Resistance and Caribbean Colonialism

In framing the discussion to follow, it is important to state that anthropologists and archaeologists have repeatedly shown that colonialism was (and is) not a one-way process of acculturation in which dominant ideals are adopted by the dominated. For instance, the “hidden transcripts” of the weak both subtly and sometimes openly challenge domination (Scott Reference Scott1990). Controlling “strategies” of production can also come to be recast as “tactics” of resistance through peoples’ “ways of using the products imposed by a dominant economic order” (de Certeau Reference de Certeau and Rendall1984:xiii; emphasis original). Thus, the everyday—the purview of archaeology—is a central arena for negotiation of power relations.

Archaeologists have described situations of domination in which Native agency is nonetheless a powerful force shaping colonial interactions (Silliman Reference Silliman2001a, Reference Silliman2001b, Reference Silliman2004), and colonized agents can be shown to have incorporated new material culture into their lives through systems that made sense to them (Dietler Reference Dietler and Cusick1998). In the Caribbean, Armstrong (Reference Armstrong1990) recognized aspects of low-fired ceramic forms used by enslaved Africans that are similar to African ones, and Goucher (Reference Goucher and Haviser1999) saw Caribbean continuities with traditional African metalworking, which is highly symbolic and meaningful in West Africa—both suggesting the continuing influences of African sensibilities long into the colonial process. Even when aspects of colonial culture are adopted by the colonized, this cannot be seen as blind following or acculturation. As Singleton points out, “Once African Americans appropriated the material culture of European Americans it became an aspect of black culture” (Singleton Reference Singleton and Cusick1998:183, emphasis original), a point echoed by Wilkie (Reference Wilkie and Haviser1999, Reference Wilkie2000) and applied by Flewellen (Reference Flewellen2019) in the nearby USVI to recognize “self-making” in choices of clothing by both enslaved and free people.

In keeping with these points, we do not suggest that the British plan for Kingstown was a monolithic force that created acculturative acceptance on the part of residents. Just the opposite. Our analysis of the specific landscape over which British ideology was cast seeks to break down “a generic” archaeology of colonialism that risks “over determining the power of the colonizers’ ideologies of landscape” (Hicks Reference Hicks2007:2). Given the results presented below, it seems clear that the Kingstown people negotiated a different path given that they successfully flourished as a community for a generation or more—something we suggest would be extremely difficult under the British plan.

Instead, we suggest that the analysis of the British plan for Kingstown allows us to understand a particular episode of colonial negotiation of the end-period of slavery in its context on the landscape. It is important in its own right to recognize the diversity of experiences of colonialism and of resistance (Odewale Reference Odewale2019), but Kingstown also reveals an inflection point that aimed to maintain British structures in the long term. Although the daily individual-scale negotiations of this system by the Kingstown people will require further archaeological work to recover, what can be analyzed from the sources at hand are (1) the limitations of the British ideals when put into practice in the BVI and (2) their tendency to introduce inequality among the Kingstown residents if they had they been implemented as designed.

Peasantry and Paternalism for Captured Africans

In 1807, the British ended the slave trade but not slavery. People were still to be treated as commodities, although those already bound could only be moved within—not brought into—British colonies. This of course led to smuggling, which, when detected, led to the confiscation of the ships involved. But what to do with the people on those ships, brought to the New World enslaved but now legally free? Over the six decades following 1807, 500 ships were taken and 40,000 Africans came to British colonial territories as nominally free people surrounded by a racist, slavery-based economy and society (Adderley Reference Adderley2006:2). Many were “sent” to service in the British Navy or Army, while other large groups were settled in the Bahamas and on Trinidad. Between 1807 and 1823, more than 1,000 individuals ended up under the control of the Collector of Customs at Tortola, BVI. These people were given a variety of names in records and recent scholarship, and the term “captured” will be used here, but contemporary British authorities called these people “liberated Africans.” This term is somewhat ironic given the various kinds of bondage to which they were subjected outside the framework of legal slavery, and it also reveals the perspectives of colonial officials, who saw them as having received a “gift” from the British state and falling under its paternal “protection.”

Historian Stephanie Camp traces the rise of such paternalism in the U.S. South: “In the decades after the [American] Revolution, proslavery ideology shifted subtly from the patriarchalism of the colonial period to paternalism, a form of social control more consistent with the humanitarian ideals of the age” (Camp Reference Camp2004:17). This attitude is well charted in the archaeology of British colonialism in the nineteenth century—for instance, being traced (along with successful resistance to it) in the Bahamas (Wilkie Reference Wilkie and Farnsworth2001; Wilkie and Farnsworth Reference Wilkie and Farnsworth2005) and described as far afield as South Africa (Lucas Reference Lucas2004:127–129). In this system, as Higman (Reference Higman1998:65–66) explains, for the colonial British, the “peasantry” represented a stable supply of labor—workers too poor to provide substantial opposition but having long-term residency, as opposed to tenants, who were less rooted in an area (and perhaps, therefore, more prone to remove their labor) and rarely supplied wage labor to plantations. Anthropologist and BVIslander Michael O'Neal provides an extensive critique of this concept and its application to the BVI in particular (O'Neal Reference O'Neal2012:65–71) and argues for the more neutral term “smallholder,” which avoids some of the pitfalls of Eurocentrism and generalization. Our use of the term “peasantry” here does not take issue with his analysis showing that such a group did not develop in the BVI, but it reflects the goals of the British colonial state under analysis. More than simply paternalism, the British looked at the arrival of these nominally free Africans in the Caribbean as an opportunity for a social experiment: “a population with which Britain might rehearse for the realities of Caribbean colonial life after slavery” (Adderley Reference Adderley2006:126–127). Quite explicitly, the goal of this experimentation included both “how to exploit the labor of Africans” after slavery and “efforts to ‘civilize’ the liberated Africans” (Adderley Reference Adderley2006:126–127). They were intended to compose a new “peasant” class maintaining British hegemony.

Clearly, however, as with those held in legal bondage, the captured Africans brought their own ideologies, goals, and plans to this interaction. For instance, running through the story of the Jamaican plantation of Montpelier, Barry Higman (Reference Higman1998) sees “a set of competing models of the appropriate relationship of people and land, people and place. Essentially, these models may be reduced to a long-term struggle between communalism and individualism” in which the planters, “a class of land-takers,” hung on to individualism through the force of the state, while the enslaved and later emancipated people of African descent “shared a communal, collective concept that was sharply opposed to the English-based legal formulation of free-hold tenure” (Higman Reference Higman1998:290–291, 303). Although Higman did not necessarily mean for this pattern to be generalized throughout Caribbean history, elements of these differing perspectives on land and land use can be traced in the present example. Historian Isaac Dookhan (Reference Dookhan1975:98) suggests that, on the whole, the British government appears to have honestly worked toward the “general welfare” of the captured Africans. Although this may be so, the cultural distinctions just described are central here, and the steps taken by colonial authorities must be recognized as culturally placed: each side had a different definition of “welfare.”

In short, then, the captured Africans of Tortola were seen by British authorities as a potential peasantry, who should be made to embrace individualist, capitalist modes of production as a way of proving the ability of African people in general to take their proper (subservient) place in a postemancipation British Empire. Aside from the different models of land-human interaction and social relations outlined above, the land itself has a role to play. The landscape of Kingstown and colonial maps laying out what this paternal, peasant ideal should look like allow us to examine this “experiment” and uncover at least some aspects of how it could have played out that are unrecorded in the written record. Mapping is itself a means of establishing and projecting colonial control, but this analysis also engages with the mapped landscape itself. In doing so, we aim to expose flaws in the British plan as a way of uncovering the potential paths of resistance in daily life open to the Kingstown People as well as barriers to their maintaining a stable community.

Kingstown, Tortola

Between 1807 and 1823, several shiploads of captured Africans arrived at Tortola, which was then a marginal and poor colony undergoing the slow collapse of the sugar economy. The history and archaeology of the BVI, of which Tortola is the main island, are better covered elsewhere (Chenoweth Reference Chenoweth2017; Dookhan Reference Dookhan1975; Harrigan and Varlack Reference Harrigan and Varlack1975; O'Neal Reference O'Neal2012; Smith Reference Smith2009), but it is relevant to note here that a sugar-based slavery economy developed later here and was always more fragile than elsewhere in the British Caribbean. By the end of the nineteenth century, plantation agriculture was entirely dead in the BVI, and it was replaced with a nearly universal pattern of small, more-or-less subsistence farms owned by the descendants of those previously enslaved.

The precise number of captured Africans who landed on Tortola differs between various contemporary accounts, but it was upward of 1,300. Parliamentary Commissioner John Dougan reports that approximately 28 ships illegally carrying enslaved people were seized at or near Tortola and brought to the Vice-Admiralty court there between 1807 and 1823 (House of Commons 1825:5). About 1,070 of these captured Africans came from four ships flying the Spanish flag—Venus, Candelaria, Manuela, and Atriviedo, all captured in 1814 and 1815—although about a quarter died shortly after arriving on Tortola, a rather shocking figure blamed on an intestinal infection passed around on the ships. Another near-third of the group was “sent” for military service, after which their trail seems to end.

Much of the remainder of the group was “apprenticed.” As with the period immediately following emancipation, also called “apprenticeship,” the idea of this innocuous term was that, like youths training in trades, the Africans would work for planters to earn their keep and, in doing so, learn skills that would allow them to function as free people. This plan, however, was fatally flawed in several ways. Aside from the obvious disadvantages of life as a person of African descent in the slavery-based Caribbean, Dookhan (Reference Dookhan1975:98) points out that these indentures were decided without any consultation of the people themselves, and their unusually long 14-year term resulted in individuals being treated more like enslaved people than free. Moreover, their status as free people paradoxically threatened them—because “their deaths constituted no financial loss to those responsible for their welfare” (Dookhan Reference Dookhan1975:100), they were often treated worse than enslaved people. This trend was perhaps worsened by the fact that many Tortola planters were themselves in dire financial straits, and it was those planters most unable to support apprentices who often accepted them, being in need of the extra labor.

The prospects of these captured Africans were worsened by tensions that existed between them and both the free Black and enslaved communities. The former considered the new arrivals lower class because they were unable to gain any employment other than manual labor, which was usually associated with enslaved status, whereas the latter were reportedly jealous of their nominal freedom (Dookhan Reference Dookhan1975:102). Dookhan's gloss may represent, to some extent, the view of whites more than the Africans themselves, as he also reported that a scheme to transfer the group to Trinidad failed because almost all refused to leave Tortola, having formed relationships with—and many having had children with—others who were still enslaved there (Dookhan Reference Dookhan1975:103). Regardless, by the time the indentures of the surviving captured Africans of 1814 and 1815 began to expire in the late 1820s, colonial authorities thought that they would stand little chance of survival on their own due to their social position, the poor state of the economy in the BVI in general, and the competition of unfree labor undercutting any efforts to find employment. Some alternate plan was needed.

Under the 1807 Slave Trade Abolition law, the Collector of Customs was responsible for dealing with the human “cargo” of seized ships. An 1830 letter to London from Tortola Collector Robert Claxton speaks positively about the group, noting that since release from their indentures, none had been arrested for a crime, and he laments that their position as farmers was precarious because of “the very doubtful tenure of the lessor of [their] lands” and their lack of recourse when their crops were damaged by “the trespassing of stock on their grounds” (House of Commons 1831:25). Claxton suggests the purchase of land to be “parceled out” to the captured Africans.

The goal of at least some members of the British colonial machinery was the “most economical mode of disposing of those persons” (House of Commons 1831:26), but the depressed state of the economy in Tortola made the purchase of land more economical there than on other islands. By March 1831, the Treasury had approved the sum of £1,025 to be spent on the purchase of 110 acres of former plantation lands in two parcels, and £5 worth of building materials for each of the 100 people to be settled (House of Commons 1831:26–27). It was made explicit that the Crown would continue to own the land and that the residents would be forbidden from transferring or selling it.

As the project commenced, the number of people to be settled on the land now known as “Kingstown” grew to nearly 300. Further grants totaling £1,300 were made in order to purchase additional land and materials (Dookhan Reference Dookhan1975:108). Several maps of the settlement were made, including the one analyzed below (Figure 1), showing the total area of the site to be about 180 acres. Twenty of these acres were used for a village, which had about 90 structures depicted on the map, including a school (apparently the old plantation house) and an Anglican chapel—St. Philip's—begun in 1834, the ruins of which still stand and were partially restored in the mid-2010s. The village clusters on the more-or-less level ground by the sea around a small brackish pond (now filled in) and the main road to Road Town, the colony's main settlement. Sixteen acres of the site were reported to be “high woods and fruit trees,” and 40 were reported to be “Guinea grass and unoccupied” (UK National Archives CO 700/VIRGIN ISLANDS/6; see Figure 1). The remaining 106 acres were divided into what were thought to be equal one-acre lots, each assigned to a particular family by lottery.

Figure 1. “Plan of Kings-town, the location of the liberated Africans in Tortola” (UK National Archives, CO 700/VIRGIN ISLANDS/6. Printed by permission of the UK National Archives, Kew).

The community appears to have survived for some time, although the archival records are limited, a reflection of both the overall colonial disinterest in the BVI (Chenoweth Reference Chenoweth2017:26–32) and the marginal success of this “experiment” for furthering British colonial goals. We do know that Kingstown had 49 day students in school in 1835 (although the average daily attendance was given at 25) who learned the “common branches and needlework” (House of Commons 1837:92–93). The August 1837 hurricane damaged much of the settlement, and the schoolhouse, which was listed as “substantially built,” was used as temporary shelter (House of Commons 1837:144). Yet, the community continues to appear in records, for instance, when the school and church both had repairs to their roofs in 1862, at the expense of the government (House of Commons 1864:94). The end of the community is unclear, but it appears that it was largely unoccupied by the turn of the twentieth century.

GIS Modeling

The map of Kingstown in Figure 1 is deceptive, presenting a more or less uniform space available for appropriation and cultivation. Figure 2, though, makes clear a defining feature of the BVI landscape: it is extremely steep. Tortola is rarely more than 3 km wide but rises to over 500 m (1,700 ft.) at Mt. Sage. In about 1 km (3,300 ft.), the land near Kingstown rises from the sea to a peak of nearly 300 m (985 ft.). The land is also uneven in terms of farming potential: although prime farming areas such as Cane Garden Bay exist, much of the island has shallow, sandy soils, and some areas are bare rock.

Figure 2. Kingstown, Tortola, from the air in 2013 (photo by John M. Chenoweth). (Color online)

The remainder of this article, then, seeks to understand how the spatial layout of the site and its landscape interacted with the colonial plan and how it may have impacted the daily lives of the captured Africans who made it home. In the next section, we explore the question of the differing agricultural potential across the site, but first we consider limitations imposed by the question of access to these fields—a central feature and limiting factor of daily life. This can be productively explored indirectly through artifact analysis but also by examining the landscape itself (Bates Reference Bates, Bates, Chenoweth and Delle2016). At Kingstown, due to the ideal of individualism inherent in the plan and the distribution of separate one-acre lots, those members of the community who were assigned fields just above the village would have been able to travel to their field quickly and easily, but those whose fields were farther away would have had more difficulties. Although it is schematic, we suggest that it is possible to quantify those differences and explore their potential implications.

In order to address these questions, the 1831 map was georeferenced and laid over the USGS Global Data Explorer 30 m Digital Elevation Model (DEM) of the area so that human movement could be studied. Although it is etic, such an analysis can speak to the physical limitations on time, productivity, and energy imposed by the landscape if used as represented in the British plan. This model and the available data have limitations connected to the resolution of the elevation model, potential modifications of the landscape, and certain assumptions used in the model, which are discussed in detail in the supplemental material available with this article (see Supplemental Text 1). We feel that these limitations do not detract from the general conclusions we draw.

Least Cost Paths and the Kingstown Fields

The generation of cost surfaces has been receiving increasing attention and application in archaeology (Hare Reference Hare2004; Herzog Reference Herzog2014; White and Surface-Evans Reference White and Surface-Evans2012). Distances between two points can be measured easily, but of course such measurements “as the crow flies” cannot account for barriers, hills, or other impediments to travel. As described most completely in Historical Archaeology by Edward González-Tennant (Reference González-Tennant2016:30), the production of a cost surface (a raster or grid of values representing the “energy required to move through an environment”) allows for the calculation of least cost paths (LCP)—the most efficient and therefore “most likely routes from one point to another across a landscape.”

As noted, the 1831 Kingstown map is deceptive in precisely this way: travel along the beach, into the hills where the fields were located, and back downhill to the village are all quite different propositions and are not differentiated by a simple measure of linear distance. ArcGIS was used to calculate the LCP from the Kingstown village (the centrally located St. Philip's Church was used as a common starting point) to each of the 106 fields represented on the 1831 map (Figure 3), producing a different path to reach each field, including the path length and elevation gain. Table 1 charts the actual distances traveled over the surface as it rises and falls, showing not only a substantial difference between the extremes, but also a wide distribution of journey lengths, which average 815 m but range between just 124 m to over 1.5 km. The difference between these trips was not only in length but also elevation: whereas some fields were just 15 m above the level of the village, others required an elevation gain of up to 220 m (Table 1).

Figure 3. Least cost paths from the center of the Kingstown village to each of the fields depicted on the 1831 map (map by John M. Chenoweth).

Table 1. Least Cost Distance, Elevation Gain, and Energy Usage for the Journey from Kingstown Village to Each Field Depicted on the 1831 Map (along with Measured Field Area and Mean Normalized Difference Vegetation Indexes [NDVI] Values for Each Field).

Note: The fields are named according to the map, with the letters representing the respective names of the original plantations combined into Kingstown: Dyer's Side Plantation (D), Gordon's Plantation (G), and Highland Grove Plantation (HG).

Time, Energy Costs, and Implications

When considering how people in the past would have actually related to this landscape, however, time might be a more meaningful unit than distance and elevation. Each person had only a limited number of daylight hours. A time-based cost raster for the area was calculated based on Tobler's Hiking Function (Tobler Reference Tobler1993), which models speed of movement as a function of slope. A map was produced with anisotropic contours—sometimes called “isochrones”—representing hiking distance at speeds of about 5 km per hour over level ground, adjusted accordingly over slopes (Figure 4). In effect, starting from the center of Kingstown village, it should take approximately the same amount of time to reach any point along any given contour shown in this image. Some of the farthest fields would take more than a half hour of steady, uphill hiking to reach, whereas those just uphill of the village could be reached in less than five minutes. Return travel downhill would have been easier, but the journey from the farthest fields would still have taken 20–25 minutes (Figure 5).

Figure 4. Map showing walking time from the center of the Kingstown village in 5-minute isochrones (map by John M. Chenoweth).

Figure 5. Map showing walking time on the return journey back toward the center of the Kingstown village in five-minute isochrones (map by John M. Chenoweth).

The figures given above for travel distance, elevation, and time have deeper meanings than simply the annoyance of long commutes. They directly impacted the ability of farmers to make ends meet—already a difficult task for a subsistence farmer in the marginal Caribbean—and these difficulties can be specified using the data just discussed. Roundtrip travel to the farthest eight fields at Kingstown totaled about 52.5 minutes, while the nearest five had a roundtrip commute of only about eight minutes. The BVI has an average of just over 12 hours per day of daylight, although the shortest days of the year are about 11 hours. Anyone who has navigated its steep, rocky slopes as night quickly descends will understand that even in familiar settings, it can be treacherous in the dark, and it is assumed that travel during the daylight would have been preferred. Those traveling to the farthest fields consumed nearly an hour in their daily commute, giving them an average of about 92.7% of the daylight available to those with the closest fields, even if they had left their homes at dawn and returned at dark.

Distance and elevation traveled can also be brought together in a single measure of energy expenditure in kilocalories. See section 2 of Supplemental Text 1 for the details of how this calculation was conducted. As with distance, elevation, and time, energy consumption for roundtrip travel to the fields varies widely (Table 1)—from 356 kcal per trip to only 27 kcal. A person traveling to the farthest, highest field assigned to the Kingstown people would—just by getting to her or his allotted ground and not including any energy used to farm it—have needed an additional 329 kcal per roundtrip journey compared to the person farming the nearest field. On a per-year basis, assuming an average of four trips to the field per week all year round (the other days being used for fishing, craft activities, or rest; this figure is supported by discussion in section 3 of Supplemental Text 1), this could be a difference of more than 68,400 kcal in energy expenditure—about 5% of the total calories needed in a year by an adult performing heavy labor, such as farming.

To put this into more meaningful terms, we can translate these raw energy figures into food: sweet potatoes, a common and nutritionally dense provision crop, were grown in the area prehistorically and are documented in the BVI in the twentieth century, so they were probably available to the Kingstown people. The difference in energetic travel cost between the farthest and nearest fields at Kingstown converts to approximately 76 kg (167 lbs) of sweet potatoes per year (see section 3 of Supplemental Text 1 for more on this calculation). In terms of time and energy usage, then, a farmer with the farthest fields has to produce substantially more food to counter energy expenditure and yet has only about nine-tenths the time to accomplish this. This information should be further contextualized by the fact that any person who has, in the course of a day, hiked for an hour with over 200 m in elevation gain will consequently have substantially less energy available to work.

Work reconstructing prehistoric sweet potato yields, also discussed in Supplemental Text 1, shows that some of the one-acre lots at Kingstown likely could not have produced enough food for a family of three. It should be noted that in most ways, these calculations assume the best case scenario for the captured Africans if they were attempting to conform to the British colonial plan. Various social and practical factors—discussed further in section 4 of Supplemental Text 1—would likely have prevented these “best cases” from being achieved in daily life at Kingstown, suggesting that these results probably underestimate the disadvantages of the upper fields and their productive limitations relative to the lower ones. Results therefore suggest that the average resident of Kingstown would have had a difficult time making ends meet and that there was a wide deviation from average. Limitations of time and energy, imposed by the differences in field placement, would have introduced substantial differences. The Normalized Difference Vegetation Indexes (NDVI) section of this article further explores these divisions through a consideration of the fields themselves and the productivity of each parcel.

NDVI and Analysis of Satellite Images

We can further consider the plan as represented in the colonial authorities’ map by examining the nature of the fields themselves. One distinction between the one-acre lots is that, when measured in GIS, they were rarely one acre (4,047 m²), with some lots as little as half an acre or less, and wide variation: the standard deviation of the measured lots is 833 m², or about one-fifth of an acre. Some of this may be due to inaccuracies in both nineteenth-century mapping and modern georeferencing and measuring, but it is still clear that these lots were far from equal in size. Moreover, it is obvious that some land is more fertile than other, and this is particularly true in the British Virgin Islands, where some areas can be excellent for agriculture, whereas others are bare rock. In order to analyze the fertility of the fields in Kingstown, high-resolution, multispectral WorldView-2 satellite imagery (obtained through a grant from the Digital Globe Foundation) was used to calculate NDVI—a measure of plant health and, consequently, soil conditions—with higher numbers suggesting good growing potential. This process is described in section 5 of Supplemental Text 1.

Table 1 shows the average NDVI value in each of the Kingstown fields. Again, there is wide variation between the fields, both in terms of size and average NDVI values. It is difficult to convert NDVI values directly into more meaningful figures, as was done above in the case of kilocalorie usage and food productivity to account for both elevation gain and distance in accessing fields. To put these numbers into context, therefore, we also ran NDVI figures for several effectively unfarmable areas on BVI out islands pictured in the same dataset, and this produced NDVI numbers ranging from just under 0.5 to about 0.6. Several of the Kingstown fields are little better than these figures (Table 1). By contrast, NDVI values for an area of Cane Garden Bay on Tortola—which, as the name suggests, has long been known as excellent farming land—are just under 0.9. Some of the Kingstown fields show strong potential for farming, with 16 having an average NDVI pixel value of 0.85 or higher.

Perhaps the best way to assess the data is to combine NDVI values with the calorie-cost figures derived above, incorporating both access and productivity into a scatterplot (Figure 6). Some fields at Kingstown would have been relatively easy to access and reasonably productive, shown in the lower-right quadrant. Yet others (upper left) would have been much more costly to access and at the same time less productive. It is interesting that there are few fields that are both unproductive and easy to access (the lower left). This may be because fields closer to the village would have also been lower on the slopes of the mountain, where soils would have been deeper and the moisture retention greater. Together, these datasets reinforce the suggestion above that the individualism of the British colonial model, in the separate working of apparently equal plots, would have necessarily produced inequality rather than provided equal opportunities.

Figure 6. Scatterplot of NDVI values and kilocalorie usage values for roundtrip access to each of the Kingstown fields.