17). It is at the interstices of material culture and daily activities that interpretations of past lives, cultural practices, and meaning come to the forefront of ar­chaeo­logi­cal study. Archaeologists have drawn heavily on theories of practice and agency, in­clud­ing the works of Roy Bhaskar (1979), Pierre Bourdieu (1977), Michel de Certeau (1984), and Anthony Giddens (1984), to develop practice-­centered approaches to ar­chaeo­logi­cal interpretation. Within archaeology, many studies of­ten conflate agency and practice, examine how practice is a form of agency, and focus on a particular aspect of agency such as identity formation or resistance. This variety of approaches represents not just tensions in commitments to specific theoretical approaches but perhaps more so demonstrates the limits of ar­chaeo­logi­cal data in making such interpretations (Dobres and Robb 2000; Hodder 1982). In fact, the “turn to agency” in archaeology has come under criti­cism (Patterson 2005). Thomas Patterson (2005) argues that this trend in ar­chaeo­ logi­cal theory coincides with the emergence of neoliberalism in the late twentieth-­century global economy and reflects our own ideas of individualism rather than an attempt to understand people’s conceptions of themselves in the past. He argues that approaches to agency reflect neoliberal concepts of individualism, in­clud­ing (1) individuals’ pursuit of their own interests, (2) the reliance on information to make rational decisions, (3) the exercise of choice in the market, and (4) that social relations are the consequences of in­di­vidual goals (Patterson 2005:376). Patterson’s point is that studies of agency are more reflective of recent ideals and models of human action, and that these draw from a specific po­liti­cal stance. He does not reject an agency approach in archaeology but instead calls for a more criti­

140 / Bardoe

cal examination of the contexts in which theories emerge and what implications these might have on our understandings of the past. One of the ways archaeologists might overcome such obstacles is to criti­cally examine human practice by contextualizing it more deeply within the social relations of the past. In doing so, archaeologists can better assess to what extent past activities are agency-­driven. For instance, archaeologists should consider the whole spectrum of daily activities in which people ­engage—­ both those that are voluntary and self-­motivated and those, such as forced labor, that are coerced and externally imposed. Within plantation contexts, labor schedules and regimes were strictly imposed on laborers, and an understanding of how those activities structured daily life is necessary to examine how practice is affected in the dialectic of structure and agency. Practice should also be approached with a consideration for how different activities take on meaning within specific cultural and his­tori­cal contexts. For instance, Jeanne Arnold (2000:14) argues that daily activities intersect to create ideologies, reproduce social relations, and maintain the social relations of power, and that all of these variables affect the lives of past peoples. In examining the shift from kin-­based economies to hierarchically based po­ liti­cal chiefdoms, Arnold (2000) examines how labor relations cut across different areas of people’s lives by arguing that “the social properties of labor relations and their important ideological, kinship, and po­liti­cal ramifications are accorded equal standing with their more obvious economic properties” (14). Arnold contends it is exactly the noneconomic properties of labor relations that allowed for the manipulation of social networks to reshift society to a hierarchically based sys­tem along the North Ameri­can Pacific Coast. Stephen Silliman (2001) takes a similar holistic and practice-­centered ap­proach to the interpretation of the processes of labor and daily practice among Native Ameri­cans in California: “To complement the existing perspectives on labor and colonialism, I submit an approach that takes into account not only the form and implementation of labor but also the small-­scale daily activities of individuals negotiating, appropriating, living in, and suffering through particular labor regimes. That is, I conceive of labor as practice” (381). Silliman is interested in the relations of labor and how people could use their labor to manipulate their social positions. Silliman conceives of labor as a form of daily practice and argues that past actors were strongly aware of the value of their labor-­power and thus appropriated it for their own use in domestic contexts and in maintaining indigenous cultural practices. While this is problematic in that it tends to draw on a rather idiosyncratic understanding of labor, Silliman points to ways in which Native Ameri­cans were able to use their labor-­power for their own benefit, outside the realm of the Spanish missions.

Changes in Afro-Antiguan Lifeways / 141

Silliman’s approach to the social relations of labor and its connections to practice and agency provide an interesting point of departure for an ar­ chaeo­logi­cal investigation of enslaved laborer lifeways at Green Castle Estate. First, labor relations and the processes of labor played an integral role in the lives of enslaved individuals as their daily schedules were, in many ways, dictated by the organization of labor and seasonal cycles of sugar production. Highlighting the nature of labor as a crucial component within the context of slavery more deeply contextualizes the daily practices of past actors. Particularly in the plantation context, the processes of daily practice and domestic activities must be interpreted in light of labor cycles and must consider how past lives were shaped by these multiple factors. In examining the material culture of enslaved and free laborers at Green Castle Estate, I highlight how the multiple roles and responsibilities of laborers influenced their daily practices and emphasize how people carved a unique niche for themselves within the regimes of sugar and slavery in Antigua.

Archaeological Evidence of Daily Life at Green Castle Estate The ar­chaeo­logi­cal deposits at Green Castle Estate demonstrate ways in which the daily lives of enslaved laborers differed from those of wage laborers and point to how the daily lives of laborers changed in the years following emancipation. Formerly enslaved individuals took advantage of the new opportunities available to them, such as greater access to markets, increased autonomy in their domestic lives, and more educational opportunities. Archaeological evidence of daily practice demonstrates how Afro-­Antiguans actively changed their lives in the postemancipation period in spite of continued coercion and restraint by the colonial legislature. Artifacts relating to foodways are ubiquitous in domestic contexts, and the ar­chaeo­logi­cal assemblages from Loci 5 and 4 at Green Castle Estate are no exception. These include faunal remains as well as vessels relating to food storage, preparation, and consumption made from a variety of materials such as ceramics, glass, and metal. Items relating to foodways comprise 77.91% (n = 3,484) of the Locus 5 (period of slavery) assemblage and 70% (n = 1,612) of the Locus 4 (postemancipation) assemblage. The differences in these assemblages help reveal how the lives of free laborers changed from those of enslaved laborers. There are differences in the distribution of items relating to foodways, as shown in Table 7.1. Different vessels, such as ceramic jar fragments, European plates, and locally produced Afro-­Antiguan pot fragments, distinguish these food storage categories. For instance, items relating to food preparation are an overwhelming 51.55% (n = 1,796) of the foodways artifacts from

142 / Bardoe

Table 7.1. Comparison of Foodways Artifacts between Locus 5 during Slavery and Locus 4 Postemancipation Locus 5 (Period of Slavery)

Food-Faunal Food Storage Food Serving Food Preparation

Locus 4 (Postemancipation)

Number of Artifacts

(%)

Number of Artifacts

(%)

148 310 1,230 1,796

 4.25  8.90 35.30 51.55

220 524 751 117

13.65 32.51 46.59  7.26

Locus 5 (period of slavery) and only 7.26% (n = 117) of the foodways artifacts from Locus 4 (postemancipation). This suggests a substantial shift in Afro-­Antiguan food preparation from the period of slavery to after emancipation, when there is a drastic change in the number of artifacts relating to food storage. Food storage items only compose 8.90% (n = 310) of the foodways artifacts from Locus 5, but they compose 32.51% (n = 524) of foodways artifacts from Locus 4. The greatest shift in foodways artifacts between Locus 5 (period of slavery) and Locus 4 (postemancipation) is from an emphasis on food preparation during slavery to food storage following emancipation. The propensity of Afro-­Antiguan ceramics, largely used as cooking pots, accounts for the emphasis on food preparation in the Locus 5 assemblage. Afro-­Antiguan wares are hand-­built, low-­fired, and locally produced earthenwares that were made and used by people of Af­ri­can descent through­out the his­tori­cal period (Ahlman et al., chapter 4 herein; Hauser and Handler 2009; Nicholson 1990; Petersen et al. 1999). Rim sherds comprise 9.53% (n = 173) of the Afro-­ Antiguan ceramic assemblage from Locus 5. As 89 everted rims suggest, the most common vessel form is a bowl with a restricted neck, most probably used as a cooking pot. Three everted rim sherds demonstrate vessels with neck diameters of 12.0 cm or less, and it is most likely these represent restricted jars used for water storage. Vessels with neck diameters ranging from 14.0 to 22.0 cm are larger, open-­mouthed vessels probably used in cooking. Evidence of postmanufacture burning lends additional evidence to the use of Afro-­Antigua ceramics as cooking vessels. Afro-­Antiguans are described as “making pots for boiling their victuals, yubbas (or frying-­pans), water-­jars, and several other utensils” (Lanaghan 1844:4; emphasis in origi­nal). Traditionally, Afro-­Antiguans preferred this pottery because it allowed them to prepare foods in traditional ways as well.

Changes in Afro-Antiguan Lifeways / 143

In their examination of slave foodways at Clifton plantation in the Bahamas, Wilkie and Farnsworth (2005) argue, “Foodways are simultaneously one of the most pliable and most conservative of cultural practices. New ingredients and techniques can be quickly incorporated to suit differing access to resources, yet meal structure, preparation, and service can remain remarkably intact” (111). In this light, enslaved laborers at Green Castle could have drawn on traditional cooking techniques and incorporated local provisions and foodstuffs to create new, creole dishes. In contrast to the assemblage at Locus 5, only 7.26% (n = 117) of the postemancipation artifact assemblage at Locus 4 consists of artifacts relating to food preparation. Like the Locus 5 assemblage, though, food preparation artifacts consist almost exclusively of Afro-­Antiguan ceramics, in­clud­ing one ceramic sherd from a coal pot. In the late nineteenth century, ceramic coal pots became a popu­lar Afro-­Antiguan style vessel (Handler 1964:151; Hauser 2008:113; Heath 1988:114). Coal pots are used as portable hearths on which coals are burned and cooking pots are placed directly into the hot coals. Coal pots similarly became common in late nineteenth-­century contexts and were used well into the twentieth century through­out the West Indies (Hauser 2008:140). The increasing use of coal pots beginning in the sec­ond half of the nineteenth century suggests a shift in Antiguan pottery traditions and a change in cooking techniques. Rather than using hearths that were embedded in the landscape of the village, the use of coal pots allowed for the movement of cooking activities through­out the landscape. Ultimately, the nature of food preparation changed following emancipation, although the relative of lack of ar­chaeo­logi­cal data relating to food preparation does not suggest that it became a less important cultural practice. While there is a considerable decrease in the percentage of food preparation artifacts between the Locus 5 and Locus 4 assemblages, there is a vast increase in the number of artifacts relating to food storage between the periods of slavery and free­dom. Artifacts relating to food storage comprise 8.90% (n = 310) of the food-­related artifacts from Locus 5. Most of these are dark olive bottle glass fragments (86.77%, n = 267) that were regularly reused for beverage storage, especially for water. Water is of­ten scarce in Antigua, and while plantations had large cisterns to supply their residents, enslaved laborers probably saved vessels to store additional water in their homes (Gonzalez Scollard 2008:83). A relatively small number of ceramic vessels were used for food storage. Fragments of stoneware beakers and storage jars, and slipware pitchers, make up only 13.23% (n = 41) of the food storage assemblage from Locus 5. In contrast, a large proportion of items relating to foodways from Locus 4 consist of storage items such as jugs, pitchers, storage jars, and a variety

144 / Bardoe

Table 7.2. Items Relating to Food Storage from Locus 4 Excavations Number of Fragments

Artifact Type

Vessel Type

Ceramic

Bottle Jug Lid Mug/Can Pitcher/Ewer Storage jar Unidentified utilitarian

  1   2   1   7   1   1  35

Glass

Bell jar Bottle, beer Bottle, case Bottle, liquor Bottle, mineral Unidentified bottle Bottle, wine style Unidentified container

  1  59   1   7  75 221 106   6

of glass jars and bottles (Table 7.2). These items compose 32.51% (n = 524) of the foodways assemblage at Locus 4 and include coarse earthenwares, stonewares, and glass bottles in a variety of shapes and sizes, such as olive green “wine-­style” bottles, case bottles, and bell jars (Figure 7.2). These shifts in foodways point to several changes in the lives of free laborers that occurred following emancipation. First, the artifacts relating to foodways reflect a much greater use of imported European ceramics such as pearlwares, whitewares, and stonewares. While this shift might reflect a greater availability and access to imported ceramics, the nature of these ceramics points to a change in the practices surrounding foodways. Second, the number of items relating to food serving and storage suggests a greater interest in these aspects of foodways as opposed to food preparation. Several changes in laborer lifeways following emancipation lend insights into why there is an emphasis on food serving and storage. Freed wage laborers could no longer rely on receiving provisions from planters as they had during the period of slavery. Although at Green Castle Estate enslaved laborers had access to provision grounds, the greater presence of artifacts relating to food storage following emancipation suggests a greater concern with preserving foodstuffs. His­tori­cal accounts also stress how formerly enslaved individuals were responsible for growing their own foodstuffs follow-

Changes in Afro-Antiguan Lifeways / 145

Figure 7.2. Storage vessel fragments from salt-­glazed stoneware sherds. (Nigel B ­ ardoe)

ing emancipation: “Now, they provide themselves with what they like; and are therefore better, if less abundantly fed” (Sturge and Harvey 1838:47– 48). Even if wage laborers were given the opportunity to grow food items they preferred, the lack of food rations received from planters was a problem that some free laborers could not overcome. At least one postemancipation court case highlights how a laborer resorted to stealing food because he had been sick and could not earn his wages, grow food items, or afford to purchase food (Sturge and Harvey 1838:35). Additionally, the challenge of providing food for elderly Afro-­Antiguans who could not work to earn a wage was a widespread social problem following emancipation: “The most painful feature in the state of Antigua at the present moment is, the destitute condition of the old and infirm, owing to the absence of a legal provision for them, and to the present distress from the long period of drought” (Sturge and Harvey 1838:33). It becomes clear that following emancipation, as laborers became responsible for providing themselves with food, a greater in-

146 / Bardoe

terest in food storage was necessary, and this is evident at Green Castle Estate by the higher occurrence of food storage vessels in the Locus 4 assemblage. Although a decline in food preparation in the postemancipation era is evidenced by only 7.26% (n = 117) of the foodways artifacts being used for food preparation, this does not mean that the importance of food preparation was undermined following emancipation. Wage laborers may have shifted the kinds of vessels they used for cooking, such as turning to cast iron pots instead of locally produced Afro-­Antiguan wares, yet still maintaining a strong tradition of food preparation. In fact, following emancipation, a greater appreciation for foodways in general may have been fostered, as it became an aspect of life now completely controlled by formerly enslaved individuals.

Conclusions The shift from slavery to emancipation in Antigua served as an experiment for the rest of the Caribbean. Foregoing the period of apprenticeship that other islands enacted, Antigua’s switch to full emancipation brought with it swift changes to Antiguan legislation to continually coerce the newly freed laborers. However, the ar­chaeo­logi­cal evidence points to ways that formerly enslaved people were able to adjust to the new conditions of life in Antigua and forge new lifeways for themselves. While the slave period assemblage from Green Castle Estate consists mainly of items relating to food preparation, a great shift occurs in the postemancipation assemblage to a greater presence of items relating to food storage. This shift coincides with the lack of food provisions planters provided in the postemancipation period, and the his­tori­cal evidence also points to widespread famine and poverty among formerly enslaved people. Despite these challenges, Afro-­Antiguans continually worked to improve their lives within the still-­coercive structures of Antiguan society. His­tori­ cal evidence points to the ways that Afro-­Antiguans took advantage of new educational opportunities provided on the island. Archaeological evidence points to increased use of imported goods in foodways that suggests greater purchasing power and access to markets. Although the material conditions of life may not have shifted drastically following emancipation, ar­chaeo­logi­ cal evidence does suggest important shifts that speak to broader changes in people’s lives in this new period of Antiguan life. As is evident, the shifts in the landscapes of Antigua and the material culture of Afro-­Antiguans over the periods of slavery and free­dom on the island are not just reflective of changes in lifeways among laborers but also highlight how labor relations and the social relations of production and con-

Changes in Afro-Antiguan Lifeways / 147

sumption between planters and laborers changed. These shifts point to ways in which Antiguans created a postemancipation society that translated the rhetoric of free­dom into the practice of emancipation on the island. The ways in which emancipation unfolded in Antigua are unique within the context of the British Caribbean and reflect the complex negotiations of social life that accompanied this shift.

8 The Military and Institutional Occupations of Charles Fort, St. Kitts, West Indies Gerald F. Schroedl and Todd M. Ahlman

Charles Fort is located on the island of St. Kitts near the town of Sandy Point approximately 30 km north-­northwest of Basseterre and less than a km west-­ northwest of the Brimstone Hill fortress (Figure 8.1). Charles Fort has two significant periods of institutional occupation: as a military post from the 1670s until 1854 and then as a leper asylum from 1890 to 1996 (Schroedl 2000a). A third use as a heritage tourism site that is now emerging indicates how important Charles Fort is to the people of St. Kitts and the Caribbean for its military history and for its postcolonial social history (see also Edward González-­Tennant and Diana González-­Tennant, chapter 9, this volume). The British constructed Charles Fort in the 1670s and 1680s as an irregular quadrilateral, enclosing approximately 1.7 ha, to protect the Sandy Point harbor. Known also as Cleverley Hill fort (see Calendar of State Papers America and the West Indies [CSPAWI] 1678, 5:599, 620, 763, 890), Charles Fort was the largest and most important fortification on St. Kitts until the mid-­eighteenth century when it was superseded by Brimstone Hill. Although Brimstone Hill was greatly expanded after 1783, Charles Fort was retained as part of the garrison well into the nineteenth century (see Schroedl 2014; Smith 1994, 1995). A 1723 site plan (Smith 1994) shows seven structures within the fort (Fig­ ure 8.2), only four of which are clearly identifiable today. So far no military buildings known to have been constructed after this date have been positively identified. The south and west defensive walls and the southeast and southwest bastions show considerable alteration, and the placement of the embrasures was changed on all the bastions after 1723 (Schroedl 2000a). The sec­ond period of occupation, established nearly 40 years after the fort closed, is its adaptive reuse as a leper asylum from 1890 to 1996. Most structures within the fort date to this occupation. In 1920, the site also became known as Hansen Home in recognition of Gerhard Hansen, the physician who identified the bacteria that causes leprosy (Anderson 2005). Historic photographs show that stone foundations were used in the asylum’s initial development. Sixteen dormitories, each with two to five 10-­by-­12-­ft.

Military and Institutional Occupations / 149

Figure 8.1. Location of Charles Fort, St. Kitts, West Indies. (Gerald F. Schroedl)

rooms, were built on stone foundations in 1890–1891 (Boon 1890:5–6; 1891). Poured concrete construction may date to as early as the 1920s, and widespread use of concrete blocks is probably more common after the 1970s. It is also evident that in two instances more recent buildings were placed on the stone foundations of earlier hospital structures (Schroedl 2000a). When some structures became uninhabitable because of neglect or storm damage, they were not rebuilt or repaired. New construction was erected elsewhere or activities relating to the building’s use were relocated to another place. As the number of patients diminished so did the number of dormitories and cottages required to house them, so that by the 1980s and 1990s many houses were overgrown and in vari­ous stages of decay (Anderson 2005; Schroedl 2000b, 2000c). The administration buildings, kitchen, medical buildings, bathroom and laundry buildings, and a few houses for the last remaining patients were maintained until the asylum was closed in 1996. With the asylum’s impending closure, discussion began among St. Kitts government officials, cultural heritage interests, and private individuals regarding the potential inclusion of Charles Fort within the Brimstone Hill Fortress National Park. This dialogue was further intensified during the nomination process under which Brimstone Hill became a UNESCO World Heri­tage site in 2000 (Brimstone Hill Fortress 2000; St. Kitts Government 1998). The purpose of ar­chaeo­logi­cal and architectural research, sponsored

150 / Schroedl and Ahlman

Figure 8.2. Plan of Charles Fort in 1723 based on an origi­nal drawing in the collection of the Duke of Buccleuch. (From Victor T. C. Smith 1994:Figure 4.)

by the St. Kitts government, was to provide information about the organization and integrity of the site and to help pub­lic officials decide how best to use it, while preserving and protecting its his­tori­cal character as an eighteenth-­century fortress and a twentieth-­century hospital complex. To provide this information, a detailed site map was made (Figure 8.3); each structure and major architectural feature was described and photographed; ar­chaeo­logi­cal test excavations were conducted in five locations; and information about the fort’s use as a leper hospital and military post was gathered from archival sources and local informants. In 2000, the effects of four years of abandonment on the site’s buildings already were evident, and local residents had begun to repurpose the site for agriculture, livestock, and charcoal production and as a source of building materials (Schroedl 2000a, 2000b, 2000c). Further deterioration to the site continued, and some buildings or ruins were no longer standing when the site was visited in subsequent years. In this chapter, we describe the buildings as they appeared in 2000. In 2012, the St. Kitts government approved a plan incorporating Charles

Military and Institutional Occupations / 151

Figure 8.3. Plan of Charles Fort, showing the locations of structures (ST) and features mapped in May 2000. Data collected with total station arbitrary metric grid. (Todd M. Ahlman)

Fort as part of the St. Christopher National Trust for management under the Brimstone Hill Fortress National Park (2012). In the late nineteenth and early twentieth centuries, leper asylums were established on many Caribbean islands, in­clud­ing Trinidad, Barbados, Domi­ nica, Antigua, Cuba, Puerto Rico, St. Croix, St. Eustatius, St. Thomas, ­Jamaica, St. Lucia, and Guadeloupe (see Abraham 1889; Brizan 1984; Colonial Office 1930; Fox 1933; Muir 1942; Ransdell 1916; Verrill 1917). Except for Puerto Rico (Schiappacasse 2011); St. Eustatius (Gilmore 2008); St.

152 / Schroedl and Ahlman

Kitts (Anderson 2005; Schroedl 2000a; Schroedl and Ahlman 2001); and St. Thomas (Barton 2012), few have received much his­tori­cal or ar­chaeo­logi­cal attention. Leper hospitals were integral to the broader concerns for pub­lic health, welfare, education, imprisonment, and hospitalization that each island addressed in the postemancipation, colonial, and postcolonial eras. In St. Kitts, for example, a newspaper, the Lazaretto, published from 1890 to 1892, was dedicated solely to the etiology and epidemiology of leprosy. For these reasons, the leper asylum occupation at Charles Fort is especially relevant to emerging studies of the archaeology of institutions and more specifically to the archaeology of confinement. These studies center on the interrogation of the social power of domination that state institutions employ to restrict the actions or confiscate the rights of individuals. Of equal importance is the power of resistance that individuals exercise to negotiate the character of their confinement (see Beisaw and Gibb 2009; Baugher 2009; Casella 2007; De Cunzo 2006, 2009). At Charles Fort consideration of the landscape, architecture, furnishings, and foods are among the elements that reflect development and changes in institutional goals and purposes of the St. Kitts government and provide the contexts in which asylum patients navi­gated their confinement. The military architecture is essential to these relationships because it formed the physical and psychological skeleton for organizing and constructing the asylum. The hospital built within this space served a full range of institutional needs, in­clud­ing medical treatment, patient and staff housing, patient hygiene, recreation and religious services, kitchen and dining areas, and cemetery grounds.

Military Occupation of Charles Fort The British settled St. Kitts in 1624 and were soon joined by the French in 1625 (Dyde 2005:21–23). The French held the island’s north and south ends and the British controlled the island’s center. The north­ern frontier was at Sandy Point, the largest British settlement on the island. Despite attempts to maintain peaceful relations, the two sides fought for control of the island during the Second Dutch War (1665–1667), King William’s War (1688– 1697), and Queen Anne’s War (1701–1713) (Marley 1998, also see Watts 1987:Table 6.3). These conflicts caused considerable property loss as well as death and displacement of the island’s population, as each side gained control of the island, only to reestablish the origi­nal boundaries when each conflict ended (Dyde 2005; Hubbard 2002; Burns 1954; Marley 1998; Crouse 1943). The Treaty of Utrecht in 1713 gave final and complete control of the island to the British (Dyde 2005). Following French destruction of English fortifications during the Second

Military and Institutional Occupations / 153

Dutch War in 1667, the British dismantled their gun platform near the north­ ern French border and began construction of Charles Fort around 1674 to guard Sandy Point and its harbor (CSPAWI 1674, 3:1201). Charles Fort became the primary focus for defending the island, and significant effort was made to finish and maintain it well into the eighteenth century (CSPAWI 1678, 5:599). The St. Kitts legislature regularly used plantation slaves for this purpose (St. Kitts Council and Assembly 1791; St. Kitts National Archives [SKNA] St. Christopher Council Minutes 1735; The National Archive [TNA], Lon­don, England, CO241/6, St. Christopher Assembly Minutes, 13 February 1745). Much of the fort was enclosed by the late 1680s, but efforts to finish bastions at the four corners and a fifth to form a pentagon were still underway in the mid-­1730s (CSPAWI 1734, 41:314ii). When completed, the fort covered about 1.7 ha and was surrounded by a moat. It was complete enough that Colonel Thomas Hill and about 400 men retreated there when the French invaded the island in July 1689, holding out for six weeks until surrendering in August (TNA CO152/20, William Matthew, State of His Majesty’s Leeward Caribe Islands in America, 1734). In July 1690, English forces retook St. Kitts and attacked Charles Fort, then occupied by 480 French soldiers. British cannon fire from Brimstone Hill helped retake the fort in less than a week (Dyde 2005:70; St. Johnston 1980:5–6; Marley 1998:204). Military engagements with the French in 1690, 1706, 1712, and 1782 precipitated repairs and renovations of the fortress walls and other features. An unspecified number of houses, for example, were destroyed within the fort during the 1690 siege (CSPAWI 8:291–294:977), and between 1706 and 1720, both frame and stone construction was used to build, renovate, or repair a storehouse, guardhouse, barracks, prison, arsenal, officers’ quarters, and a well. By 1723 a sally port, arsenal, cistern, magazine, a guardroom, and two unidentified structures were present within the fort, and within a year the fort was armed with 27 cannons (TNA CO152/4, Queries proposed by the Right Honorable the Lords Commissioners for Trade and Plantations to John Hart Esquire His Majesties Captain General and Governor of the Leeward Carribbee Islands in America and his Answers thereto; Humbly Offered to their Lordships Consideration, 10 July 1724). In 1734 Charles Fort had “a small cistern, a house of two rooms, one of which serves for a little arsenal, in which are thirty one small arms, in but indifferent order, a good stone kitchen, two large ovens, a small guard room, a small prison for the soldiers, two rooms for the gunner, and a pretty large magazine arched to be bomb proof as all the magazines on the island” (CSPAWI 1734, 41:314ii). Additional construction and renovations continued through the 1750s, when, for example, in 1751 it was recom-

154 / Schroedl and Ahlman

mended to spend £1200 to construct sentry boxes and to repair the officer’s guard house, sick room, cook room as well as the prison and the well (TNA CO241/6, Assembly Minutes, Report on Fortifications, 31 May 1751). Simi­ lar repairs were suggested in 1771, 1773, 1775, and 1778 (TNA CO241/14, Assembly Minutes, Report of the Committee to Inspect the State of Fortifications, 19 March 1771; TNA CO241/16, Assembly Minutes, Report of the Committee to Inspect the State of Fortifications, 25 No­vem­ber 1773; TNA CO241/16, Assembly Minutes, Report of the Committee to Inspect the State of Fortifications 18 De­cem­ber 1775; TNA CO241/17, Assembly Minutes, Report of the Committee to Inspect the State of Fortifications, 12 June 1778). A proposal in 1779 called for building a hospital “upon old walls now standing” and to build a roof “upon the wall of the old kitchen” (TNA CO241/17, Assembly Minutes, 11 August 1779) and to complete construction of a blacksmith’s shop, four additional sentry boxes, and a shade for the carpenter’s shop. A gunner and up to eight matrosses were stationed at the fort, and there was room to house about 30 to 100 men (CSPAWI 1685, 7:388; CSPAWI 1706, 18:195i, 341; CSPAWI 1720, 27:251; TNA CO241/3, Assembly Minutes, 26 Sep­tem­ber 1730; TNA CO241/4, Assembly Minutes, 20 June 1740; TNA CO155/28, A Report of the Condition of his Majesties Garrison of Brimstone Hill in the Island of St. Christopher’s with Ordnance, Ammunition and Other Warlike Stores Belong­ing to it, 15 De­cem­ber 1755; CO241/7, Assembly Minutes, Report on Forti­fi­cations, 14 April 1759; TNA CO241/10, Assembly Minutes, 26 April 1764; TNA CO241/12, Assembly Minutes, 28 June 1770, 6 May 1772; TNA CO241/ 16, Assembly Minutes, Report on Fortifications, 25 No­vem­ber 1773; TNA CO241/17, Assembly Minutes, 9 Sep­tem­ber 1778). Following the French invasion and occupation of 1782–1783, the British on St. Kitts continued to maintain Charles Fort but devoted most of their energy and resources to enlarging and garrisoning Brimstone Hill. Proposals after 1784 were made, for example, to construct a hospital and to repair buildings for use as a barracks. In 1813 Charles Fort came under the control of the colonial government, although full transfer was not completed until 1844 (TNA CO241/18; James Lees to the Assembly, 31 May 1785; TNA WO44/334, G. Juxon to Board of Ordnance, 10 Sep­tem­ber 1813; TNA WO44/474, N. Nugent and R. Wills to Board of Ordnance, 10 May 1844). By 1838 it was reported that all the forts and batteries on St. Kitts, except Brimstone Hill, which remained under the Ordnance Department, were all more or less in ruins (TNA WO44/448, Office of Ordnance, Forts and Batteries on St. Christopher, 30 Janu­ary 1838). In 1839, the colonial government passed a resolution to use Charles Fort as a poor house, and in 1846 the idea of using Charles Fort as a lunatic asylum was suggested (TNA CO241/30, Assembly Minutes, 21 March 1839; TNA CO241/34,

Military and Institutional Occupations / 155

Rules for the Lunatic Asylum, August 1846). By 1854, when Brimstone Hill was closed, the colonial government took control of the property and declared that this would allow “the long contemplated design of the fitting up of Charles Fort for a Lazerato” (TNA CO241/36, Assembly Minutes, 12 Oc­to­ber 1854). Not until May 1889 did a committee appointed by the governor specifically recommend establishing a leper asylum at Charles Fort, and legislation for doing so was approved in July 1890 (TNA CO241/43, TNA CO241/43, Bill to Create a Leper Asylum, 2 July 1890).

Archeological Assessment of the Military Occupation Comparison with the 1723 plan shows that the bastions now exhibit changes in the location of the gun embrasures (Schroedl 2000a). For example, where the northeast bastion once had four embrasures, there are now single embrasures. There is now a single embrasure in the north bastion wall and two in the east wall where none is shown in 1723. The southeast bastion shows similar differences. The northwest bastion has three embrasures on the north-­facing wall, where the 1723 plan shows none, and there is a single embrasure on the east-­facing wall where the 1723 plan shows two. This is consistent with a 1778 recommendation for adding two north-­facing embrasures to this bastion (TNA CO241/17, Assembly Minutes, Report on Fortifications, 12 June 1778, 1 Sep­tem­ber 1778). The seaward-­facing portion of this bastion has been completely lost to erosion, as has most of the south­west bastion. The defensive walls also show modifications (Schroedl 2000a). The southeast bastion wall and an undetermined portion of the connecting south defensive wall were nearly doubled in width at some time in the fort’s history. The interior of the south wall, in addition, is braced with a series of seven short buttresses. The west, or seaward-­facing, defensive wall, as shown on the 1723 plan, has a slightly irregular outward curve, while a 1782 siege map exhibits a distinctive angle at the wall’s midpoint that remains evident today (Library of Congress 1782). This is consistent with the fort’s description as a pentagon and with a probable firing platform constructed at this juncture in 1734 (CSPAWI 1734, 41:314ii). Alterations to the defensive walls and bastions are compatible with regular reports through­out the eighteenth and nineteenth centuries calling for their rebuilding and repair (TNA CO241/9, Assembly Minutes, Report on Fortifications, 4 June 1764; TNA WO55/945, J. C. Smyth to Sir Charles Smith, 28 March 1826). Besides the bastions and connecting curtain walls, recognizable features associated with eighteenth-­and nineteenth-­century military occupation are two guardhouses or prisons (Structures 5 and 6), the sally port (Structure 29), the arsenal (Structure 32), and the cistern (Structure 35). Twenty 1-­by-­1-­m

156 / Schroedl and Ahlman

test units were excavated to further document these features. In the northeast bastion six pits revealed a stone floor where a later structure, possibly associated with the leper asylum, was built. Two contiguous excavation units also determined that a similar stone floor covered the southeast bastion’s interior. These floors replaced the grenadier’s shown in the fort’s 1723 plan and may be associated with the 1778 proposal to reconfig­ure the embrasures. A foundation where the fort’s arsenal is shown on the 1723 map suggested that this might be the building’s location. A 120-­cm-­deep excavation revealed six distinctive deposits. The upper 70 cm of fill represent nineteenthand twentieth-­century deposition. Below this were ash-­and charcoal-­laden sediments covering a mortared floor and a stone support pier or partition. These lower deposits contain artifacts dating to the fort’s eighteenth-­century military occupation. The large cistern located on the north­ern side of the fortress is one of the origi­nal structures associated with Charles Fort, although it was used and probably rebuilt or repaired in the nineteenth and twentieth centuries. Artifacts recovered from the excavation date to the mid-­twentieth century when the cistern was abandoned or needed less of­ten once piped water was introduced to the leper asylum perhaps as early as the 1890s but certainly by the 1920s (TNA CO241/43, Committee to Select a Locality for a Leper Asy­lum, 18 May 1889; SKNA Records of the Public Works Department, April 1924). A defensive ditch once surrounded the fort on the north, south, and east sides but is no longer visible at the surface. A single test pit 170 cm deep was placed near the east defensive wall and entrance gate to investigate its deposits. Artifacts suggest that most of the fill accumulated since the mid-­ nineteenth century may represent debris from squatters’ huts built in the 1840s (TNA WO44/474, T. Clark to A. Bynum, 6 August 1845, TNA MPH 1/1126/1, St. Christopher Plan of Forts and Batteries, 1845). Excavations halted when an adult human burial was encountered in a pit that originated circa 140 cm below the present surface. The bones were left in place and the individual’s age and cultural affiliation are undetermined. It is unlikely, from its depth and location, that this is a European soldier, an enslaved Af­ ri­can, or even a leprosy patient, but instead a person interred between the time the fort was largely abandoned (ca. 1838) and the leper asylum’s establishment (1890).

Leper Asylum 1890 to 1996 The leper asylum occupation at Charles Fort is represented by 54 distinctive buildings, structures, or walls. Six buildings, undetected during mapping, are seen in an 1893 historic photograph or air photographs taken in

Military and Institutional Occupations / 157

1946 (United States Air Force [USAF]), 1954 (United States Navy [USN]), 1956, 1968, or 1982 (Directorate of Overseas Surveys [DOS]). The asylum is generally organized into (1) an administrative and food preparation area, (2) a patient treatment and hygiene area, (3) separate areas to house male and female patients, and (4) a cemetery and church area (Tables 8.1 and 8.2). This arrangement as shown in Fig­ures 8.3 and 8.4 was established during the initial construction and maintained through­out the hospital’s history. The administrative and treatment buildings were constructed in place and some patient housing was prefabricated and imported from an undetermined manufacturer in the United States in the 1890s (TNA CO241/43, Assembly Minutes, 22 April 1890). Some were built as one-­or two-­room cottages and others, as shown in Figure 8.5, were built as multiroom dormitories. The interior grounds were landscaped as shown so dramatically in a 1920s-­era postcard (Fig­ure 8.6). A fence divided female and male living areas, and an additional security fence on top of the fortress walls enclosed the men’s living area (see Figures 8.3 and 8.4). The administrative and food service area consists of the kitchen (Structure 4), storage facilities using the former guardhouses (Structures 5 and 6), and administrative facilities and staff quarters located at the fort’s east side and entrance (Structures 7 and 11). The kitchen was built by 1893 and was rebuilt or renovated several times. The south and east walls were replaced and a wood frame ell added to the rear of the building, and a brick fireplace and chimney were later attached to the ell. Structures 5 and 6 (the origi­nal guardhouses) situated on either side of the fort’s main entrance exhibit a variety of frame and concrete modifications for conversion to food storage as well as the concrete bath installed in Structure 5. The administrative building (Structure 7) was positioned along the top of the defensive wall above the fort’s main entrance and the two storage buildings. It had two rooms for living quarters and one for an office. Stairways to the building were placed on either side of the storage rooms. Structure 11 was a nearly identical single-­ story structure built on a stone platform that housed orderlies and nurses. Concrete pillars support a sec­ond story addition to which was attached a concrete shower and toilet. Both structures were built before 1893. The patient treatment and hygiene area includes the assembly hall (Structure 8), clinic and pharmacy (Structures 9 and 10), toilet (Structure 24), two laundry and bath buildings (Structures 25 and 28), and septic tank (Structure 26). These structures and a steel post and wire mesh fence anchored in a continuous concrete footing separate the north or male side of the hospital from the south or female side. This fence was origi­nally two closely spaced parallel sections of sheet iron anchored at either end to a treatment building and picket fence (see Figure 8.4). The buildings are connected with a

Table 8.1. Description and Dimensions for Individually Numbered Structures at Charles Fort Structure Description  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

Dimensions (m) width × length

Stone foundation, probable patient housing 5.00 × 9.20 Stone foundation, probable patient housing 6.00 × undetermined Standing structure, patient detention 3.70 × 6.20 Standing structure, kitchen 7.10 × 10.00 Standing structure, guard house and asylum storage 5.80 × 6.85 Standing structure, guard house and asylum storage 5.90 × 6.85 Standing structure, staff housing and administrative offices 6.90 × 13.60 Standing structure, assembly and recreation hall 6.55 × 12.35 Standing structure, medical clinic 6.00 × 9.25 Standing structure, pharmacy 3.65 × 3.75* Ruin, hospital staff housing 4.60 × 12.25† Standing structure, patient housing 5.20 × 60 Standing structure, patient housing 5.10 × 6.25 Standing structure, patient housing 5.10 × 6.25 Standing structure, patient housing 5.10 × 6.25 Standing structure, patient housing 5.10 × 9.50 Standing structure, bathroom 2.20 × 2.45 Standing structure, patient housing 5.10 × 9.25 Stone foundation, probable patient housing 5.10 × 6.25 Stone foundation, probable patient housing 4.90 × 9.70 Stone foundation, probable patient housing 5.10 × 6.25 Stone foundation, probable patient housing 4.90 × 9.25 Standing structure, fountain 2.70 m diameter Standing structure, bathroom 2.85 × 3.70 Ruin, bath and laundry 3.75 × 9.40‡ 2.75 × 3.40 Standing structure, septic tank attached to Structure 24 Ruin, probable patient housing 4.90 × 15.40 Ruin, bath and laundry 3.75 × 7.75** Standing structure, sally port 2.85 × 4.60†† Stone foundation, probable patient housing 4.90 × 15.55 Stone foundation, probable patient housing 4.90 × 15.85 Stone foundation, probable arsenal undetermined Stone foundation, probable patient housing 5.20 × 15.80 Ruin, patient housing 4.90 × 6.50 Standing structure, cistern 5.25 × 14.50 Ruin, patient housing 5.00 × 12.50 Stone foundation, library 5.00 × 9.40 Standing structure, chapel building 6.00 × 12.00

*Foundation measures 3.65 × 9.20 m †Attached bath measures 2.15 × 3.40 m ‡Foundation measures 3.75 × 9.40 m **Foundation measures 3.75 × 14.20 m ††Width and height of opening

Military and Institutional Occupations / 159

Table 8.2. Description of Individually Numbered Walls at Charles Fort Wall

Description

 1

Wall connecting the northwest and southeast interior corners of the northeast bastion

 2

Wall in front of Structure 4 running from the northeast corner of Structure 8 to Structure 37

 3

Short wall connecting Structures 4 and 5

 4

Retaining wall in front of Structure 11

 5

Retaining wall connecting Structures 6 and 11

 6

Wall attached to the southwest interior corner of the southeast bastion

 7

Concrete foundation for the fence running from Structure 4 to the west defensive wall and dividing the male from female sides of the hospital grounds

 8

Concrete block wall separating the southwest bastion from the fort’s interior

 9

South cistern catchment retaining wall

10

North cistern catchment retaining wall

drain that runs from the kitchen (Structure 4) to the septic tank. When patient segregation was discontinued, the fence was removed and concrete steps were installed between the pharmacy and one of the baths for ease in crossing the fence’s footing. The assembly and recreation hall, built in 1926 (SKNA, Records of the Public Works Department, 26 March to 2 July 1926), was a wood building with a concrete floor and men’s and women’s entrances on either side. An interior wood rail dividing the building lengthwise further restricted patients from intermingling. Patients could purchase personal items here and listen to the radio or watch television. Some ate here or in their cottages. As residents grew older and more infirmed, church services also were conducted here (Anderson 2005). The clinic (Structure 9) was a concrete block structure divided into a nurse’s treatment and doctor’s examination rooms. The toilet (Structure 24) is a partitioned, poured-­concrete building with separate access for men and women. The pharmacy (Structure 10) was a small wood structure sitting on a tile floor slightly larger than the building’s footprint, and the laundry and bath is a cut stone and concrete building (Structure 25) erected on a stone foundation. These structures are where two probable treatment buildings stood in the 1890s (see Figure 8.4).

Figure 8.4. Panoramic photograph taken 1893, showing leper asylum, with six cottages for female patients to the left (left to right: unnumbered and Structures 18 through 22) and dormitories for men to the right (left to right: unnumbered cistern, Structures 27, 30, 31, and unnumbered housing unit at far right). Note two treatment buildings at center, security fences, pathways, and garden in front of Structure 27. (TNA CO152/187, Leeward Islands, St. Kitts and Nevis, Despatch No. 140, July 10, 1893.)

Figure 8.5. Patient housing, Structure 31, March 1899. Note the veranda, stairs, and entrances to in­di­vidual apartments. (C. C. Lyon, from the collection of Todd M. Ahlman.)

Figure 8.6. Postcard ca. 1920 showing structures in the northwest corner of the asylum (left to right: Structures 31, 33, and 34) and extensive landscaping and gardens. Roof at the lower right is the kitchen (Structure 4). (V. E. John, from the collection of Todd M. Ahlman.)

162 / Schroedl and Ahlman

Female patient housing was on the south side of the asylum grounds and includes four foundations and five standing structures of four nearly identical post and beam or concrete construction divided into two rooms with their own entrances from a concrete veranda (Structures 12, 13, 14, and 15). The fifth standing building (Structure 18) is 3 m longer and has three doorways, two entering the larger of two rooms. It was built after 1946 (USAF 1946; USN 1954). Three stone foundations (Structures 19, 20, and 21) consist of two or three identical bays, and the fourth (Structure 22) has a single bay. A patient cottage that once stood in the southeast bastion was replaced with two smaller houses before 1946 (Figure 8.4; USAF 1946). Both were demolished before 1954 (USN 1954) when a concrete block reinforcing wall (Wall 8) was constructed diagonally across the bastion. The barred windows indicate that Structure 16, built on an elevated stone platform, was a detention building. A concrete toilet (Structure 17) near Structure 18 was built in 1973 for the last few patients living at the hospital (Anderson 2005:183). A concrete fountain (Structure 23) was built in 1929 as part of the decorative landscaping that characterized the asylum grounds (SKNA Records of the Public Works Department, May 1929). The north or male half of the asylum compound consists of nine r­ uins or foundations (Structures 1, 2, 27, 30, 31, 33, 34, 36, and 37) and one standing building (Structure 3). Structures 29 (sally port), 32 (probable arsenal), and 35 (cistern) are associated with the fort’s military occupation. Two other structures, a cistern next to Structure 27 and a dormitory building in the northwest bastion were present in the 1890s (see Figure 8.4). Structures 1 and 2 are low stone foundations, and Structure 3 is a two-­room concrete detention building with barred windows located in the northeast bastion. Both men and women were held in Structure 3 after use of Structure 16 was discontinued (Anderson 2005:166). Structure 37, the library, sits on an elevated stone foundation, with three massive cut stone stairways ascending to the structure’s front porch and another set of steps connecting it to the cistern on one side. Its use as a library was discontinued in the 1950s, and reading materials were placed in Structure 5, one of the storage buildings (Anderson 2005:65–66). Structures 27, 30, 31, and 33, are comparable cut stone foundations with one or five bay floor plans and a continuous veranda on one side, with steps at both ends (see Figures 8.5 and 8.6). Structures 27, 30, and 31 and a dormitory in the northwest bastion were occupied in the 1890s (see Figure 8.4). Structure 34, a smaller single-­story house, and Structure 36, a larger two-­ story building, both used concrete piers and were certainly occupied circa 1920 (see Figure 8.6).

Military and Institutional Occupations / 163

To prevent male patients from escaping the asylum, a galvanized steel picket fence was constructed in 1928 on the men’s side of the compound along the top of the fort’s defensive wall from the middle of the west wall to the north wall’s east end (SKNA Records of the Public Works Department, De­cem­ber 1928). It does not enclose Structures 1 and 2, suggesting these buildings were not in use at this time, but curiously the detention building is not enclosed either. It is unknown whether an earlier security fence existed. The open areas between the service and administrative structures and men’s and women’s housing were used for walkways and gardens that included ornamental plants and flowers as well as space for fruit trees and vegetables. The landscaping that included the fountain (Structure 23) was organized with bold arrangements of white painted stones forming decorative borders enclosing floral and geometric patterns (see Figure 8.6). A barbed wire fence enclosing the asylum grounds extended 50 to perhaps 200 m beyond the fortress walls. Most of this area was used for farming, and the hospital chapel and cemetery were also located here. The masonry church or chapel (Structure 38) is located near the southeast bastion and exhibits a dedication stone dated 1926. It had a steeple and the interior was furnished with a pulpit, pews, and railings (Schroedl 2000a). The asylum’s cemetery is 30 to 50 m wide and about 100 m along the fortress’s south wall. No grave markers currently survive, so estimating the number of burials is difficult. A small building measuring circa 4 by 5 m and perhaps associated with mortuary activities is visible at the cemetery’s edge near the chapel in air photographs taken between 1946 and 1968 (USAF 1946; USN 1954; DOS 1956, 1968). The size of the cemetery as well as death and burial records from the 1890s to 1930s and 1940s to 1990s, although incomplete, suggest as many as 200 to 400 individuals. For example, there are 148 deaths reported for 1892 to 1916 and 57 burials recorded from 1947 to 1994 (Anderson 2005:Appendix C; TNA CO157/6–24,29, Leeward Islands Blue Books, Leper Asylum, General Statistical Tables, 1892–1911, 1916).

Social Dimensions of the Asylum Occupation A primary consideration for selecting Charles Fort as the ideal location of the leper asylum, as was the case elsewhere in the Caribbean, was the need to isolate, control, and confine the infected and segregate men from women to prevent pregnancies and possible sexual transmission of disease. Moderate success was accomplished by putting a barbed wire fence around the fort, placing treatment and living areas within the forts walls, erecting a fence to separate men and women, building a sec­ond fence to prevent men from

164 / Schroedl and Ahlman

leaving, and erecting two detention buildings. The social, medical, and psychological concerns of patients, their friends and relatives, medical authorities, and government officials of­ten made these objectives difficult to justify or enforce, and the practice was discontinued after 1978 (Anderson 2005). When created in 1890, leper asylum capacity was estimated at about 100 patients. From 1892 to 1916, there were between 60 and 73 patients living at the asylum, with men outnumbering women on average 37 to 28. In 1920 and 1925 the population increased slightly to 79 and 76 individuals, with men again outnumbering women (TNA CO157/6–24, 29, 34, 29, Blue Books, General Statistical Tables, 1892–1911, 1916, 1920, 1925). Rec­ords for 1930, 1935, 1940, 1942, 1943, and 1945 indicate a steep decline to an average of 48 individuals and men again outnumbered women (TNA CO157/44, 49, 54, 59, Blue Books, Leper Asylum, General Statistical Tables, 1930, 1935, 1940, 1945; Muir 1943, 1944). Muir (1943) also reports that in 1942 six patients “lived in a small settlement just outside the walls” of the asylum (31). In the 1950s and 1960s the population ranged from 25 to 40, and after 1971 there were no more than 11 patients, with a single remaining resident when the facility closed in 1996 (Anderson 2005:Tables 7.2 and 7.3). At first patients received two meals a day, but this was quickly increased to three and by 1898 four meals were served. Changes in food preservation technology, local food production, and patterns of food imports affected the types of food products consumed and the variety of meals that were prepared. Menus published between the 1890s to 1920s show that breakfast included milk, sugar, bread, and tea, coffee, or chocolate. Served at the evening meal were milk, sugar, and crackers, while the main midday meal consisted of salt pork, salt beef, or salt fish, bread, barley, and potatoes (Irish, yams, and sweet potatoes), and tannias (dasheen). Fresh fish or mutton and occasionally chicken were eaten. Tapioca, sago, cheese, and eggs also were consumed, and there was an allowance made for extra fat from butter, lard, and pork (TNA CO157/6–24, 29, Leeward Islands Blue Books, Leper Asylum, Dietary Scales, 1892–1916). These meals are consistent with published tenders for supplying the hospital and with lists of food actually provided to the asylum patients (Charles Fort 1907, 1908, 1909; TNA CO242/2, 3, I (3), 10 March 1904; II (9), 2 March 1905; III (9) 1 March 1906; and IV (10), 7 March 1907). Two additional diets were created in the 1920s. One contained less variety and more restrictions on substituting one item for another, while the other mandated daily milk consumption (Anderson 2005:178). These were abandoned within a few years, after which all inmates received similar meals. Fresh fruit, vegetables, fish, and meat came from local markets, but the patients also raised

Military and Institutional Occupations / 165

these commodities on the hospital grounds or obtained them through informal means. Lists of foods purchased in the 1950s and 1960s suggest that meals probably differed from what they had been at least through the 1930s because of overall changes in the kinds of foods available, refrigeration, and the fewer number of patients being served. Staples such as bread, flour, rice, and Irish potatoes were provided and so were salt pork and salt fish, but also being consumed were, for example, cooked ham, sausage, red beans, unspecified frozen meat, pig snout, pig tails, and chicken wings (Charles Fort 1956, 1964). The chief medical officer at Sandy Point administered the leper ­hospital and the nearby general hospital. The asylum staff included a senior attendant, cook, and at least one maid, orderly, nurse, and guard (Muir 1943). When on duty they lived at the hospital and looked after the personal, medi­ cal, and social needs of the patients. Family, friends, and a variety of charitable and religious individuals regularly visited the hospital as did a group of official dignitaries in­clud­ing the governor general in 1931 (St. Kitts Union Messenger, 13 February 1931:3). Patients spent much of their time tending their vegetable gardens, flower beds, and fruit trees (Muir 1944:41). Substantial purchases of cloth, sewing needles, and thread show that making clothes, curtains, and furniture coverings was an important activity in the 1950s and 1960s (Charles Fort 1956, 1960, 1961, 1962, 1967, 1968, 1969). They received shoes, clothing, bedding, dining utensils, and personal items such as toothbrushes and were given a monthly allowance or could earn additional money for doing extra work around the hospital grounds (Anderson 2005:172; TNA CO242/2, Gazette, IV (10), 7 March 1907; V (10), 5 March 1908; TNA CO242/3, Gazette, VII (10), VIII (12), 9 March 1911; IX (10), 7 March 1912). The stigma of leprosy was enormous because of its disfiguring and disabling effects, misunderstood means of infection and transmission, and until the 1950s ineffective means of treatment. For these reasons, the whole purpose of the Charles Fort leper asylum was to confine the afflicted to a defined space. The social dimensions of the disease centered on the degree to which confinement was compulsory, with patients treated like criminals with mandatory confinement, to a forceful but more compassionate approach where medical officers quarantined and persuaded the afflicted to seek treatment or where individuals admitted themselves voluntarily. Although each approach was used in St. Kitts, for most of the asylum’s history its patients were treated humanely and given adequate medical and personal care. The approach to care and confinement became progressively more benign and less restrictive as patient numbers declined and those that remained hospi-

166 / Schroedl and Ahlman

talized grew older, less mobile, and more debilitated. The decline in numbers was correlated with an improvement in epidemiology, therapies, and medicines after the 1940s.

Summary and Conclusions Charles Fort is important in the archaeology and history of St. Kitts and the Caribbean for its military occupation and its adaptive reuse as a leper asylum. The fort was constructed in the late seventeenth century as a direct response to persistent conflict with the French over control of the island. The fort retains its seventeenth-­century plan, and its massive bastions and curtain walls, although repaired and damaged by seaward erosion, are among the very best preserved of any early colonial fort in the West Indies. Small-­ scale changes to existing buildings and new construction in the eighteenth and nineteenth centuries deviated little from the fort’s internal organization evident in 1723 and well described in 1734. The sally port and entranceway guardhouses are evident, but other interior buildings, such as the arsenal, only retain ar­chaeo­logi­cal integrity. It is possible some leper asylum buildings were erected on the footprint of what may represent structures dating to the military occupation. By the time military use of the fort ceased, proposals had been made to convert it to a leper asylum. Doing so, however, was delayed until 1890 when the hospital began operation with about 70 patients. Most of the 48 structures and other features visible in 2000, in­clud­ing reuse of the guardhouses and the cistern, are attributed to this period in the fort’s history. The hospital retains its origi­nal configuration with administrative, treatment, food preparation, segregated housing, and extramural agricultural and mortuary areas. The earliest buildings surely were erected on stone foundations. Concrete construction dates to the early twentieth century and likely includes, with subsequent use of concrete blocks, successive building activities dating at least to the 1970s if not later. The clinic, assembly hall, laundry, and toilet buildings collectively exhibit the greatest evidence of modern construction because all these facilities had to be maintained no matter how few patients there were. As their numbers declined so did the need for housing, and some accommodations were reduced to ruins or foundations long before the asylum was closed. As important as any medical or personal care was concerned, the primary purpose of the leper asylum on St. Kitts was to confine and control its occupants. This was unambiguous in laws regulating the asylum, made explicit by designating the residents as inmates, and in the leper asylum’s landscape. To maintain internal control a substantial fence, monitored by a guard and

Military and Institutional Occupations / 167

regulations for passing from one side to the other, separated male from female housing areas. Community buildings had separate entrances or like the assembly hall were divided by a partition. Access to water and sewage disposal made it economical and practical to situate duplicate laundry, bathing, and bathroom facilities along the fence. Separate detention buildings were deterrents to internal transgressions and escape. External controls to prevent escape included the barbed wire fence surrounding the asylum, the fortress walls, and the steel picket fence atop the fortress wall on the men’s side of the compound. External controls also assured regulated and monitored interaction with designated and approved outsiders. These barriers were relaxed or eliminated as fewer cases were diagnosed, admissions declined, and residents became increasingly feeble from age and the disease or died. Societal fear of leprosy drove the harshness of control and confinement, but compassion for the afflicted created genuine concern for their welfare. The variety and quantities of food items and meals prepared for the patients appear comparable to a diet available to and consumed by most of the island’s population. The provision of clothing, shoes, bedding, towels, soap, toothpaste, serving dishes, and tableware helped maintain in­di­vidual pride and personal well-­being. The patients helped build and maintain the hospital grounds, and this served to reduce the boredom of confinement, stimulate social interaction, and improve self-­esteem. Social interaction also was promoted through the library, the assembly hall, and chapel. The chapel, because of its location outside the fortress walls, contributed to in­di­vidual religious faith and provided a symbolic connection to the rest of society. Its proximity to the cemetery also reminded individuals of their mortality, their deceased companions, and the hope of physical healing in the afterlife. Charles Fort is an enduring symbol of the plantation era when military defense was so crucial to the survival of British colonial life on St. Kitts. From the 1670s to the 1690s it stood alone as the island’s primary place of refuge and deterrent to foreign invasion. Its importance was diminished but never dismissed as Brimstone Hill assumed preeminence in the island’s defense. The adaptive reuse of the fort as a leper asylum is no less significant as a material manifestation of the late nineteenth-­and twentieth-­century colonial and institutional history of the island. Fortunately, in establishing this sec­ ond and very different use, the essential integrity of the fortress as a military site was maintained and the integrity of the hospital remains largely intact.

Acknowledgments We thank G. A. Dwyer Astaphan, minister, and Mrs. Hillary E. Watley, permanent secretary of the St. Kitts Ministry of Tourism, Information, Tele-

168 / Schroedl and Ahlman

communications, Commerce, and Consumer Affairs for sponsoring ar­chaeo­ logi­cal and archival research at Charles Fort. We also greatly appreciate the help and encouragement of the Brimstone Hill Fortress National Park Society and its former manager, Larry Armony. Additional interest and support came from Jacinth Henry Martin, minister of culture; Creighton Pension, director, Office of Culture; and Jacqueline Armony, St. Kitts National Trust. Victoria O’Flaherty, Department of Archives, made available numerous rec­ ords relating to the leper asylum occupation, and Victor Smith helped with archival work in England relating to the fort’s military occupation. We are especially grateful to Malcom Govia who shared his personal knowledge of the leper asylum from his nearly 25-­year employment at the hospital.

9 Caribbean Heritage in 3D New Heritage and His­tori­cal Archaeology in Nevis, West Indies Edward González-­Tennant and Diana González-­Tennant The application of digital technologies to document and interpret the his­ tori­cal archaeology of the Caribbean is part of a new heritage (González-­ Tennant and González-­Tennant 2016) that offers an opportunity to share ar­chaeo­logi­cal work with the public. New heritage centers on the use of 3D technologies such as photogrammetry and virtual reality by archaeologists and other heritage workers. Although new heritage shares commonalities with virtual archaeology, digital archaeology, and cyberarchaeology, these other named approaches differ in their focus on specific technologies, or how archaeologists use them. We begin the chapter with a brief description of new heritage to clarify the shared and unique aspects of this approach. We see our work in Nevis as part of a tradition of methodological experimentation in Caribbean his­tori­cal archaeology, which continues to support innovative treatments of the region’s colonial past. This includes the use of macroscopic and chemical composition studies to document the production and trade of Afro-­Caribbean ceramics (Ahlman et al. 2008; Ahlman et al. 2009; Hauser et al. 2008; Heath 1988, 1999; Kelly et al. 2008; Petersen et al. 1999), the increasing use of geographic information systems (GIS) to document and interpret colonial landscapes (Armstrong et al. 2008; Leech 2008), and a reorientation bringing his­tori­cal documents into a fuller conversation with artifacts and ruins (Armstrong 1985, 1990, 2003; Wilkie and Farnsworth 1999, 2005). These studies highlight the ability of archaeologists to embrace a wide range of evidence when investigating the his­tori­cal period of the Caribbean. Many of these studies have become models for archaeologists working in other areas such as North America, but the rapidly expanding practice of new heritage has yet to significantly affect Caribbean his­tori­cal archaeology, and we hope to encourage more of this experimentation by discussing our preliminary work in this regard. This case study is based on three seasons of work at the site of Fort Charles, Nevis. This site represents one of the earliest and longest-­lived British forts in the region. New digital technologies are central to our investigation of this site’s unique history. Our work combines perspectives from qualitative GIS

170 / González-Tennant and González-Tennant

(Elwood and Cope 2009) with immersive virtual reality simulations in archaeology (Carrozzino and Bergamasco 2010) to produce what we are calling an immersive qualitative GIS (IQ-­GIS). Our central goal with this sys­ tem is to reach a broader pub­lic by democratizing access to ar­chaeo­logi­cal data and interpretations by placing this information within a framework that is increasingly intuitive, particularly as video games and other new media become more popu­lar around the world. This approach provides users with access to vari­ous data in an accessible and intuitive way. Such an approach allows archaeologists to represent the site at multiple points in time, in­clud­ ing the juxtaposition of present ruins alongside ar­chaeo­logi­cal interpretations in real time. The ability to share our investigations in this way also speaks to pub­lic archaeology at the site.

Digital Approaches to Representing Archaeological Heritage This section offers a brief overview of the vari­ous ways archaeologists have used new media in recent years. (See González-­Tennant and González-­ Tennant [2016] for a more exhaustive treatment of these topics.) Archaeologists tend to be early adopters of new technologies. Our discipline was one of the first social sciences to embrace the use of GIS (Kvamme 1999). Similarly, as 3D modeling and virtual technologies emerged in the late 1970s and 1980s, archaeologists quickly adapted them to heritage work. The term “virtual archaeology” emerged at this time to denote the use of 3D models to represent ar­chaeo­logi­cal contexts (Reilly 1990) and quickly came to focus on this to produce still images of past sites for interpretative publications and pub­lic outreach (Forte 1997). Early virtual archaeology focused almost exclusively on monumental or prehistoric archaeology, and particularly on sites associated with Greek and Roman history from across Europe. The late 1990s represent a crucial period for the use of digital technologies within archaeology. The increasing affordability of computing hardware allowed more archaeologists to experiment with 3D projects (Koller et al. 2009:73). The use of these 3D technologies was a minor aspect of the overall digitization of archaeology, which grew out of the increasing adoption of GIS, global positioning systems (GPS), and remote sensing technologies during the 1980s and 1990s (Kvamme 1999; Zubrow 2006). Several archaeologists have come to refer to the use of any or all of these approaches as part of a broader digital archaeology, which explores “the basic relationships that archaeologists have with Information and Communication Technology (ICT) and digital technology to assess the impact that such innovations have had on the very basic ways that archaeology is performed and considered” (Daly and Evans 2006:3). Digital archaeology offered a powerful way to investi-

New Heritage and Historical Archaeology / 171

gate the past and was not restricted to one theoretical viewpoint. Processualists embraced GIS and simulations to model past environments and vari­ous forms of human action. Postprocessualists drew on digital technologies to develop reflexive field methods (Hodder 2000), and those embracing phenomenological approaches viewed the technology as a way to develop a deeper awareness of the experiential aspects of past landscapes. If virtual and digital archaeologies center on the documentation and display of ar­chaeo­logi­cal contexts, then cyberarchaeology focuses on the immersive aspects of online worlds (Forte 2010:13), and quickly came to represent “a new way of understanding virtual communities through the study of their cultural artifacts” (Harrison 2009:4). Harrison believes that cyberarchaeology “has the potential to provide insights into the ways in which the notions of heritage are transforming in the early twenty-­first century” (2009:16) as researchers watch the process of heritage creation unfold before their eyes. In this way, cyberarchaeology differs from virtual and digital archaeology because it is less concerned with technological experimentation and more concerned with the ways people use virtual technologies to interact with one another and to create objects virtually that represent ar­chaeo­ logi­cal artifacts, sites, and contexts. New heritage is the intersection of new media technologies and cultural heritage (Kalay et al. 2008). New media is the “translation of all existing media into numerical data accessible through computers” (Manovich 2001:​ 20) and includes the digitization of analog materials (e.g., photographs, movies, and records) as well as computer images and 3D models. We view new heritage as a covering term bridging the range of practices within virtual, digital, and cyberarchaeologies. Without advocating for a hierarchical ordering of these concepts, new heritage does highlight a distinctive ­practice central to representing ar­chaeo­logi­cal contexts in time and space. New heri­ tage easily accommodates methodologies from a range of disciplines, in­ clud­ing his­tori­cal archaeology, oral history, ethnography, and the digital humanities. This approach supports the use of vari­ous technologies to further the goals of collaboration, pub­lic outreach, and social justice (González-­ Tennant 2013). Another benefit of using the term “new heritage” addresses un­intended meanings that may arise when researchers qualify their methodology as “virtual,” which many see as reducing the value of such work by conflating the term “virtual” with notions of “less real” and therefore of less value (Boellstorff 2008:65–66). However, new heritage does not have the same cultural capital as more common terms like “virtual” or “digital,” and it does not always resonate with the pub­lic in the way these other terms do. In other words, our choice to frame our work as new heritage is meant to highlight a practice rather than to overshadow or downplay other approaches.

172 / González-Tennant and González-Tennant

This brief overview is offered to frame and understand trends associated with the ways archaeologists are exploring the use of these digital technologies. Other researchers have transitioned from one term to another over the course of their careers, usually to highlight some aspect of their evolving approach. Maurizio Forte, a pioneer in virtual archaeology (Forte 1997) describes his recent work as cyberarchaeology (Forte 2010). This transition emphasizes Forte’s shift from an approach centered on representing ar­chaeo­ logi­cal contexts in 3D to his recent work focusing on the ways people interact with ar­chaeo­logi­cal contexts and one another through virtual technolo­gies. The four terms discussed here can be quickly characterized on the bases of methods and theories that scholars use. Virtual and digital archaeologists primarily focus on the methodological potentials that arise from new technologies. Cyberarchaeology and new heritage theorize the application of new media. These approaches specifically focus on the ways people in the present use these technologies to make sense of the past. We view new heritage as embracing a wider practice than cyberarchaeology by underscoring how tangible heritage (e.g., ar­chaeo­logi­cal sites) only “becomes ‘heritage’ when it becomes recognizable within a particular set of cultural or social values, which are themselves ‘intangible’ ” (Smith and Akagawa 2009:6). This type of work requires heritage researchers to employ an ethnographic engagement within online worlds, mirroring concerns at physical sites that gauge how vari­ous groups value local heritage resources (Meskell 2005; González-­ Tennant 2014). The application of these technologies to Caribbean his­tori­ cal archaeology is in its infancy. Following a brief history of Nevis and Fort Charles, the remainder of this chapter explores the potentials that new heritage holds for investigating and interpreting Caribbean his­tori­cal ar­chae­ ology.

Brief History of Nevis and Fort Charles Nevis is in the north­ern arc of the islands known as the Lesser Antilles. The island was initially settled by aceramic peoples migrating from South America more than two thousand years ago, although the greatest Pre­colum­bian settlements did not occur on the island until the period between AD 600 and 1500 (Wilson 1989, 2007). The island was seen by Christopher Columbus during his sec­ond voyage in 1493, but European colo­nists did not establish a permanent settlement on Nevis until the 1620s (Hubbard 1996:23). Originally attempting to settle on the northeast coast of St. Kitts, and after visiting nearby Barbuda, Anthony Hilton led a small group of settlers to Nevis on July 22, 1628, where they settled the area that would become Charlestown, which is today the island’s capital city. In 1629 a Spanish fleet attacked

New Heritage and Historical Archaeology / 173

Charlestown, and cannon were installed at Pelican Point (now Fort Charles) to defend the colony. This was ultimately a futile attempt, and the Spanish captured the island, burned fields and houses, and sent settlers back to England. Hilton returned to Nevis the following year (1630) to find that other displaced settlers were already rebuilding the settlement (Dyde 2005:32–38). By 1667 the colo­nists in Nevis were increasingly turning away from tobacco in favor of sugar. The economy quickly prospered, and Nevis became the seat of the Leeward Islands Colony from 1670 until approximately 1700. During this time, Nevis’s economic and po­liti­cal growth was dramatically affected in 1706 when a French fleet of nearly 50 ships landed troops on the island’s southeast coast. The troops made their way to the west­ern side of the island and took possession of Charlestown. The French burned a considerable portion of Charlestown and the island’s cane fields as well as seizing all property, in­clud­ing the island’s slave population. In a surprising turn, considerable numbers of slaves retreated to the slopes of Mt. Nevis during this time and mounted an armed but unsuccessful resistance to the French. Nevis’s decline was exacerbated over the coming decades as St. Kitts entered its own heyday of sugar production beginning in the 1730s (Meni­ketti 2006, 2009, 2015). During this time and through­out the eighteenth century freed Af­ri­cans continued to leave their mark on Nevis. This included owning estates, operating small businesses, and joining the British military (Dyde 2005:106–107). Following the 1706 French invasion, the citizens of Nevis spent several decades attempting to construct a large fortification on the south­ern side of the island. The result was the Saddle Hill fortifications, built by slave labor but most likely never completed (González-­Tennant 2014). During this time, the nearby site of Indian Castle, a partially fortified port, was also active on the island’s southeast side. This site acted as a sec­ondary port during the late seventeenth and early eighteenth centuries in addition to the port in Charlestown. Through the eighteenth century, and while the Revolutionary War continued in North America, the French harassed British colonies across the Caribbean and occupied Nevis for nearly two years in the 1780s. They removed signal cannon from Saddle Hill, spiked cannon at Fort Charles, and stationed a garrison in Charlestown during this time. Otherwise, the French largely left Nevis alone during this period. One possible explanation for this is revealed by Captain Horatio Nelson, who described many of Nevis’s merchants as greater rebels than those in America (Dyde 2005:116). While the Amelioration Act of 1798 advocated for more humane treatment of slaves in the British West Indies, it was not until the Slavery Abo­ lition Act of 1833 that Afro-­Caribbeans in Nevis began to enjoy greater free­doms. While other British colonies such as Jamaica instituted appren-

174 / González-Tennant and González-Tennant

ticeship systems, this failed in Nevis where freed Afro-­Caribbeans successfully lobbied for more free­doms. As in other locations (Delle 1999), Nevi­ sian merchants were able to restore their control of society by the 1840s through a variety of land control tactics. A growing absentee landowner population further undermined Nevis’s economy through­out the sec­ond half of the nineteenth century, leading to widespread poverty and an island population largely devoid of European colo­nists and their descendants by the early twentieth century. Fort Charles sits south of and overlooks Gallows Bay and Charlestown (Figure 9.1). The fort is an irregularly shaped polygonal fort with several uneven bastions where upwards of 26 cannons were placed as of the mid-­ 1600s. The site covers approximately 1.5 acres. Still present are the remains of a structure that served as a guardhouse and armory, the cistern, and portions of the ravelin. The fort’s west­ern and north­ern walls are located above areas that are rapidly eroding into the sea, although discussions with local Nevisians suggest that the cutting away of land here did not begin until the 1950s or later following a series of earthquakes in the region. It is one of the earliest British forts in the New World and occupies a central role in the military history of Nevis. The fort’s location was home to some of the central conflicts between the British settlers and other colonial nations in the West Indies. Sightings and engagements with the Spanish, Dutch, and French have occurred at or near the fort. His­tori­cal documents record multiple construction phases increasing the size of the fort and resulting in the construction or improvement of numerous buildings within its walls. In recent decades, the faced stone of walls have been subjected to looting, which in turn destabilizes the structure in vari­ous places. The looting is regrettable, but it does provide researchers an opportunity to better record the multiple building phases at the site. During the seventeenth century, Fort Charles served as Nevis’s central defensive site and engaged in numerous encounters that successfully repelled the ships of other nations. For instance, in 1673 the fort’s cannons accurately hit several passing Dutch ships “so smartly that we could perceive people going overboard with plugs to stop their leaks” (William Stapleton to the Council of Plantations, quoted in Machling 2012:145). Plans from the early 1700s document a sizable and well-­designed fort. During this time, slave labor was of­ten sought to improve vari­ous elements of the fort, and these workers may have spent considerable time at the site (Machling 2012:150). This aspect of the fort’s history may support the investigation of nonplantation experiences of Afro-­Nevisian lives, providing a comparative context for similar studies on neighboring St. Kitts (Schroedl and Ahlman 2002). The his­tori­ cal documents also record a well-­provisioned fort during the eighteenth cen-

Figure 9.1. Archaeological work at Fort Charles. (Edward González-­Tennant)

176 / González-Tennant and González-Tennant

tury, and numerous structural improvements are known to have taken place at the site during this time (Machling 2012:150–155). A standing guard of men was initially placed at the fort in the 1600s and was maintained until the French occupied Nevis between 1782 and 1784. Fort Charles was central to Nevis’s defense, but it was not the sole military installation on the island. Indeed, it was only one of 14 military installations that eventually ringed the island (Machling 2012). These included a redoubt located along the island’s north­ern shore that was eventually destroyed to make way for an airport runway, a series of gun emplacements along the island’s west­ern shore, considerable armaments at Indian Bay along the island’s south­ern shore, and the aforementioned Saddle Hill. The majority of these installations were in operation by the late 1600s, with Saddle Hill following decades later in response to the French invasion of 1706. Of the vari­ous military facilities in Nevis, Fort Charles operated for by far the longest period of time. It continued operation as a basic military facility until the 1870s as documented by salary payments to military personnel until then (Machling 2012:160). Anecdotal evidence suggests that Fort Charles remained in use until the 1890s, ending its life as a customs fort. Archaeological Investigations of Fort Charles Significant ar­chaeo­logi­cal investigations at the site of Fort Charles began with our 2013 his­tori­cal archaeology field school. Our work seeks to address several themes in Caribbean his­tori­cal archaeology. Fort Charles provides an important case study for investigating long-­term aspects of colonial life in the region. Research at Caribbean military sites addresses key deficiencies in the his­tori­cal archaeology of the region (Watters 2001). For instance, our project seeks to document the experiences of soldiers and other colonial citizens between initial settlement and the late nineteenth century. While investigations of plantation sites continue to represent most his­tori­cal ar­chaeo­logi­ cal projects in the Caribbean (Armstrong and Hauser 2009), archaeologists have increasingly turned their attention to nonplantation contexts, in­clud­ ing a growing literature associated with Afro-­Caribbean experiences at military sites (Ahlman et al. 2008; Ahlman et al. 2009; González-­Tennant 2014; Goucher 1999; Schroedl and Ahlman 2002). Preliminary investigations at Fort Charles reveal a potentially unique pattern of ceramics in regard to the interaction between the island’s British and Afro-­Caribbean populations. Our work combines recent advances in the British empirical tradition of landscape archaeology (Leech 2008) with emerging theoretical studies refocusing attention on interaction and transculturation (Curet and Hauser 2011). Investigating the changing landscape of Fort Charles reveals changing Caribbean lifeways during three centuries, and our project aims to join

New Heritage and Historical Archaeology / 177

a growing list of ar­chaeo­logi­cal projects investigating identity (Wilkie and Farnsworth 2005), power (Delle 1998), and interaction while contributing to recent landscape studies in the British West Indies. Excavations have focused on two contexts at the site (Figure 9.1). The first is a section of stone flooring along the north­ern wall. This area of the fort is in greatest danger of eroding into the sea, and our excavations here seek to document vari­ous architectural features. The sec­ond area focuses on the only remaining building, now in ruins. This building is recorded on maps dating to the late 1600s and was most likely constructed during a series of site improvements at that time. The two-­room structure served as an officers’ quarters and armory. British export and Afro-Nevisian ceramics make up the largest percent­ age of diagnostic artifacts. This includes 1,004 sherds from the 2013 season and 661 sherds from the 2014 season. The 2015 season is still being analyzed. Refined earthernwares account for 35% (n = 612) of the 1,704 diagnostic ceramics currently analyzed. The full range of British export ceramics are present given the long-­term occupation of Fort Charles between the early 1600s and late 1800s. This includes delftwares, creamwares, slipwares, pearlwares, and whitewares. Similarly, the full range of British stonewares, from Rhenish blue and gray to white salt glazed, Nottingham, and ginger beer bottles are present at the site. Stonewares account for 10% (n = 177) of the sherds recovered in 2013 and 2014. Figure 9.2 includes a sampling of these ceramics from the 2013 field season. Although Nevis remained a British colony well into the twentieth century, there is evidence of trade with the United States in the form of yellow wares and alkaline glazed stonewares. This is unsurprising given the attitude many Nevisian traders and merchants had toward Britain in the late eighteenth century, leading to Captain Nelson’s remark that the island’s citizens were more rebellious than those even in the Ameri­can colonies (Dyde 2005:116). Afro-­Nevisian ceramics are one of the most common artifact types found at Fort Charles. They account for 49% (n = 830) of the current ceramic assemblage. Our analy­sis of these ceramics follows previous studies in Nevis and the nearby islands of St. Eustatius (Heath 1988, 1999), Anguilla, Antigua, Barbuda, Montserrat (Petersen et al. 1999), and St. Kitts (Ahlman et al. 2008; Ahlman et al. 2009; Ahlman et al. chapter 4 herein). Most of the Afro-­Nevisian ceramics fall into a generic hollowware category, but a small number of sherds represent cooking pots, yabba, and griddle vessels. In many ways, the ongoing analy­sis of materials from Fort Charles is typi­ cal of a colonial British military site. In addition to the fortifications and standing ruins, ordinance, pipes, and British export ceramics all conform to a site established in the early to mid-­1600s with an occupation lasting until

Figure 9.2. Selection of ceramics recovered through excavation at the site of Fort Charles. Top row, from left to right: hand-­painted blue delftware, polychrome pearlware, polychrome pearlware, blue shell-­edged pearlware. Second row: polychrome pearlware, hand-­painted blue delftware, Staffordshire-­type slipware, white salt-­glazed stoneware. Third row: Staffordshire-­type slipware, porcelain, transfer-­printed blue on white pearlware, white salt-­glazed stoneware with basket weave design. Fourth row: Staffordshire-­ type slipware, porcelain, transfer-­printed blue on white pearlware, brown salt-­glazed stoneware. (Edward González-­Tennant)

New Heritage and Historical Archaeology / 179

the late 1800s. The main exception to this traditional military site pattern is the presence of an unusually large percentage of Afro-­Nevisian ceramics. Our continuing work acknowledges that economic growth and developing colonial societies were structured in unclear and locally unique ways (Kelly 2004; Curet and Hauser 2011). One important avenue in this line of research centers on the ways archaeologists promote these expanded histories of local contexts. Our work at Fort Charles draws on new heritage to accomplish this. New heritage, and particularly the use of game engines to construct immersive virtual world environments, provides a powerful tool for revealing the evidence drawn on by archaeologists in the discipline’s interpretive work. The remainder of this chapter examines the preliminary ways we are using these technologies to create a digital record of our ar­ chaeo­logi­cal work at Fort Charles, in a form that invites exploration by a larger segment of the pub­lic than is traditionally possible at Caribbean ar­ chaeo­logi­cal sites.

New Heritage as Immersive Qualitative GIS The digital reconstruction of ar­chaeo­logi­cal contexts is accomplished by one of two primary methods: the use of 3D capture technologies to automatically measure extant or ruined objects and structures, or the use of 3D modeling software to reconstruct vanished objects as interpreted through ar­ chaeo­logi­cal investigations (Koller et al. 2009:2). This later method is sometimes referred to as hand modeling because the designer must create the 3D image by hand using software. The first method can be divided into two subcategories. The first involves the use of high-­priced 3D scanners capable of capturing entire landscapes or less expensive and smaller units for scanning artifacts. The sec­ond subcategory is photogrammetry and refers to the use of software to extract spatially accurate 3D models from a series of photographs. Our work at Fort Charles draws on this sec­ond type, and specifically the use of Agisoft’s Photoscan software to aid in documenting the site. This chapter’s case study provides a Caribbean example using both scanning technologies and 3D software to digitally document the site and reconstruct it based on documentary research and ar­chaeo­logi­cal fieldwork. Our exploration of new heritage at Fort Charles also examines different ways of interacting with the resulting virtual content. Regardless of whether using 3D scanning or hand modeling, all virtual content is manipulated within 3D software as part of any new heritage workflow. This allows for the integration of captured and 3D models produced by hand using software. The 3D data created by scanning technologies requires post-­processing, of­ten by the same programs used for hand modeling (e.g., Blender). As such, the use of

180 / González-Tennant and González-Tennant

3D software remains central to new heritage projects that seek to produce interactive and immersive versions of reconstructed ar­chaeo­logi­cal contexts. Unfortunately, very few archaeologists are well versed in the use of these programs. This typically requires archaeologists to work with specialists outside the field. This is not intrinsically problematic, but archaeologists who learn this software provide an important bridge between traditional ar­chaeo­logi­ cal practice and digital technologies. We rely on photogrammetry to help us digitally document different aspects of Fort Charles. Photogrammetry uses a series of photographs taken at regularly spaced intervals around and over the object or context. An article by Porter and colleagues (2016) is an excellent overview of integrating photogrammetry into ar­chaeo­logi­cal fieldwork. Although this article specifically deals with scanning artifacts, the principles are the same for re­cord­ing larger objects and contexts. Our primary use of this technology centers on recording our excavation units and the site’s ruined structures in three dimensions. Figure 9.3 shows an intermediate step in this process, the creation of a point cloud in Agisoft’s PhotoScan software (Agisoft 2018) based on a series of photographs (represented as labeled squares in the image), the positions of which are calculated by the software. The resulting model is a spatially accurate 3D model that can be integrated into a virtual world environment. The process of using 3D software to model ar­chaeo­logi­cal contexts (a.k.a. hand modeling) involves five general steps. The first step centers on the collection and organization of supporting evidence. The sec­ond step in hand modeling ar­chaeo­logi­cal contexts involves “blocking out” the general layout of the virtual reconstruction (Figure 9.4). The blocking-­out phase at Fort Charles relies on data from field mapping that is first processed using GIS software. Alternatively, a scaled map image can be exported from the GIS and used as a base map in the 3D software. The third step involves adding details to the model. The standing ruins at Fort Charles present a challenging object to hand modeling. As such, we have reconstructed a conjectural version representing what the structure likely looked like during occupation (Figure 9.5). A version of the ruined structure is also available based on photogrammetry (see Figure 9.3). The current working 3D model of Fort Charles attempts to represent the site as it would have existed following improvements in the late 1600s. The fourth step involves adding textures to the 3D model (see Figure 9.5). Texturing refers to the placement of images on the surfaces of 3D models. Adding an image or photograph of stone walls or plaster provides a more realistic appearance. The fifth and final step focuses on sharing the virtual reconstruction. Virtual and digital archaeologists typically create a series of still images while

Figure 9.3. Documenting ruined structures at Fort Charles with PhotoScan software. (Edward González-­ Tennant)

Figure 9.4. Blocked-­out model of Fort Charles. (Edward González-­Tennant)

New Heritage and Historical Archaeology / 183

Figure 9.5. Textured guard house. (Diana González-­Tennant)

cyberarchaeology and new heritage tend to explore immersive and interactive ways of sharing 3D content. This of­ten involves the creation of content in online worlds (e.g., Second Life) or virtual world environments. The Rosewood Heritage Project is one project specifically using both approaches to bring the history of an Af­ri­can Ameri­can town in Florida to life. The combination of online worlds, virtual world environments, and other new media technologies (e.g., digital storytelling) tell the tragic history of Rosewood, a black town that was completely destroyed during a 1923 race riot (González-­Tennant 2013). The use of virtual reconstructions at Fort Charles produces traditional content such as still images as well as immersive virtual world environments. This facilitates sharing the Fort Charles reconstruction using game engines, programs that allow users to rapidly create and deliver video games. Our game engine of choice is Unity 3D (Unity 2018). Our future work seeks to take this a step further and draw on the interactive potentials of virtual worlds to deliver information that archaeologists typically communicate through images and lengthy written reports.

The Fort Charles IQ-­GIS Previous research into the use of virtual world environments for his­tori­cal archaeology concentrates on the creation of still images and videos or rudimentary uses of game engines for heritage visualization (see González-­ Tennant and González-­Tennant 2016 for a review). New heritage work at Fort Charles seeks to more fully use the potential of game engines to de-

184 / González-Tennant and González-Tennant

liver ar­chaeo­logi­cal interpretations to a wider audience. We are orienting this work to simultaneously draw on a growing tradition of qualitative GIS (QualGIS), which represents a methodological intersection in the use of geospatial technologies for qualitative research (Kwan 2002) and is rooted in a “hybrid understanding of GIS as technology, methodology, and situated social practice” (Elwood and Cope 2009:3). QualGIS requires and represents a mixed methods approach. It treats knowledge as partial and situated, echoing feminist scholars of science (Haraway 1988). QualGIS recognizes that epistemology does not determine methodology and acknowledges how vari­ ous data are open to multiple interpretations. The perspective of QualGIS practitioners in neighboring disciplines echo ideas of contingency and context at the heart of postprocessual archaeologies (Hodder 1982, 1991; Barrett 1987; Hegmon 2003). In practice, QualGIS situates multiple forms of quantitative and qualitative data alongside one another. This requires a nuanced attention to detail as the act of translating his­tori­cal knowledge into geospatial formats can result in the loss of qualitative meaning (Schuurman 2009). As such, QualGIS as a methodology provides a useful perspective when thinking about how archaeologists can represent their vari­ous data in ways that make sense to nonspecialists. The use of game engines supports the creation of a new form of GIS, one that is simultaneously immersive and qualitative. Our experimentation with IQ-­GIS allows users to explore a reconstructed landscape while interacting with the vari­ous lines of evidence informing ar­chaeo­logi­cal interpretation. As one moves through the virtual landscape, a series of menus allows users to access vari­ous lines of evidence. For instance, while exploring Fort Charles’s guardhouse and armory, users can choose to display historic maps, excavation units, artifact counts by layer, and so forth. The context for much of this interaction is framed with 3D models created using photogrammetry (Figure 9.6). This method of interaction also supports engaging with the location and types of artifacts found across the site. The resulting experience actively translates months of ar­chaeo­logi­cal research into an accessible format that is deployed via a web interface or laptop installed at a local museum.

Concluding Thoughts The Fort Charles IQ-­GIS combines traditional his­tori­cal ar­chaeo­logi­cal approaches in the Caribbean with new information and media technologies. Archaeological investigations began at the site after consulting with a wide cross-­section of members of Nevisian society as part of a collaborative project exploring Nevis’s colonial heritage (González-­Tennant 2014). The spa-

New Heritage and Historical Archaeology / 185

Figure 9.6. Fort Charles IQ-­GIS prototype showing interactive elements alongside 3D models of excavations and ruins created with photogrammetry. (Edward González-­ Tennant)

tial layout of the site is interpreted based on historic documents, ar­chaeo­ logi­cal excavations, and field mapping. This information is managed and interpreted using GIS, which provide a starting point to craft an authentic virtual reconstruction of the site. Our ongoing work uses game engines to provide a powerful solution to sharing ar­chaeo­logi­cal work with a wider pub­lic by allowing users to interactively explore vari­ous datasets within a virtually reconstructed version of the site. The ability to represent a site at vari­ous points in time—in vari­ous forms and combinations—is a particularly powerful aspect of new heritage and allows researchers to present their findings not only in three dimensions but also across time. This adds an im­ portant and of­ten neglected fourth dimension to ar­chaeo­logi­cal interpretation, which of­ten represents static moments in the past as representative of complex histories. This is typically a function of physical reconstruction, a limitation we can escape with virtual technologies. This chapter began with an outline of four approaches to using digital and virtual technologies for archaeology. While these approaches share a general interest in the use of digital tools to document, interpret, and represent ar­chaeo­logi­cal contexts, they differ in their focus on the application of technologies and their relationships with users. Qualitative GIS provides an additional focus for archaeologists who embrace a mixed methods approach to investigating Caribbean history. This approach situates qualitative information alongside traditional, quantitative information. This approach neatly

186 / González-Tennant and González-Tennant

intersects the established tradition of documentary archaeology in the Caribbean. An IQ-­GIS allows researchers to deliver information that is typically displayed via static means (e.g., images, reports). Users of our immersive virtual experience find themselves situated within a reconstructed landscape representing his­tori­cal sites at multiple points in the past. González-­Ruibal (2008) has called for precisely this type of work as part of a larger discussion regarding possible alternatives to ar­chaeo­logi­cal narration. He argues that while narration and storytelling remain dominant forms of dissemination among academics regarding their research, archaeologists need to take seriously the exploration of alternative forms of dissemination. Our approach to new heritage at Fort Charles allows users to step inside a 3D site map and access multiple datasets driving ar­chaeo­logi­cal interpretations. In effect, this allows the pub­lic to reveal the onion skin of history in much the same way as professional archaeologists do. At Fort Charles, virtual technologies allow us to position alternative interpretations of the site alongside one another, even in the same virtual world environment. This is accomplished by providing users the ability to choose, in real time, vari­ous reconstructions of the site. These alterna­tive interpretations replace one another in the virtual world environment or are superimposed as partially transparent or outlined structures on the reconstructed landscape. This approach allows archaeologists to address the po­liti­cal nature of their work. Multiple interpretations of a site through time aid archaeologists in the creation of a more inclusive interpretation of the past. The number of British export ceramics in relation to the number of to Afro-­ Caribbean ceramics at Fort Charles suggests a nonplantation context, an aspect of Caribbean his­tori­cal archaeology receiving increasing attention (as evidenced by the chapters in this volume) (Ahlman et al. 2008; ­Ahlman et al. 2009; Armstrong 2001, 2003, 2006; González-­Tennant 2014; Goucher 1999; Meniketti 2006, 2009; Schroedl and Ahlman 2002). Whether these archaeologists explicitly frame their work as po­liti­cal or not, they do share a commitment to expanding the ar­chaeo­logi­cal investigation of Afro-­Caribbean life to include nonplantation contexts. In the British Caribbean, many colonies began importing Af­ri­can slaves in the 1600s, who were freed by emancipation in 1834. This was of­ten followed by a period of apprenticeship in some colonies. As such, in the sec­ond decade of the twenty-­first century, many Afro-­Caribbean communities have been free for as long if not longer than they were enslaved. The creation of projects exploring the ever-­growing postemancipation period will become increasingly important if his­tori­cal archaeologists wish to fully represent the breadth and width of Afro-­Caribbean experiences. IQ-­GIS translates this type of work for a broader audience. In this way, nonplantation sites that reveal important insights into Afro-­Caribbean

New Heritage and Historical Archaeology / 187

l­ifeways can be explored by vari­ous stakeholder communities across the region and beyond. This supports a more inclusive archaeology. New heritage holds great potential for Caribbean his­tori­cal archaeology, but these technologies thus far have received relatively little attention in the region. Our work at Fort Charles was designed to center these interests. A dedication to technological experimentation must begin during the planning stages of a new project. If they are not, vital opportunities are lost. While traditional ar­chaeo­logi­cal methods of­ten produce sufficient data to virtually reconstruct most contexts, the 3D recording of excavations is impossible once units have been backfilled. Moreover, without a dedicated research plan incorporating new heritage approaches from the beginning, vital information may not be recorded. As a majority of his­tori­cal archaeology in the Caribbean is still undertaken by international teams, not identifying sufficient data collection strategies may result in unnecessary delays to sharing ar­chaeo­logi­cal investigations. The investigation of new heritage at Fort Charles situates emerging digital technologies within a growing tradition investigating nonplantation contexts for historic Afro-­Caribbean communities. His­tori­cal archaeologists are increasingly moving away from the region’s plantation contexts. This trend will continue as Afro-­Caribbean communities seek to gain additional knowledge and recognition for postemancipation experiences in the region. New heritage provides a method supporting the pub­lic engagement and a more intuitive exploration of ar­chaeo­logi­cal interpretations. In closing, we are not positioning the Fort Charles Archaeological Project IQ-­GIS as the central, reproducible model of new heritage in the Caribbean but rather encourage others to experiment with these technologies. Ultimately, a diversity of approaches will result in the most productive application of new heritage to the region. The process of experimentation is crucial to the successful implementation of new technologies. Although archaeologists tend to be early adopters of new digital technologies, failure to share results retards the process of implementing them. This is regrettable and avoidable. This chapter represents an early attempt at incorporating new heritage techniques within a traditional ar­chaeo­logi­cal investigation. It draws on previous research carried out by the authors at sites in the United States and offers a novel approach to investigating, recording, and interpreting the region’s cultural heritage. The time investment required to bring these approaches into a project is not so significant as to overwhelm the traditional goals of ar­chaeo­logi­cal investigation. The additional field time required to digitally document sites is nominal. Ten to 20 minutes is sufficient to photograph an excavation unit for photogrammetry. Standing ruins and structures can also be documented

188 / González-Tennant and González-Tennant

in a matter of hours. The processing of these data is straightforward, and typically 10 minutes are needed to complete a unit or structure. The greatest time investment required is the construction of 3D models representing vanished contexts and structures, and the construction of the virtual world environment. Incorporating new heritage within ar­chaeo­logi­cal practice opens new collaborative abilities with students as well. It represents a powerful way for academics to engage students who are more familiar with new media technologies. This may take the form of partnerships with computer science, graphic design, and digital art faculty and students, or it may entail simply encouraging archaeology students who are drawn to these technologies and providing them with increased opportunities to explore those interests. This speaks to much broader concerns regarding training and professionalization for a discipline that regularly comes under pub­lic scrutiny. To realize the potentials of new heritage for the Caribbean, it is necessary for others to incorporate these methods within their research. We look forward to such work with anticipation and await the impressive results it will surely produce.

10 Current and Future Directions in the His­tori­cal Archaeology of the East­ern Caribbean Paul Farnsworth The discovery and colonization of the Caribbean islands were primarily undertaken for economic reasons by European capitalists motivated solely by profit. As a result, his­tori­cal archaeologists in the Caribbean have been drawn to study the institution responsible for the primary source of production of that profit: the plantation powered by enslaved labor. Although a range of commodities was produced on Caribbean plantations, sugar was the most economically important crop. However, despite a common perception, his­tori­cal archaeology in the Caribbean is not just about sugar plantations. They are but one aspect of Caribbean his­tori­cal archaeology. As demonstrated by the chapters in this book, exciting research is being conducted on a wide range of sites through­out the region.

The Frontier One new direction focuses attention on the colonial Caribbean frontier, as in Stephan Lenik’s chapter. Until recently, his­tori­cal archaeologists have not thought too much about the period when the Caribbean was a frontier. The rapid depopulation and virtual extinction of the indigenous population on many islands after Spanish contact and settlement, and especially the genocide in the Greater Antilles, have obscured the fact that indigenous people still inhabited the Lesser Antilles when other European nations settled those islands, and initial European settlement created a frontier situation between indigenous people and colonizers. Treaties between the European powers protected the Kalinago (Carib) on some islands such as Domi­nica (as well as St. Vincent, St. Lucia, and Tobago), and gave at least temporary respite from overwhelming European settlement, prolonging these frontier situations for another century (1660–1763). In all, from Columbian discovery until colonization, on some Caribbean islands there was a frontier period of over 250 years, a period as long as the colonial and postcolonial periods combined. As Lenik points out, historians have generally seen these frontiers as eco-

190 / Farnsworth

nomic resources to be exploited for profit (as did the colonial powers themselves). However, these frontiers were more than zones of yet-­to-­be-­fulfilled economic potential. They were also spaces where indigenous, Af­ri­can, and European peoples interacted on the margins of colonial control, where different priorities and potentials were also important. For the Kalinago, the priority was maintenance of their way of life and, ultimately, their survival. Lenik identified one such strategy at the Indian River settlement in the use of temporary coastal trade settlements where goods could be exchanged with Europeans and subsequently transported to safer inland settlements. The temporary settlement also allowed the Kalinago to observe European visitors and decide whether to engage them in trade, remain hidden, or flee to a place of refuge if necessary. One such place of refuge was the Grand Fond site in an inaccessible coastal location. Unfortunately, Lenik was unable to determine whether this place of refuge was used by Kalinago, European, or people of mixed ancestry, but its location with access to fresh water and food, as well as surveillance of the approaches, was used by a small group of people choosing to stay beyond the reach of colonial control. Despite the treaties that stipulated only Kalinago should inhabit Domi­ nica, in the late seventeenth century French-­speaking settlers with en­slaved Af­ri­cans began woodcutting, growing tobacco, and provisioning estates along the island’s coasts, living outside the law. Missionary activities were allowed for conversion of the indigenous peoples, and these could be funded by plantations on the islands. Grand Bay was one settlement of French-­speaking people, and in 1747 French Jesuits established a parish church, factory, and plantation growing cash crops to pay off the Jesuits’ debts from their activities in Mar­ti­nique (and thus also outside legal settlement). Lenik’s research at the Grand Bay site revealed how the parish served free and enslaved people of French, Af­ri­can, and mixed ancestry in this frontier setting. Importantly, he demonstrated that the frontier setting facilitated the dynamic multiethnic Catholic society. Lenik’s research demonstrated the considerable ar­chaeo­logi­cal potential for future research and study of the frontier period in the Caribbean, which forms the precursor to the plantation societies that subsequently developed on many of the islands. Although his research focuses on Domi­nica, it is clearly relevant to the other Neutral Islands, and there was a frontier period on all of the islands of both the Greater and Lesser Antilles. This may have been brief on some islands, and will be hard to identify ar­chaeo­logi­cally, but the period between first contact and economic subjugation is an important one that has received far too little research thus far and would reward additional attention from his­tori­cal archaeologists in the future.

Current and Future Directions / 191

Urbanism The development of urbanism in the Caribbean has been a major theme in the his­tori­cal archaeology of the region. Possibly the earliest his­tori­cal archaeology in the Caribbean was the study of La Isabela in the Domi­ nican Repub­lic by the Domi­nican Junta para la Celebración del Centenario in June 1892 (Deagan and Cruxent 2002:79). Sevilla la Nueva in Jamaica was the subject of early his­tori­cal ar­chaeo­logi­cal research on urban sites beginning in 1937 (Cotter 1948, 1970), as was Caparra, Puerto Rico, also in the 1930s (de Hostos 1938, cited in Ewen 2001), Nueva Cadiz, Vene­ zuela, in the 1950s (Cruxent 1955), and Port Royal, Jamaica, beginning in the late 1950s (Link 1960). More recent studies at La Isabela (Deagan and Cruxent 2002) and Port Royal (Hamilton 2006), as well as Havana, Cuba (Davis 1996; Domínguez 1981, 1984; Ewen 2001; Fernández Pequeno and Hernández 1996); Puerto Real, Haiti (Deagan 1995; Ewen 1991); Spanish Town, Jamaica (Mathewson 1972a); Sevilla la Nueva, Jamaica (Woodward 2006, 2011); San Juan, Puerto Rico (Joseph and Bryne 1992; Pantel et al. 1986; Solís Magana 1988); Oranjestad, St. Eustatius (Barka 2001; Dethlefsen and Barka 1982; Gilmore 2006b); Charlotte Amalie, St. Thomas (Armstrong 2011; D. Armstrong and Williamson 2011; Williamson and Armstrong 2009, 2011); and Bridgetown, Barbados (Smith and Watson 2009; Welch 2003), among others, have continued this research theme. Jamestown, on Nevis, is important as a modern study of early urbanism in the Caribbean, albeit urbanism that ultimately failed. By finding the truth behind the romantic legend of the seventeenth-­century town that was destroyed by a tsunami, Carter L. Hudgins, Eric Klingelhofer, and Roger H. Leech have brought an important piece of history back into the present. Investigating romantic legends, such as the one about Jamestown, is something that attracts the general pub­lic to archaeology and helps to create an image of the discipline in popu­lar culture. As such, projects of this nature can be an important means to connect with the pub­lic and draw them into, and hopefully develop an appreciation for, the discipline. However, as the research by Hudgins, Klingelhofer, and Leech reveals, the reality behind the legend is of­ten very different, and considerably less romantic, as is the research process that reveals it. Jamestown on Nevis was not engulfed in a tsunami, the population was not drowned, and this was not another Port Royal, Jamaica. Instead, the town died, not in 1690 but around 1720 as a result of French invasion, economic and demographic chaos exacerbated by drought, potential slave rebellion, and increased economic competition from neighboring islands. There­after, it was decidedly unromantic soil erosion from sugarcane fields

192 / Farnsworth

upslope on Mount Nevis that slowly buried the site, followed by nineteenth-­ century construction that masked evidence of the earlier settlement. The truth is much less romantic and dramatic; it is nonetheless a fascinating story about the development of a frontier settlement that grew in prosperity in the seventeenth century, only to fail in the early eighteenth century. Its late eighteenth-­to early nineteenth-­century rebirth as a suburban villa for aristocratic relaxation was followed by a sec­ond failure, as the renaissance of the sugar industry was ended by emancipation of the enslaved labor force on which it depended. Jamestown and the island of Nevis were not alone in following these cycles of growth and failure. They are another feature of the history of many islands in the Caribbean, and one for which his­tori­cal archaeology can document the effects on the people who lived there from research at a variety of sites, be they urban or rural.

Trade and Freedom One strategy to ameliorate the cycles of economic boom and bust associated with many of the sugar islands was to develop as a node in the trade networks, serving many islands as middleman and transshipment point for goods from many nations and their colonies. St. Eustatius, discussed by R. Grant Gilmore III in this volume, was the foremost of these ports, becoming the first free port in modern times in 1754, which propelled the Dutch island to be the center of international Atlantic trade in the sec­ond half of the eighteenth century. However, first British occupation in 1781 and then French in 1795 introduced regulations that strangled the trade on which the island depended, leading to a steep economic decline into the early nineteenth century from which the island has never recovered. Although there were many plantations on St. Eustatius, sugar monoculture there contributed little to the economy of the island. The trade network centered on the island affected all levels of society from merchants and planters to free and enslaved blacks. The enslaved people created wealth for their owners and themselves in Oranjestad. Indeed, all residents of the island participated in the free trade of St. Eustatius, and opportunities for self-­manumission were plentiful, creating a large free-­black population on the island. Both enslaved and free people enjoyed significant mobility not only on the island but also to travel to others in pursuit of trade. Enslaved Af­ri­cans were at the core of the trade of St. Eustatius and were able to purchase their free­dom more easily than on other islands, but having done so, many stayed on Statia to take advantage of the social and economic circumstances unique to the island.

Current and Future Directions / 193

Gilmore’s intensive ar­chaeo­logi­cal investigation of a Caribbean free black community located on the outskirts of Oranjestad revealed three domestic structures, animal burials, boundary walls and ditches as well as an Af­ri­can-­ style shrine. Gilmore interprets the sequence of domestic structures to represent the increasing purchasing power of the free-­black occupants derived from trade activities. These changes reflect the upward mobility of Statia’s free blacks over the 150 years the site represents. The shrine provides evidence that elements of West Af­ri­can spirituality were maintained by the free black population of St. Eustatius. The earthen mound shrine has direct parallels with those built by a number of groups in West Africa. Also, being in the open, it appears that on Statia, at least for free blacks, the overt practice of West Af­ri­can religion was permissible. The lack of artifacts found in the spaces between buildings and boundary walls also suggests ties to West Af­ri­can traditions of the use of space, in­ clud­ing yards swept clean of vegetation and trash, leaving artifacts concentrated along boundary walls. Thus, while free blacks on St. Eustatius were participating in the European economic sys­tem that brought them wealth and social mobility, at the same time they were practicing West Af­ri­can traditions and culture. The investigation of free black settlements, both urban and rural, is an important direction for research in the Caribbean as we seek to understand the development of island cultures and identities from colonialism to the present.

Trade, Identity, and Class The prime beneficiary of the decline of St. Eustatius in the nineteenth century was probably Charlotte Amalie on the Danish island of St. Thomas. By the beginning of the nineteenth century, Charlotte Amalie was a densely settled urban port connected to other islands in the Caribbean and to Europe, the Americas, Africa, and the Far East. The Bankhus was a residence occupied by a series of merchants and their families in the nineteenth c­ entury. The artifacts recovered from the Bankhus by Christian Williamson and Douglas V. Armstrong came from a series of contexts, some representing the gradual accumulation of the detritus of daily life but most associated with destruction caused by a series of hurricanes and tropical storms that shook the island in the nineteenth century. Subsequent repair and refurbishment created an updated material culture at the site awaiting the next cycle of destruction and repair, providing the archaeologist with a sequence of material culture documenting change through time at the site. Hurricane or tropical storm destruction followed by rebuilding is common through­out the Caribbean and is far more so than by earthquake and

194 / Farnsworth

tsunami, even though that may appeal to the more romantic. The cycle of storm damage and repair, as well as resulting relocation and/or abandonment, is a feature of the history of many islands in the Caribbean and can provide a sequence of ar­chaeo­logi­cal deposits through time that are useful for exploring a range of issues. For example, the dating of the features themselves, together with the evidence of the degree of destruction and devastation they represent, could be used with his­tori­cal records to refine the storm history of each island and be used in the development of climate models to document and predict the frequency and severity of storms through­out the region in the future. With global warming, sea level rise, and climate change being such important problems facing modern Caribbean populations, such nonanthropological uses of ar­chaeo­logi­cal data represent an important contribution for archaeology in the region. Danish colonial neutrality and a pragmatic approach to colonial trade allowed Charlotte Amalie to thrive in the nineteenth century with its expansive trade connections and diverse population. The Bankhus was typical of the diversity of the population, as over time it was home to merchants/ bankers from Germany, France, Denmark, and St. Thomas, although their families were almost exclusively from St. Thomas. They were white, black, and mixed and represented religious affiliations in­clud­ing Jewish, Tora­elite, Lutheran, and the Dutch Reformed Church. Their servants and support staff were mostly from the islands, St. Thomas primarily of course, but also from St. John, St. Eustatius, St. Croix, St. Domingue, Tortola, Saba, and St. Martin, and their religions were as diverse as their masters’: Protestant, Lutheran, Dutch Reform, Episcopal, Methodist, and Roman Catholic. Yet, it is a major concern that the materials recovered ar­chaeo­logi­cally displayed little evidence of the differing origins or beliefs of the vari­ous occupants over time. This demonstrates the challenges those of us interested in exploring Carib­ bean identities face, such as geographic origins, ancestry, or religion through the ar­chaeo­logi­cal record. Do we have the theoretical tools and the creativity to develop the methods necessary to reveal and explore the identities of those we wish to study through the ar­chaeo­logi­cal record? Certainly there are studies both in this volume and published elsewhere that demonstrate that we can do so under certain circumstances, but clearly this is not always the case, and one that archaeologists working in the Caribbean and beyond have to address. The ar­chaeo­logi­cal materials also exhibit little differentiation between contexts associated with class, especially between the merchants and their servants, even though Danish society was divided as deeply by class as any other colonial society. More signs of wealth were associated with the main house, to be sure, but the more distinct, lower-­cost material record associ-

Current and Future Directions / 195

ated with the servants at the neighboring Magens House compound was not found at the Bankhus (Armstrong 2011; D. Armstrong and Williamson 2011; Armstrong et al. 2009, 2013; Williamson and Armstrong 2009, 2011). This demonstrates that in urban sites it is frequently difficult to separate the material culture of the masters and the servants, be they free, as in this case, or enslaved, and the challenges faced by archaeologists interested in exploring issues of class through the ar­chaeo­logi­cal record of these sites.

Free Urban Populations The presence of only one enslaved person in the history of the Bankhus property is a reminder that on St. Thomas the majority of people of Af­ri­can descent were free by 1803, and the port of Charlotte Amalie was a haven for free blacks, even though slavery continued to play a major role in agriculture and the production of wealth in the Danish West Indies until 1848. This study of St. Thomas, together with that by R. Grant Gilmore III on St. Eustatius, illustrates that there was a large, free, laboring class in all the urban centers of the Caribbean whose lives are just as deserving of ar­chaeo­ logi­cal attention as their wealthy employers, despite not leaving a record of prominent houses, costly goods, and extensive documentary and pictorial evidence. Be they black, white, or mixed, these free (legally or otherwise), urban populations were a very important component of Caribbean society in the colonial period, and afterward. Yet, the study of the lives of the many urban residents who were neither wealthy nor enslaved is still an underrepresented aspect of Caribbean his­tori­cal archaeology. This is a topic that takes on increasing importance after emancipation as rural populations were drawn to economic opportunities in the towns and ports. For many stakeholders in the Caribbean today, research on nonwealthy, urban populations is likely to be of more interest and importance to them than rural, enslaved plantation populations. Unfortunately, the opportunities for such urban investigations are of­ten limited, and the techni­cal, methodological, and financial challenges of urban archaeology in the Caribbean make such research extremely difficult. However, as in the United States and Europe, urban archaeology must become a routine element of development projects in modern Caribbean towns and cities if a better understanding of the development and characteristics of Caribbean life and cultures is to be obtained. The materials recovered ar­chaeo­logi­cally from the Bankhus display the access to goods from diverse sources and the wealth that would be expected of Danish colonial merchants. Some artifacts tie directly to the Danish presence, while many show connections to Danish trading partners in the Far

196 / Farnsworth

East and the presence of the Moravian missionaries in the Danish islands. Yet the origins of most of the artifacts are from a greater diversity of sources, Europe, the Far East and the Americas as well as regional Caribbean trade. Despite this seemingly open attitude to diversity and the goods of all nations, underlying Danish control was exhibited in the photographic record of the Bankhus, with its thick and imposing exterior walls, interior rooms filled with fine furnishings and antiques paying homage to colonial life, and walls covered with symbolic representations of Danish power and authority.

Afro-­Caribbean Wares Ports also played an important role in the distribution of Afro-­Caribbean wares. These wares have been the subject of study for over 50 years in the Caribbean (e.g., Handler 1963, 1964), in part because they are one of the few items of material culture that survive in the ar­chaeo­logi­cal record that were made and used by people of Af­ri­can descent outside of European control. These wares are a direct connection to the material culture of the enslaved and freed Af­ri­cans who worked on the plantations of the Caribbean. While most previous studies of these wares have focused on intra­island trade, Todd M. Ahlman, Gerald F. Schroedl, Barbara H. Heath, R. Grant Gilmore III, and Jeffrey R. Ferguson have revealed evidence for interisland trade in these wares. Ahlman, Schroedl, Heath, Gilmore, and Ferguson’s results from instrumental neutron activation analy­sis (INAA) of Afro-­Caribbean ware sherds and clay samples from 24 sites on seven different islands showed that there were numerous paste recipes being used in pottery manufacture by enslaved and freed Af­ri­cans. They conclude that there was likely more than one manu­ facturing center on each island, and that some people had access to markets that offered pottery from more than one source, be the source local or not. They document that interisland trade in Afro-­Caribbean ceramics took place. Ahlman and his coauthors believe that Nevis was the major center of production in the north­ern Lesser Antilles, with very few vessels from elsewhere appearing on the island. In contrast, St. Eustatius (Statia) appears to have had little significant on-­island pottery production, with pottery from eight different sources being found there, but none matching the two clay sources from the island that were tested. Thus, it appears that pottery from across the north­ern Lesser Antilles entered St. Eustatius and was used by those in many households on the island. Of course, the slave and free black community on St. Eustatius was unlike any other in the New World, having free access to a trade center that was unparalleled in the Caribbean. St. Kitts had pottery from five different sources, with four of those sources

Current and Future Directions / 197

being represented at the Brimstone Hill Fortress. At least some of the sources of these pots were on other islands; however, at the nearby Earl Romney slave village at the base of Brimstone Hill, only two sources are represented, and despite its proximity to the port at Sandy Point, these were on either St. Kitts or nearby Nevis. Not everyone on St. Kitts had access to the inter­island market. Sites in more remote parts of the island also had pottery from only a small number of sources and used locally made pottery or traded exclusively with Nevisian potters. Ahlman, Schroedl, Heath, Gilmore, and Ferguson also found that 14 of their 24 sites had sherds from just one clay source, with a couple of other sites having sherds that could not be assigned to any compositional group. So, at two-­thirds of their sites, only one clay source was represented. Therefore, while the study provides clear evidence of interisland trade in Afro-­ Caribbean ceramics, it also shows that local production and the internal market on most islands dominated the trade in pottery among enslaved and free Af­ri­cans in the north­ern Lesser Antilles. This study presents important new evidence for both inter-­and intraisland trade, but it only represents the beginning of the research that needs to be carried out. Similar studies using this new technology must be conducted on sites from these islands and, especially, on other islands in the region to enable us to understand the networks of trade and exchange that existed. This approach will allow archaeologists to study the social relationships that the possession and movement of these ceramics reflected through­out the region. Studies are needed that compare and contrast ceramic distribution on different islands, that are not limited by colonial or modern national boundaries, or bodies of intervening water, and not limited to just local ceramics, but other locally produced materials and artifacts. Each island has its own unique history and culture, and such materials-­identification studies will be important to help define what makes the people and their cultures different from their neighbors’ in all realms of their lives. Afro-­Caribbean ceramics played a major role on many Caribbean plantations, as at Habitation Crève Cœur on Mar­ti­nique studied by Diane Wall­ man and Kenneth G. Kelly, and Green Castle Estate on Antigua studied by ­Samantha Rebovich Bardoe. Both were large eighteenth-­through nineteenth-­ century sugar plantations. The archaeology of plantations, and especially sugar plantations, has been a major theme in Caribbean his­tori­cal archaeology since the 1970s, especially on the Anglophone islands (e.g., Armstrong 1990, 1992, 2011; Armstrong and Kelly 2000; Delle 1998; Handler and Lange 1978; Farnsworth 1996, 1999, 2001; Hicks 2007; Higman 1998; Pul­si­pher and Goodwin 1999, 2001; Reeves 2011; Wilkie and Farnsworth 2005), although research on the Dutch (Barka 1985, 1993, 2001; Haviser

198 / Farnsworth

1999, 2001, 2013; Haviser and DeCorse 1991; Haviser et al. 2010; Heath 1988, 1999; Hofman and Haviser 2015) and French (Bain et al. 2011; Bonnissent et al. 2012, 2013; Brunache 2011; Croteau 2004; Delpuech 2001; Garros et al. 2013; Gibson 2009; Kelly 2004, 2008, 2009, 2011; Kelly and Bérard 2014; Tomadini et al. 2014; Wallman 2014) islands is beginning to come more into prominence. This is criti­cal as scholars seek to understand the full range of experiences of the Af­ri­can-­descendant population of the islands. The two studies in this volume demonstrate general similarities but also differences that reflect the unique context of each island. Both studies also focus on the changes that occurred with emancipation, reflecting a recent, growing trend in the ar­chaeo­logi­cal study of Caribbean plantations. Afro-­Antiguan wares form a major proportion of the household assemblage excavated from the period of enslavement, primarily as food preparation vessels, but after emancipation the proportions of these vessels, and ceramic food preparation vessels in total, decline drastically. Instead, more imported European ceramics are found, perhaps reflecting a greater availa­ bility and access to these goods, but the vessel forms are those used for food serving and storage, not food preparation. Bardoe suggests that cast iron cooking pots (which do not appear in the ar­chaeo­logi­cal record with the same frequency as fragile ceramic cooking pots) were used after emancipation as part of a shift to more mobile cooking activities using coal pots made of Afro-­Antiguan ware, instead of fixed hearths in cabins. At Habitation Crève Cœur on Mar­ti­nique, coco neg was also primarily used for food preparation vessels. It was used consistently by the enslaved population, but its use decreased prior to emancipation and then became even less common during the sharecropping period. Wallman and Kelly also see the shift to more imported European ceramics with emancipation, but they do not see the same change in vessel forms. They see more direct functional replacement of locally made wares with imported European wares. Ceramic cooking pots formed the majority of the assemblage at both pre-­and postemancipation houses excavated at Habitation Crève Cœur. However, at the Green Castle Estate on Antigua, after emancipation, food preparation vessels represent a small proportion of the assemblage, having been the majority prior to emancipation. On Antigua we see a shift to an emphasis on food storage with emancipation that reflects the uncertainty of food supply after planters were no longer obliged to provide the people with the provisions as they had during enslavement. While the enslaved had access to provision grounds prior to abolition, their produce could be traded or bartered at island markets for goods or other food supplies, and the planter was obliged to provide basic levels of provisions. For those who were elderly or infirm, these supplies were criti­cal. With emancipation, the planters re-

Current and Future Directions / 199

stricted island markets, and it proved difficult for many Antiguans to grow enough basic provisions, and the plight of those who could not work became especially dire.

Challenging the Documentary Record One important role of his­tori­cal archaeology in general is to challenge, instead of reinforce, the documentary record. Using material evidence to interrogate inferences from documentary sources remains an important aspect of the discipline, and one that must continue to be an important goal for his­tori­cal archaeology in the Caribbean. Caribbean historiography is based on documents written by and for the colonizers and presents its perspectives on what they viewed as important. In this postcolonial era, the material record provides an alternative source of data that is not limited or tainted in the same way (its interpretation, of course, is another matter). Wallman and Kelly’s analy­sis of the faunal remains from Habitation Crève Cœur provides a detailed view of the subsistence patterns on a French plantation that contradicts expectations based on his­tori­cal accounts. Their results document the importance of pig, sheep, and goat raised on the plantation, illustrating that these animals were largely consumed there, and that the best cuts were not traded away as his­tori­cal accounts suggest, or were only consumed on the plantation during celebrations. The consumption of imported barreled, salted beef distributed either as rations or purchased directly by the enslaved laborers was also identified in the ar­chaeo­logi­cal rec­ ord, but not to the extent anticipated from his­tori­cal accounts that suggest it was an important staple of enslaved people in the French West Indies. There is scant mention of hunting in the documentary record, and yet vari­ous wild mammal species were also recovered, with opossum, agouti, and mongoose being consumed. A wide range of fish species were also recovered, clearly indicating that they formed an important part of the diet, which again contradicts the his­tori­cal record that implies that fishing played a very marginal role in the diet of enslaved people. Equally, while the his­ tori­cal record suggests that salt fish was an important staple, no evidence for its consumption was found at Habitation Crève Cœur. Shellfish were also an important part of the diet of the enslaved people at Habitation Crève Cœur, which is not in accord with his­tori­cal accounts, which emphasize the importance of crustaceans, particularly crabs, but not shellfish. Indeed, the wide range of local food species consumed at the site in addition to the introduced domestic species suggests a significant adaptation to the local environment by the enslaved population that is not apparent from the documentary record alone.

200 / Farnsworth

Wallman and Kelly’s study at Habitation Crève Cœur is also important in that it demonstrates the changes in diet that occurred with time, again something that is impossible to identify in the documentary record. There is a tendency to assume that conditions on the plantation were static through the period of enslavement, and archaeologists of­ten present a static picture of what was a dynamic situation. Wallman and Kelly’s study illustrates that changes were taking place through­out the period of slavery. In this case, there were significant changes in the procurement of fish and shellfish. After emancipation there is a change in preferred species as well as procurement strategies, perhaps because the free laborers had more time to exploit a wider range of habitats. Habitation Crève Cœur also provides evidence of different household choices, with enslaved households occupied at similar times making slightly different choices in their food and ceramic acquisitions, another aspect that is impossible to see in the documentary record. While some of these decisions may have been driven by in­di­vidual household choice, others differences provide material evidence of the hierarchy within the enslaved population of the plantation. Certain households used low proportions of coco neg and much higher proportions of imported ceramics than others. The recovery of luxury commodities and sewing materials also suggests differential access to goods. Increased access to goods was accompanied by a diet in­clud­ing more fish, a wider diversity of shellfish, and more bird than other households.

Military and Civilian Interactions A major theme in his­tori­cal ar­chaeo­logi­cal research in the Caribbean has been European military sites, many of which are enduring symbols of colonial power and control over the islands. Military fortifications were constructed from the earliest European settlement in the Caribbean at La Navidad (Deagan 1987) through the colonial era, as each European power sought to protect its assets from the others. These fortifications required constant upkeep and repair as well as upgrading of the defenses as military methods and techniques evolved and colonial politics and allegiances shifted over time. Yet eventually, the military usefulness of them all came to an end. Some fell into disrepair and were left to return to nature, while others were repurposed, with many currently being used for tourism purposes. Many of the excavations at military sites have taken place in the context of architectural preservation as these sites became a focus for tourism. Gerald F. Schroedl and Todd M. Ahlman’s research at Charles Fort on St. Kitts is an example of such a project. Charles Fort is one of the best pre-

Current and Future Directions / 201

served early British forts in the Caribbean. First constructed in the 1670s, it was the largest and most important fortification on St. Kitts until it was superseded by Brimstone Hill Fortress in the mid-­eighteenth century, although it continued to be used by the military until 1854. The ar­chaeo­ logi­cal research has documented the changes in the fortifications and their chronology. Edward González-­Tennant and Diana González-­Tennant’s research at Fort Charles on nearby Nevis is focused on another early, long-­ occupied, British fort. Originating in 1629 with a Spanish attack on the island, Charles Fort was the most important defensive site on Nevis, continuing in use through the 1870s, and possibly until the 1890s. One focus on the González-­Tennants’ work has also been fully documenting architectural features, especially those in danger from erosion. Their work also seeks to document the experiences of soldiers and other colonial occupants of Fort Charles on Nevis, as has been done at Brimstone Hill Fortress on nearby St. Kitts (Ahlman et al. 2008, 2009; Klippel 2001, 2002; Klippel and Schroedl 1999; Schroedl 2014; Schroedl and Ahlman 2002). The soldiers and support personnel associated with military sites were a significant element of the colonial population on many Caribbean islands, especially in the ports and towns, yet one that had been little studied prior to the beginning of the current century. Much of the construction work at British military sites was undertaken by enslaved laborers, and this was the case at Fort Charles. There is a growing literature on Afro-­Caribbean populations associated with military sites, and the excavations at Fort Charles have already contributed to the investigation of Afro-­Caribbean peoples’ lives in this context. Comparisons with artifacts recovered from Brimstone Hill Fortress on St. Kitts reveal that a significantly different pattern of interaction between the British and the Afro-­ Caribbean population was taking place at Fort Charles. At Brimstone Hill, Afro-­Caribbean wares appear to have been primarily used by enslaved Af­ ri­cans, and were relatively few in number. In contrast, at Fort Charles, there is a much higher proportion of these wares. The González-­Tennants argue that this was through choice, not necessity, as it is constant through­out the site’s occupation, both in periods when Nevis was a dominant player in the sugar trade as well as during periods of economic decline, when local supply was a less costly, or possibly easier, option. The exact nature and meaning of this interaction between the British population and the Afro-­Caribbean population must be further explored, and questions about whether it was purely an economic relationship or reflects social interaction need to be answered. The work at Fort Charles and Brimstone Hill shows that the colonial societies that developed on the islands did so, and were structured, in ways that

202 / Farnsworth

were uniquely shaped by the particular island’s context and population, even though they were all bound up in larger regional and international developments over which they had little control. This suggests that an important direction for future ar­chaeo­logi­cal research will be the study of fortifications focused on the people who built and occupied them, and their interactions with nearby communities, be they rural or urban, and not just the architectural details and military aspects as has been the practice so of­ten to date.

Marginalized Populations A feature of modern his­tori­cal archaeology is the recognition of the importance of studying populations who have been marginalized, be it geographically, socially, or physically. Indeed, most of the chapters in this volume do just this. The study of the late nineteenth-­through twentieth-­century reuse of Charles Fort in St. Kitts by Gerald F. Schroedl and Todd M. Ahlman examines a population that was physically marginalized through their incarceration due to having contracted Hansen’s disease (leprosy). Leper asylums were established on many of the islands in the Caribbean as postemancipation, colonial, and postcolonial governments tried to address the pub­lic health issues that Hansen’s disease presented. However, there has been very little his­tori­cal or ar­chaeo­logi­cal research devoted to these institutions in the Caribbean. The study of the Charles Fort asylum is, therefore, a major contribution to the archaeology of these institutions in the Caribbean, and beyond, as well as the more general archaeology of confinement that has become an important theme in the broader discipline. The study of the postmilitary reuse of Charles Fort provides a significant and fascinating contribution to the growing body of ar­chaeo­logi­cal research on the late nineteenth and twentieth centuries of St. Kitts. The focus on the postemancipation use of the fort is significant as many historic structures through­out the Caribbean went through different phases of use and disuse during the colonial period and rarely remained unaltered through time. While archaeologists of­ten privilege the origi­nal purpose of a particu­ lar site or structure in their research, its subsequent uses may be as important as its colonial origins. It is certainly conceivable that for some sites and structures their postcolonial use is more significant to the current inhabitants of the island.

New Heritage In general, archaeologists working in the Caribbean have not been slow to apply new methodologies to the study of the region’s ar­chaeo­logi­cal past, as

Current and Future Directions / 203

Edward González-­Tennant and Diana González-­Tennant point out. However, they think that archaeologists have been slow to adopt “new media” in research. They use their research at Fort Charles on Nevis to illustrate the application of “new heritage” methods and demonstrate their potential for future studies. By “new heritage,” the González-­Tennants mean a range of digital methods and technologies for documenting and interpreting cultural heritage resources. In particular, their focus is on the use of virtual world environments to represent artifacts, structures, and landscapes. The González-­Tennants are using new heritage as one method of allowing the public, locally and internationally, to explore the unique history of Fort Charles. They do so using game engines to construct virtual world environments. This approach allows users to interactively explore vari­ous datasets within the virtually reconstructed version of the site, getting away from static reconstructions of the past. Their approach also provides the ability to represent the site at vari­ous points in time, bringing in the fourth dimension that is such an important element of archaeology. However, this can only be carried out if the necessary information is recorded from the start of the project. To effectively use these approaches they have to be included in the initial planning of the project and the necessary data collected in the field. Thus, for example, excavation units have to be photographed from multiple angles to render a three-­dimensional reconstruction in the computer. This is within the abilities of any archae­olo­ gist, but is more than what is traditionally recorded, and will take more time in the field. Artifacts similarly have to be photographed from multiple angles, if no 3D scanner is available. Similarly the other steps in the process that the González-­Tennants describe are not necessarily beyond the abilities of most archaeologists, but they do represent an investment of time. I think this is, in part, why new heritage has not received as much attention in the Caribbean as it has elsewhere. As the González-­Tennants point out, the majority of his­tori­cal archaeology in the Caribbean is still undertaken by international teams, not island residents alone. As such, time in the field is both precious and expensive for most researchers and there is always more that could be done than there is time to do. Archaeologists have to be convinced that the benefits of this approach justify the additional time required, compared to the other uses that this time could be put to, and, importantly, the costs involved. I think another aspect that archaeologists have to consider in this calculation is how the intended audiences will access these virtual reconstructions. Is a laptop at a local museum an adequate outlet for the work involved in such virtual reconstructions? What is the level of household internet access and speed, as well as computing power, among the general population in

204 / Farnsworth

the Caribbean, if that is the intended audience for this work? In the United States there are clear disparities in the levels of internet access between ur­ ban and rural populations as well as rich and poor and black and white. It is criti­cal that in our enthusiasm for the possibilities of new technologies we do not create a new heritage that is inaccessible to the very stakeholders in the region we most want to engage, thus rendering them, albeit inadvertently, sec­ond-­class citizens in the virtual world, even though these methods may well engage those outside the region very effectively. The González-­Tennants are clearly aware of this problem and are working hard to try and make their results available through multiple institutions on Nevis and are working with Nevisians to share their heritage from the research at Fort Charles. The ever-­increasing use of smart phones may also allow increased access to this “new heritage,” provided the technologies can be accessed through them. Another cause for concern is the need for the involvement of local stakeholders in the creation of these representations of their past to avoid the imposition on them of a new heritage manufactured by international researchers. The González-­Tennants are working closely with Nevisians, as part of a collaborative project, to avoid these issues, and future researches must equally work with local stakeholders as they develop their projects. These new approaches to presenting and exploring the results of ar­chaeo­ logi­cal research in the Caribbean are exciting and important for fellow his­ tori­cal archaeologists and other heritage professionals. With careful planning and appropriate caution, they will also be extremely valuable tools for sharing the region’s unique heritage both internationally and with the stakeholder communities across the region.

Local Stakeholders One question that archaeologists working in the region, and especially international researchers, have to try and answer is, who are the local stakeholders. Today, every island has a diversity of residents, some who may claim ancestral occupation going back several thousand years, many others hundreds of years, and still others a few decades or less. None of these groups is monolithic in its opinions, far from it, yet each has claims to aspects of their island’s past. Archaeologists have to build coalitions with a range of local groups and learn to respect the views of others who may not share the same goals. Several of the chapters in this volume report results from projects that have built successful coalitions with local groups, NGOs, and governments, and it has become essential for international researchers to do so. The growth of new resident populations on some of the islands can pre­ sent particular challenges for international researchers. These groups may

Current and Future Directions / 205

have shallow roots on the island, but they of­ten have more economic power, higher levels of education, and, sometimes, po­liti­cal influence. Often being origi­nally from west­ern Europe or the United States, some of these new residents have more exposure to archaeology and, of­ten, more time to pursue an interest in it than do other residents. As a result, they of­ten play important and active roles in local ar­chaeo­logi­cal and his­tori­cal societies. For international researchers, these groups can of­ten be those to turn to as representing local stakeholders, as they are of­ten the most receptive to international ar­chaeo­logi­cal researchers. However, archaeologists must recognize that they represent just one group of local stakeholders and must resist the temptation to ignore other groups who are less economically powerful, understanding of the goals, or welcoming of outsiders. Island governments are, of course, another local stakeholder that has to be included, or at least consulted, in ar­chaeo­logi­cal research projects, but they also present significant challenges for international researchers. Legal mandates aside, government officials, be they elected or appointed, represent constituencies among the local population. While economic issues, such as direct or indirect benefits through tourism, or potential economic problems such as impacts on construction and development projects, can be prominent in such discussions, archaeologists have to understand that these are just one aspect of complex local po­liti­cal situations of which they frequently have far too little knowledge and understanding. Becoming too entangled in local politics may bring short-­term joy but long-­term grief as po­liti­cal fortunes wax and wane and governments change. Involving local stakeholders in his­tori­cal archaeology is also vital to protecting their cultural heritage and preserving their unique sites. Archaeological sites are threatened across the region, and many are being destroyed by development as well as natural processes (of­ten aided by human actions, such as mining beach sand). Protecting and preserving his­tori­cal ar­chaeo­ logi­cal sites have become the major concerns among many Caribbean archaeologists, and efforts toward educating the pub­lic about their cultural value, pressing for legislation to protect sites where none currently exists, and/or enforcing existing legislation where it does exist will be ongoing priorities for archaeology in the Caribbean for the foreseeable future (Siegel and Righter 2011).

The Postcolonial Period and the Future of His­tori­cal Archaeology in the Caribbean The evaluation of sites and structures that are threatened will soon lead archaeologists to consideration of a period of archaeology that has received little attention to date. That is, the archaeology of the postcolonial period.

206 / Farnsworth

For Haiti and the Domi­nican Republic, the archaeology of the last two centuries is essentially postcolonial archaeology, although there has been little attention paid to it. For Cuba it has been over a century since independence from Spain. For many former British colonies, independence was granted in the 1960s or early 1970s, so that they have either passed, or are approaching the half century of their independence. For those islands that did not receive full independence from Britain, as well as for the French, Dutch, and Ameri­ can islands, significant measures of self-­government were granted around the mid-­twentieth century. For many archaeologists, especially those used to heritage management in the United States and its territories, 50 years marks the age at which sites are considered to be historic, and while the historic preservation legislation in the different islands may or may not recognize the significance of this particular time marker, for his­tori­cal archaeologists the immediate postcolonial period has become, or will soon become, a legitimate topic for research on many islands. This opens up a range of potential new research avenues and topics as well as sites built or occupied in the immediate postindependence period. Importantly, it opens up research arenas and topics that are likely to be of more direct relevance and interest to today’s residents than the painful memories of enslavement, colonial exploitation or neglect, and economic servitude. However, defining what archaeology can contribute to the study of the mid-­to late twentieth century that cannot be found in documents, oral histories, and other twentieth-­century media will certainly not be easy, and persuading others both in academia and the pub­lic that it has any value may be harder still. For example, the Anguillan Revolution (1967–1969) is probably considered the most important event in Anguilla’s history by the current population. However, I am currently struggling to find ways in which archaeology can contribute to our understanding of the Anguillan Revolution and its impact. But finding ways that his­tori­cal archaeology can contribute to studies of the postindependence period will be vital in order to continue to contribute to this vibrant region of the world.

Closing Thoughts It is true that there are still many sites and subjects that can be studied ar­ chaeo­logi­cally from the colonial period, and I am not suggesting that archaeologists abandon that research. In many cases, local stakeholders have been involved and remain interested in these topics. However, archaeologists need to respect the interests of local stakeholders, as much, if not more, than our own private research agendas. If we are to successfully protect ar­chaeo­ logi­cal resources through­out the region we need the support and interest of

Current and Future Directions / 207

the people who live there. Only then will governments act to preserve and protect ar­chaeo­logi­cal sites and direct some of their limited resources to doing so. By opening ourselves up to the study of the postcolonial (or self-­ government, where appropriate) period, archaeologists can forge relationships and alliances with Caribbean people and their governments. By studying sites and subjects from whatever time periods that are of relevance to local stakeholders, and involving them in the research process in as many ways as is possible, archaeologists can build local coalitions to support research and heritage preservation as well as develop research questions and programs that respond to and reflect regional interests. In doing so, archaeologists may well find that fascinating new topics or questions are raised that had not occurred to them to study ar­chaeo­logi­cally before. This will be difficult. But, just as we must learn how best to use the technologies of “new heritage” to further ar­chaeo­logi­cal research and communication of our results, we must learn to adapt our research foci to the needs of the postcolonial/self-­government populations of the region and meet the demands of the local stakeholders whose islands we have the privilege of studying.