reading response

Louis P. Nelson

Buildings & Landscapes: Journal of the Vernacular Architecture Forum, Volume 21, Number 1, Spring 2014, pp. 88-125 (Article)

Published by University of Minnesota Press DOI: 10.1353/bdl.2014.0006

For additional information about this article

Access provided by University of Virginia Libraries __ACCESS_STATEMENT__ (Viva) (2 Sep 2014 16:09 GMT)

http://muse.jhu.edu/journals/bdl/summary/v021/21.1.nelson.html

88 | 

louis p. nelson

Architectures of West African Enslavement

The gleaming white castle rises dramatically from the promontory overlooking the coastal town of Dixcove, Ghana, much as it has for more than three centuries (Figure 1). Visitors arrive in town and park at the base of the hill, climbing to the castle on foot. The path wends up the side of the hill and diverts to the west toward the small parade ground that stands between the front of the castle and the sea. Two massive diamond- shaped bastions stand to either side of a heavily rusticated door, which opens through the solid, unbroken masonry of the outer wall into the front courtyard of the castle. In the back corner of an interior courtyard is an arched opening into the northeast bastion, accessed by a range of three steps. The small chamber behind is closed by a heavy iron gate (Figure 2). No more than two hundred square feet, this was the slave hole. Merely 2 percent of the total square footage of the whole compound, this small cell was a criti- cal component of the sequences of such spaces in English castles along the west coast of Africa. It played a vital role in the highly lucrative system of transatlantic slavery that defined the coastline of West Africa and that sustained the sugar pro- duction of the British West Indies from the late seventeenth century to the abolition of the slave trade in 1807.1

The slave holes of the English castles lining the west coast of Africa were and are horrifying.2 “There was nothing to be heard but the rattling of chains, smacking of whips, and the groans and cries of our fellow men,” reported Ottobah Cugoano.3 But these chambers were not the only spaces experienced by enslaved Africans in the

long journey from Africa to the Caribbean. While historians have dealt extensively with many di- mensions of the slave trade—and this article de- pends on the published scholarship of a number of excellent historians—architectural historians have left the spaces of enslavement largely un- touched. This article situates the historical nar- ratives of the slave trade not in buildings but in spaces, some permanent, some temporary, some floating, and some created by implements of bondage. This understanding of space depends on that formulated by Henri Lefebvre in which space is at once physical, socially produced, and imagined.4 When possible, physical spaces have been documented, recorded in their surviving conditions as artifacts of slavery.5 But these spaces were very much the product of those social and economic relationships governing the capture (and resistance to capture), containment, trans- fer, and sale of people. They are also products of the British imagination, changing as their in- tended function changed and changing relative to other institutional architecture. Understood as agents in the economic and social relationships of exchange, these spaces—in buildings and in canoes—are components of a machine of pro- duction dedicated to the generation of “the slave” as a social, historically contingent and situated subject.

In tracking sequences of spaces, I do not pre- tend to suggest that this article reports a typical experience; variations over space and time and among personal dispositions mean such an at- tempt would be folly.6 But the telling that follows was in its component parts the reality for many

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 89

who were transported against their will from Af- rica to the Caribbean by British slavers, a term used historically to identify both the men who enslaved and the boats that carried their victims. Based on fieldwork in Ghana and on a careful examination of documentary and visual records in England, Ghana, and the United States, this article reconstructs the spatial experience of the enslaved, examining when possible not just the physical spaces but also the spatial experience of the senses, so powerfully captured by Cugoano.

By engaging sequences of spaces, this article embraces four methodological convictions. The first is that the meanings of spaces are not always determined by their physical production. Some- times the experience of a space, if such an experi- ence can be reconstructed, is far more important than its material making. Another conviction is that the writing of history benefits from the rigor- ous examination of the spaces where events took place; too frequently, historians generate disem-

bodied narratives that divorce events from place. And this article also presumes that the most im- portant meaning of architecture is often found in the various human networks—social, economic, political—that tied those spaces one to the next. Derived from these three convictions, this article is a history of architectural experience.

The fourth and last methodological convic- tion is potentially the most important. West Af- rica’s coastal castles and their associated spaces are rightly understood as spaces engaged in the economic processes of slave making—of trans- forming a person into a commodity. It is no ac- cident that the definitive study of the Royal Af- rican Company—the organization that oversaw England’s African trade until 1750—was written not by a social or a political historian but by a historian of business in a series on emergent international capitalism.7 The decisions made by Europeans along the west coast of Africa and by their African partners were motivated by

Figure 1. View of the castle at Dixcove, Ghana. Photograph by Louis P. Nelson, July 2012.

90 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

economic self-interest, and over time it became increasingly clear that the highest profits lay in the production of slaves. This economic machine matured through the eighteenth century, gen- erating finely tuned processes of exchange and function-specific architecture to support that machine.

Understood as an economic process, slave making had a number of important factors that drove the decisions of enslavers. One was valua- tion; central to this process was the recognition that individuals became more valuable as they

traveled further in the process from capture to final sale in the New World. Allied with this was the recognition that some goods had higher value potential over other goods; in slave making these criteria were typically gender, age, and health. A younger healthy male had far greater value potential than an older sickly female. Another reality was that damaged goods were less valu- able, creating an incentive to protect investment from damage. While deeply inhumane, these were the factors that drove many decisions in the early modern West African trade; to ignore this reality is to write poor history. But it is also poor history to write about these spaces under the presumption of isolation. In some ways, the trad- ing castles transformed in the later eighteenth century to support the slave trade corresponded with European prisons, schools, and institutions of reform, which also transformed and matured in these same decades.8 The fundamental differ- ence was, of course, that prisons and schools ap- pealed to the humanity of their subject to enact reform (however insufficiently), whereas the cas- tles in West Africa and their allied spaces were more concerned with the economic processes of commodification.9 These differences will be- come clear as the narrative unfolds.

Capture and Coffles The first act of enslavement was capture. En- slaved Africans were sometimes the result of wars. In 1682 Barbot reported that one English trader recently purchased three hundred Afri- cans “almost for nothing, besides the trouble of receiving them at the beach in his boats.” This was a huge number, the result of a recent war: “The Commenda men brought them from the field of battle, having obtained a victory over a neighboring nation, and taken a great number of prisoners.”10 Nonetheless, as recent histori- cal scholarship has made clear, most Africans sold into slavery in the eighteenth century were not legitimate prisoners of war but the result of internal legal processes, debt, famine, or, most commonly, kidnapping.11 One English observer wrote, “We often saw several thousand men in full war paint march inland where they stormed the villages, set the houses ablaze and caught

Figure 2. View of the slave hole gate at Dixcove Castle, Ghana. Photograph by Louis P. Nelson, March 2011.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 91

the inhabitants.” Francis Owen, another English trader, described how an African trader “went out with his troops, set three sides of a village ablaze, placed his soldiers at the forth side to grab those inhabitants who ran out of the flames.”12 Especially as the eighteenth century unfolded, kidnapping became the primary mode of slave production across West Africa as African trad- ers worked to meet the escalating demand for enslaved Africans from their European trading partners.

Surviving African accounts of capture are rare. That written by Ottobah Cugoano, pub- lished in 1787 as part of the emerging critique of the slave trade, offers a rare glimpse into the experience of an African boy.13 Playing with friends, Cugoano was assaulted by “ruffians”

who claimed the boys had “committed a fault against their lord” and must travel with them to answer for this fault. Threatened with weapons, the boys submitted and walked for two days to what was likely a market town, where they were divided among the houses of locals, colluders in the plot. After spending six days in a house, Cugoano was eventually conveyed under the direction of another trader to a town populated with white men near the coast. The next morn- ing, the trader took him into the castle under the presumption of obtaining some trade goods, but it was in the castle that young Ottobah came to fully understand his kidnapping.14 Asa-Asa, an- other African boy, relayed how his hometown was attacked and burned by invaders. They pillaged for two days and then carried off all the survivors

Figure 3. Map of the West Coast of Africa identifying the locations of various sites mentioned in the text. Drawn by Jason Truesdale.

1 2 3 4

5

Gambia River

Cross River

Benu e Rive

r

Lake Volta

Black Volta

White Volta Red Volta

Senegam bia

Sierra Leone

Windward Coast Gold Coast

Bight Of Benin

Bight Of Biafra

Nig er R

iver

Kumasi

Salaga

Sankana

Gwolu

Idah

Timbuktu

Old Calabar

Ouidah

1 Nzulezu 2 Dixcove 3 Cape Coast 4 Anomabo 5 Accra

Bauchi

Atlantic Ocean

92 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

in a coffle of about twenty.15 It is not surprising that these accounts came from children. The increasing numbers of older children sold into slavery over the course of the eighteenth cen- tury has led some historians to see kidnapping of youth supplanting war as the primary means of introducing persons into the slave trade mar- ket, especially as the value of adolescent males— unwounded by war—surpassed the value of all other categories of potential slaves.16 The result was the delivery of millions of Africans to the coast, some as the result of war or crimes in their village but, as the century unfolded, most deliv- ered to coastal factories by armed raiding parties through a method called “grand pillage,” which simply meant people stolen from their homes as they slept, worked, and played.17

Many African communities fiercely resisted raiders.18 Oral histories from the Kasena, Bulsa, and Chiana peoples have preserved stories of at- tack and resistance against African slave traders.19 The Mfantses still use a greeting that echoes re- sistance to slaving: “If there is someone chasing you, let us know so as to hide you.”20 Victim com- munities also enacted architectural strategies of resistance. The Daagare people of Sankana have an oral tradition that reports cow horn alarms were used to warn nearby villages of a slaver attack. That same community also points to a series of three caves that were used as places of refuge.21 With appropriate provisions these caves provided sanctuary for weeks at a time. One of their songs begins, “When the raiders came, the people of Sankana stood firm” (Figure 3).22

The people of Gwollu offered a different strat- egy for defense: they built a wall encircling their village. The village elders reported that the wall “was constructed at the time of the notorious raiders,” in response to “the disappearance of people.” They found that their people were being trapped in nets in the forest. But then the raid- ers began attacking the village, coming at night, usually burning houses to drive out the inhabi- tants. The village elders offered a number of solutions, including (1) building roofs of baked bricks that could not burn as easily; (2) installing impediments to block or trap an intruder; and (3) installing narrow doors, often low enough to

require bending over to gain entry, that would allow only a single person access to the interior. Yet even after these were put in place, the raids continued. So the final solution was to erect a vil- lage wall with small holes for surveillance. After two years the wall was complete and brought to the village “limited peace.”23

Sometimes, resistance resulted in even more drastic measures. Research in Dahomey (now Benin) has determined that the region’s many lacustrine (waterborne) villages were part of a defensive strategy of relocating whole villages to the edges of lakes that were both visually shel- tered and largely inaccessible by land, thereby more easily defensible.24 The appearance of such lacustrine villages across West Africa, from Benin through at least western Ghana, suggests the possibility that this practice was widespread (Figure 4). Such strategies required, of course, learning new architectures and embracing dif- ferent foodways. Isolation meant safety. From building walls to building water villages, Afri- cans responded to the raids of the slavers. Ac- tively attending to these architectural strategies reminds us that capture was itself a stage in the process and that victim communities were ac- tively engaged in acts of resistance. There was an architectural history to African-on-African and, eventually, European-on-African violence.

Once captured, the enslaved African was bound to a coffle in the forced march from the point of capture to trading posts along the coast. A remarkable early representation of such a gang appears in a watercolor sketch drawn by Captain Samuel Gable into the logbook of the Sandown, an English slaver ship that slaved in Sierra Leone in 1793. Entitled, in part, “Representation of a Lott of Fullanis bringing their slaves for sale to the Europeans,” the image shows a collection of drivers armed with bows and arrows marching seven enslaved in a gang along the horizon of the picture plane (Figure 5). The implements of restraint are the diagonal yokes that reach from the waist of one captive to the neck of another. Other slaves are carrying large baskets on their heads, likely filled with food and essential sup- plies or sometimes other trade items like ivory tusks, which had to be carried with the coffle dur-

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 93

ing the days, weeks, and sometimes months of a march.25

Such yokes were the target of much criti- cism among abolitionists by the late eighteenth century. Thomas Clarkson’s Letters on the Slave Trade, published in 1791, includes two plates illustrating three variations on the yokes (Fig- ure  6).26 In the simplest version, Clarkson re- ports, Africans were bound around the neck by two pieces of wood fastened to each other at the ends. Such individual devices prevented escape by creating an impediment to the escapee, who could not quickly run through the dense for- est. A second version shows a single yoke with crutches, one at each end to secure two people. The third has a crutch at one end and a twisted rope at the other, by which it was hung around the neck. “It is reported to be so heavy,” Clarkson writes, “that it is extremely difficult for the per- son who wears it to walk, much less to escape or run away.” The only way to manage the weight

of the log was to rest the crutched end of the yoke on the shoulders of the person before. In Clarkson’s account the straight end of the yoke was tied around the person’s neck, and in the captain’s sketch it appears to be secured around the waist. One such yoke, probably from the early nineteenth century, survives in the collections of the International Slavery Museum in Liverpool. The base of the yoke has an indentation a few inches from the end that would have been used to secure a cord between the yoke and the neck or waist of the captive. The crutched end of the yoke does not have any evidence of such mechanisms for fastening, suggesting that it simply rested on the shoulders of the person before. Some ac- counts indicate that Africans were bound neck to neck in chains with their hands fastened behind their waists or to their sides.27 Bound in one of these ways, most coffles walked from their point of capture to the coast. Evidence of the torture of this march was clear to Carl Wadstrom when he

Figure 4. The isolated lacustrine village of Nzulezu, Ghana. Photograph by Louis P. Nelson, March 2011.

94 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

Figure 5. “Representation of a Lott of Fullonis . . . ,” Samuel Gable Logbook, 1793. Image courtesy National Maritime Museum, Greenwich, England.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 95

noted the “mangled bodies of several [Africans who had just arrived], whose wounds were still bleeding  .  .  . a most shocking spectacle.”28 The march to the sea was gruesome. The estimates of mortality vary widely, but it is absolutely clear that not all who began the journey lived to see the ocean.29

Although evidence is scant and uncertain, his- torians generally agree that while most Africans came from within a few hundred miles of the coast, some traveled great distances. They could hail from as far as Timbuktu, approximately one thousand miles from the closest castles or facto- ries on the Gold Coast.30 The complexity of Af- rican trading meant that the majority of those Africans sold into slavery experienced complex African trading networks that brought captives from across the interior.31 Ghanaian historian Akosua Perbi has determined that early mod- ern Ghana had at least nine major trade routes by which captives were channeled to the coast.32 Most of the interior routes passed through Ku- masi; one of the major spurs from Kumasi led di- rectly to Cape Coast. Those traveling substantial distances likely passed through one of the inte- rior’s many market towns. Perbi has identified as many as sixty-six market towns in Ghana, most of which probably engaged in the trade of slaves.33 As a result, captives became commodities in the complex trading network that spread broadly across west and central Africa. Asa-Asa, for ex- ample, reported that he had been sold six times over the course of about six months, “sometimes for money, sometimes for cloth, and sometimes for guns,” before he arrived at the coast, where he was sold again for exportation.34

The largest and most important market town in what would become modern Ghana was Sal- aga, located approximately three hundred miles inland from Cape Coast. This critical town en- joyed a very important position. Situated between the two branches of the Volta River, Salaga was easily accessible to the coast. The town was the terminus for two branches of the trans-Saharan trade routes and was the market town with the greatest access to North Africa.35 Salaga was also politically important as the major market town under the rule of the Asante, after their capture

of the town in 1744. The Asante were one of the major African tribes that provided Africans for the transatlantic market, and as a result the Sal- aga market became a major point of exchange.36 Recent archaeological investigations have uncov- ered evidence for the courtyard of a large slave warehouse, as well as water dams and cisterns used to bathe slaves before sale.37 As described by Perbi, the Salaga market was divided into two sections, foodstuffs and slaves. The majority of the foodstuffs were sold in the morning, while the sales of people usually took place in the af- ternoons, a lag that gave traders time to organize their “stock” by gender and age and to assess their value. Some period accounts suggest that a town crier announced the daily opening of the slave market, having been paid by traders to launch sales.38 This was an experience often re- peated; many if not most people would be housed in slave camps at some point along the journey.39

Research on the internal trade in the Bights of Biafra and Benin, the regions that would become southeastern Nigeria, suggests that traders in these places also depended on a vast network that penetrated deep into the African interior. The diary of Antera Duke, one of the leading African traders in Old Calabar in the Bight of Biafra, in- dicates that he depended on a trading network

Figure 6. Representation of a yoke as printed in Thomas Clarkson, Letters on the Slave Trade (London, 1791).

96 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

that ranged over 30,000 square miles and that most of his slaves came from market towns near the fringes of that region; many came from the Cameroon grasslands, most having already trav- eled great distances.40 David Northrup’s work on this region suggests that many captives, if not a majority, were marched in coffles from at least as far as Bauchi, a major market town more than 500 miles from the coast, although the last 150 miles was surely by canoe on the Niger or Cross Rivers.41 Many of these enslaved Africans also passed through the market town of Idah, near the critical junction of the Niger and Benue Riv- ers. The market in Idah occurred every ten days, was attended by as many as six thousand people, and witnessed the sale of approximately three hundred enslaved Africans at each market (see Figure 3).42

If travel by land in a coffle was the experience of most enslaved Africans, some were also trans- ported from market to market via canoe. This was certainly the case for those in Gambia and the Bight of Biafra, both regions dominated by major river systems. Fashioned from the trunk of a single enormous tree, canoes, up to eighty feet in length, could hold scores of paddlers, raid- ers, and captives.43 When preparing for a raiding expedition, canoes were fitted out with a range of weaponry, including “two guns, which were three pounders, fixed upon a block of wood; one in the canoe’s stern and one in the bow.”44 Fleets of up to twelve canoes would depart on expedi- tions that could last as long as three weeks, re- turning with as many as thirty enslaved Africans per canoe. These canoes were such a critical component of slaving on Old Calabar that Olau- dah Equiano noticed that some people seemed to live out of their canoes, which were fitted out with “household utensils and provisions of all kinds. . . . [Some] staid and cooked in theirs, and laid in them all night.”45 In such canoes the initial captives of a raid traveled for weeks, as described by one sailor, with “their Arms generally pin- ioned behind their backs with Grass Ropes. They are made to lie down in the Bottom frequently of a wet canoe.”46 In Gambia and the Bights of Benin and Biafra, most of the enslaved arrived at their final African destination via canoe; along

the Gold Coast that arrival was on foot as part of a coffle.

Before arriving at a coastal castle, most coffles would pass through the castle town: “Their streets and roads . . . are only poorly kept clean so there is a right good smell,” wrote Johannes Rask in the early eighteenth century. But he then asserted that their courtyards were “usually very neat and clean” and that trees planted along the streets “provide a comfortable shade and a lovely appearance so that every negeri looks like a small green grove.”47 Occupants of the town included the majority of the castle slaves, or those slaves owned by the castle—not intended for resale— purchased to undertake work in the castle. One report from 1749 lists 379 castle slaves for Cape Coast Castle, all identified by name and occupa- tion, which ranged from washerwoman to brick mason.48 But living in the town was also an as- sortment of both African and European traders, many of whom had houses and offices in town; they frequently paid a regular fee to house in the secured slave holes of the castle those slaves they intended to sell to ship captains. Centuries-old multistory masonry buildings are likely the rem- nants of such traders’ houses (Figure 7). Often predating the erection of the castle, the town was a socially and economically complex entity that was thoroughly integrated into the workings of the castle.49

At some point the coffle came in sight of the castle itself. As described by one European ob- server, the castles appeared to be “chalk moun- tains, especially when the sun shines directly on them.”50 The first sight of the castle was often a terrifying prospect, for at that moment enslaved Africans came to understand their fate. As one firsthand European observer wrote, “To be ex- ported is, for them, synonymous with being murdered.”51 He continued, “Sometimes a gag is placed in the slave’s mouth, which is forced wide open, in order to keep him from shouting.  .  .  . This intractability among the slaves as they ap- proach the coast also has its origin partly from their conviction that they are going to be shipped out to places where they will be eaten.”52 Oth- ers, like young Cugoano, had been tricked into approaching the castle: “Those [slaves] who are

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 97

brought [to the castles] without knowing the fate that awaits them there . . . are violently fallen upon in the forts and dragged into the slave holes. . . . They, especially the women, throw themselves to the ground, shriek, stretch their hands over their heads and turn their eyes to heaven.”53

The Royal African Company The early history of the British castles along the west coast of Africa falls into two major eras: their formation under the Royal African Com- pany and their restructuring and expansion under the Company of Merchants Trading to Af- rica, which took over control of the fortifications in May 1751. Under the direction of the Royal Af- rican Company, the amount of space dedicated to the containment of enslaved Africans at Dixcove and most other castles was never very large. In part, this is because, at the point of their con- struction in the later seventeenth century, the trade in slaves was only one growing dimension of the broader African trade undertaken from these castles.

While the English had been involved in trade along the west coast of Africa since a 1618 expe- dition up the Gambia River, the establishment of a permanent well-organized presence on the coast did not materialize until the construction of the castle at Cormantin in the 1630s.54 But in these early years the trade was driven almost en- tirely by the search for gold; the trade in slaves was dominated by the Portuguese, who enjoyed an exclusive position of selling slaves to Spanish plantations in the Americas. The revolt of the Portuguese against the Spanish in 1640 opened the door, however, to alternative European sup- pliers of enslaved labor. Paired with the rapid ex- pansion of sugar production in the newly settled English colonies in the eastern Caribbean, the middle decades of the seventeenth century set the stage for English involvement in the African trade in slaves. While English slaving had taken place for decades under various structures, this country’s involvement in the African slave trade took on an organized structure in 1672 with the foundation of the Royal African Company, which received the English monopoly on the African trade for a stretch of five thousand miles of the

African coast.55 By that time human trafficking was assumed to be an integral part of the African trade.

The Royal African Company took over man- agement of the various settlements and fortified castles already present on the West African coast. The oldest settlement, the castle at Cormantin, had been begun by the English in the 1630s but was lost to the Dutch in 1665. The most impor- tant settlement was Cape Coast Castle, which was begun by the Swedes in 1652 but taken by the English in 1664. The castle at Cape Coast stood “on a round head jutting out into the sea . . . and its being encompassed on that side  .  .  . by several rocks and the sea itself, render it inacces- sible on that side. The only landing is just under the fortress.”56 The Royal African Company dra- matically expanded the Swedish castle at Cape Coast from 1674 to 1682.57 After that expansion the castle included a number of large warehouse spaces, apartments for the governor, barracks for soldiers, and a chapel, in addition to the various towers and bastions defining its perimeter. But much of its power derived from its appearance, not from its actual constructed fabric; as with most early European construction in Africa, the walls were largely rubble contained in two ve- neers of whitewashed stone. Since castles were often poorly maintained, castle governors regu- larly complained that the heavy rains would liter- ally wash the building away.58

Given its remarkably advantageous site, the

Figure 7. European slaver’s house, Elmina, Ghana. Photograph by Louis P. Nelson, March 2011.

98 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

castle at Cape Coast was from the 1670s onward the center of English trading interests along the West African coast. Through the seventeenth century, the castle housed a garrison never smaller than fifty men and often over one hun- dred. The castle also boasted an officer corps of ten to fifteen and as many as thirty castle slaves.59 The Royal African Company lost its monopoly on the Africa trade at the end of the seventeenth century, but it continued to oversee the mainte- nance and construction of castles along the West African coast, including the one at Cape Coast, until the middle of the eighteenth century (see Figure 3).60

From Cape Coast the company established along the coast a number of factories—small, intentionally experimental, and often imperma- nent settlements where traders could engage in business. These were unfortified outposts of a few men intended to establish a successful trade treaty with a local chief.61 If they succeeded, the location might become a castle; if they failed, the company had lost little capital investment. The letterbooks of the Royal African Company offer brief glimpses into the material realities of these factories. In 1681 James Nightingale wrote from a remote location to the castle at Cape Coast. He notified the castle authorities that he had made arrangements with the local chief to build the factory. “They demand ten Anchors of brandy for ye whole charge of building ½ in hand ye other halfe when it is built,” a small charge for the con- struction of a building, he thought. He wanted immediate approval to move forward “to have it finished before Ye Raines.”62 The factory was to be built out of posts, mud, and thatch, the stan- dard West African building materials, excepting “doors, doorcases and windows,” which were to be imported to the site.63

Letters from other factories make it clear that the most important building was the warehouse. Ralph Hassell requested “a new good large lock for the warehouse door.” But he also requested “some irons for slaves for we have none here: I bought three men and three women yesterday: I am afraid of their running away for want of irons to put them in.” Those enslaved Africans, it seems, were stored in the warehouse, together

with other trade goods.64 These impermanent factories were a reflection of the reality that trade in any one location fluctuated enormously and that the management of some established forti- fied sites and some portable sites simply made more economic sense than the construction of many immovable, capital-intensive castles in locations that might prove unprofitable.65 Espe- cially in the seventeenth century, the supply of enslaved Africans to the trade was closely linked to African wartime activity, and if a region was relatively peaceful, the supply of prisoners of war to an established castle could dwindle quickly. As a result, smaller factories could be moved up and down the coast as the market required, channel- ing victims to Cape Coast by canoe or sloop.

Even so, castles were essential. Physically dom- inating the landscape, castles created the sense of a secure site for the high-stakes buying and trading in slaves and other goods. A substantial percentage of indoor space in most remote castles was given over to warehouses for storage of the various European and East Indian goods used in the African trade.66 One account by Dane Hans Christian Monrad, penned between 1805 and 1809, reported that the castle warehouses stored those things considered most valuable by African traders, especially “guns  .  .  . gunpowder, shot, flint stones, iron, lead, swords, knives of various kinds, all manner of cotton, calico, salem puris [a cotton cloth from India], silk cloth, wollen caps, quantities of beads, mirrors, and  .  .  . tobacco, rum, brandy and cowries.”67 While the particu- lars vary from account to account, it is quite clear that the most valuable trade item across Africa was guns.68 It was critical that castle warehouses remain filled with ready trade items to quickly satisfy the demands of African traders.69 But some early factories and castles also offered a dedicated containment space for small lots of captives during the negotiations between castle governors and African traders. Through the sev- enteenth and into the early eighteenth century, this was most commonly accomplished through the construction of a purpose-built “slave house,” sometimes called a “slave booth,” often erected outside the walls of the castle or factory.70

Because the grim setting offered safety and

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 99

easy access to provisions, purchasing slaves from the castle meant prices higher than those when purchasing slaves from coastal impermanent factories.71 If they were up to the challenge, sla- ver ship captains could negotiate and purchase Africans directly from independent African or European traders along the coast, who often sig- naled passing ships that they had Africans for sale.72 But such purchases from the coast were more dangerous in wartime; mooring offshore and out of site of a friendly castle for long periods of time left a slaver ship much more vulnerable to attack.73

Some of those factories at geographically or politically advantageous positions eventually became fortified outposts; the castles at Kom- menda, Anomabo, Secondi, and Dixcove were all built later in the seventeenth century to replace earlier factories. Early plans from Dixcove, begun in the 1690s, offer an example of the castles built by the Royal African Company. The shel- tered cove at a place called “Dick’s” had been a favored landing place for canoes and small boats for centuries.74 The English had established a fac- tory there by 1684, but in 1692 they began the construction of a castle, which took four years to complete. In 1711 it was described as a large square castle of stone and lime with two round flankers and two square bastions.75 A 1727 sur- vey of Dixcove completed by William Smith for the Royal African Company documents the early rectilinear core of the castle, with three diamond- shaped bastions and one round tower, suggest- ing that one of the round towers had by that point been replaced (Figure 8). The open courtyard in the interior was ringed by buildings providing some warehouse storage, apartments for officers, and barracks for soldiers.

Careful examination of early plans and the surviving fabric of Dixcove and many other castles makes it clear that the builders of Dix- cove never dedicated a large space for the hous- ing of enslaved Africans, a view contrary to the standard interpretation of these coastal forts. In fact, Dixcove’s placement insured its importance not so much as a slave castle—although it was a slave castle—but as a guard over a supply of essential natural resources. A heavily forested

site provided lumber; a natural bed of lime- stone sourced masonry building materials; and a natural spring made it an excellent location for ship refreshment and repair.76 As a result, its role in the slave trade was largely as a point of temporary holding and transfer.77 Of course, captives passed through this and other fortifica- tions and factories along the coast, but they only passed through. Records of the Royal African Company from 1678 suggest that between April and December of that year, Cape Coast housed 1,854 Africans for sale, quite obviously not all at the same time. Three hundred sixty-six arrived from Anamabu; 330, from Egya; 166, from the castle at Accra; and the same number, from Win- nebah. The rest are unaccounted for but presum- ably came from small factories along the coast or from traders delivering directly to Cape Coast.78

Unlike the smaller fortifications, Cape Coast had underground containment chambers with substantial capacity.79 These chambers were cre- ated by the English expansion of the castle in the 1680s; they utilized the remains of the quarry from which the stone for the massive exterior walls and new bastions had been extracted. The massive hole, it seems, was vaulted and paved above to create a central parade ground.80 One ac- count from 1682 describes the vast chamber as a “mansion . . . cut out of the rocky ground, arched and divided into several rooms; so that it will con- veniently contain a thousand Blacks, let down at an opening for the purpose.” The advantage of keeping the enslaved underground was “good security to the garrison against any insurrec- tion.”81 Inspection of the surviving fabric of the castle suggests that some portion of the series of chambers under the courtyard (spaces now used as cisterns) were first used for containment. These chambers had no natural light except that leaking in through the single openings near the roof vaults, through which slaves were raised and lowered (see Figure 9). A later 1756 plan of the castle shows a number of ventilators in the floor of the courtyard, extending all the way across the parade grounds, presumably inserted to deliver more light and fresh air into the spaces below. The ventilators suggest that these spaces were vast indeed, spanning across the width of the

100 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

parade ground. While period descriptions report that these cells were intended to hold up to one thousand captives, records reveal that they actu- ally held far fewer. One account from the turn of the eighteenth century records four hundred Africans being held in the containment cham- bers of Cape Coast.82 With or without the venti- lators, the conditions must have been unthink- able. Even above, in the best apartments for the captain and his guard, one factor wrote in 1708, “there is never a dry room to lye in.”83

The Company of Merchants Founded in 1750, the Company of Merchants Trading to Africa took over management of the forts and instituted a major restructuring of the West African slave trade. This change was driven largely by the press to undermine the monopoly

on the slave trade held by the Royal African Company at its founding.84 As early as the late seventeenth century, merchants—many from Bristol—began to launch voyages independent of the Royal African Company, a process that slowly weakened the company’s monopoly and caused its slow demise.85 The newly founded Company of Merchants was open to anyone who wanted to enter the slave trade. Parliament agreed to pay a yearly grant to the company for the maintenance of the British castles along the coast of West Africa. The London-based African Committee, consisting of three men, one each from London, Liverpool, and Bristol, received the monies, gath- ered the necessary materials, men, and other resources, and sent them to Cape Coast to then be distributed to the various castles as necessary.

Through the 1750s and 1760s, the company

Figure 8. Plan of Dixcove Fort as it appeared in 1727. Drawing by Jason Truesdale based on William Smith, “Survey of Dixcove Fort, 1727,” appearing in Arnold Walker Lawrence, Fortif ied Trade-Posts: The English in West Africa, 1645–1822 (London: Cape Publishers, 1969), 188.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 101

spent a great deal of time and resources assess- ing and improving the physical condition of the existing fortifications and building new forts. In 1756 the company informed the governor of Cape Coast Castle that “the ruinous condition of the Forts in general, is a matter of great concern to us.”86 That same year, the company sent John Appleby, an “engineer,” to Africa to oversee the reconstruction of the castles, with the instruc- tion that he was to be in charge of all designs, “ex- cept you shall find it necessary from the opinion of the governor and council at Cape Coast Castle to make any Alterations in the Lodgements for Slaves or any of the inner Apartment.”87 The com- pany wished the fortifications to be made sub- stantial, “making good use of lime, and not clay as formerly,” objecting to the mud-walled con- struction of so much of the castle construction under the Royal African Company.88 A few years later, in 1758, the company required reports from the chiefs of each castle assessing the material conditions of the outposts along the coast. Let- ters sent in the 1760s suggest that those chiefs had not repaired or upgraded the conditions of the castles to the expectations of the company.89

Under the Royal African Company the forti- fications had been understood as trading posts where Europeans could engage in business trans- actions with African traders. One of the increas- ingly lucrative markets was the trade in slaves. In the seventeenth and early eighteenth centu- ries, many castles housed slaves not in dedicated rooms in the castle but in stockades called “slave yards” outside the castle walls.90 Under the Com- pany of Merchants, the fortifications became installations dedicated almost exclusively to the trade in slaves. For the first time castle plans care- fully illustrated slave containment chambers, called “slave holes” on most mid-eighteenth- century plans. Most commonly, these slave holes were built within the thick walls of a newly erected bastion, as in the case of Dixcove’s slave hole. The use of the term hole might draw from the square ventilators that often opened through the pavement of the bastion down into the cham- ber (Figure 9). It might also draw from the an- cient use of the term hole as a verb meaning “to oppress.”91 The appearance on the 1750s plan of

Cape Coast Castle of a “Black Hole,” a deep, nar- row chamber clearly used for punishment, sug- gests that hole certainly retained its association with explicit oppression in the middle decades of the century. Plans also demarcate slave yards— transformed into a place for daily exercise—and identify ventilators for the holes within bastion walls. Records from the second half of the cen- tury demonstrate that the sick from among the enslaved were sent out of the castle into town to recover, so as not to contaminate the rest of the African captives. One 1753 letter from the cas- tle governor to a ship’s captain reports that the eighteen Africans being held in the castle were all in good health but that three had been sent to town. A 1770s listing of castle rules reports that no sick slave was to remain in the castle.92 Such attention to conditions was quite obviously driven not by general concerns for welfare but the economic incentive to preserve the value of an investment in human capital. Remote castles were refitted with secure slave holes for more ef- fective temporary containment; new castles were built at points central to the slave trade; and a new set of holding chambers was added at Cape Coast, which remained the flagship fort. Under the Company of Merchants, the castles became slaving machines.

Changes to the existing fortifications are well represented by the new work at Dixcove Fort. The expansion of the new spur had, in fact, begun prior to the economic restructuring and so cannot be claimed as improvements by the Company of Merchants. A plan of the castle produced in 1750 represents the new-built spur and barracks— constructed of locally fired bricks—but notes three old bastions out of repair.93 Significantly, a 1750 plan drawn under the directorship of the Royal African Society does not differentiate any one of the three ruined bastions as any differ- ent from the others. This is significant because a 1756 plan of the same fort—now produced under the aegis of the Company of Merchants— reveals that the small slave hole, the same space that survives to the present, had been installed in the northeast bastion—and was called the “slave room” on the plan (Figure 10). Bastions, of course, lent themselves well to such a function.

102 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

They were segregated from the rest of the spaces of the castle interior. They were usually very sol- idly built. And they usually had only a single en- trance that was easily surveilled from many loca- tions within the fort. H. C. Monrad, writing from personal experience, reported that in such spaces slaves “spend the night on wooden planks, one over the other, [suggesting wall-mounted shelves for sleeping] in dark vaults where only a little air can seep in through a square hole that has been made in the door.” He continued by remarking his surprise that the enslaved didn’t suffocate, “since the heat, augmented by the evaporation from many people, caused by that of which one must at times, rid one’s self, and which are col- lected in large tubs, produce an extremely abhor- rent, mephitic stench.” The odor, he continued, was at times so strong that “in the morning when the rattling doors are opened, (especially if there are many slaves), [it] spreads over the entire fort and corrupts the air.”94 By the second half of the eighteenth century, all English castles were out- fitted with similar slave holes.

Soon after taking control in 1750, the Com- pany of Merchants began a campaign to construct a new castle at Anomabo, a lucrative factory since the sixteenth century.95 Surviving records sug- gest that Anomabo had become a favored slaving locale for many English slave captains.96 The pri- mary motivation was to secure the site for Eng- lish trade and to provide defense against a feared French invasion.97 The agreement with the local chief allowed the English to build “a fort or ware- house of whatever kind they choose,” as well as to cultivate ten acres for a garden.98 By 1753, Par- liament, which was charged with paying for the castle at Anomabo, finally consented, and the foundations of the castle were laid in August of that year. Built right on the edge of a beachside shelf of rock, the building is a large square hulk with massive diamond-shaped bastions at all four corners (Figure 11). Intent on building a cas- tle that did not require constant repair, all of the building materials were shipped from England. More than two million English bricks traveled to Africa as ballast, approximately twenty thousand

Figure 9. Courtyard at Cape Coast Castle, Cape Coast, Ghana. The square opening to the left of the stairs is an entrance into the early containment chambers. The ventilators—two small openings, framed in stone and cut into the pavement—can be seen to the far right on the edge of the courtyard. This castle is a World Heritage Site and as such has been restored, unlike most others along the coast. Photograph by Louis P. Nelson, August 2012.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 103

sent in each of some one hundred ships. Since the ships were unable to dock, all of these materi- als had to be brought ashore by canoe. As in the retrofitted forts, the company built a dedicated slave hole into one of the bastions, although here the chamber was much larger than allowed by smaller preexisting bastions (Figure 12). The cas- tle was operational but still not complete by 1760. The efficiency of castles like Anomabo is made evident by the fact that while Cape Coast contin- ued to be populated by an ex-patriot community of up to one hundred, these smaller castles were often manned by as few as six.99

A final act of reconstruction undertaken by the Company of Merchants was the extensive expansion of the castle at Cape Coast, especially the construction of a new set of slave prisons; no

longer warehouses, these spaces were purpose- built, more closely aligned with the prisons of England and Europe than with the warehouses and quarry spaces of the earlier castles. Designs for the new set of prisons in a defensive spur were prepared in 1768, and construction was com- pleted soon thereafter. Five vaulted chambers were built into the large new double bastion that overlooked the sea (Figure 13).100 As evident in the section through the bastion, the five chambers were stepped to accommodate a central brick gutter that encouraged human waste to flow down and out through the exterior wall of the lowest chambers. Presumably, the gutter would also drain washing water, although there is no evidence that the spaces were regularly cleaned. A parade ground door opened into a tunnel that

Figure 10. Plan of Dixcove Fort as it appeared in 1756. Notice the “slave room,” labeled 1, in the upper bastion. Drawing by Jason Truesdale based on original plan published in Arnold Walker Lawrence, Fortif ied Trade-Posts: The English in West Africa, 1645–1822 (London: Cape Publishers, 1969), 198.

104 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

led down into the chambers, connecting with the second chamber in the sequence of five (Fig- ure 14). A 1972 archaeological excavation of the prison chambers unearthed food debris, body or- naments, and other evidence that demonstrated that all five chambers were used for human con- tainment.101 The first chamber was structurally isolated from the rest by a small opening that originally had a lockable door. Unlike all the others, that chamber had a brick-laid floor and a much more sophisticated drainage system for more effective cleaning out of waste (Figure 15).102 The last of the chambers has evidence of a lockable iron gate and evidence for shelving; the three middle chambers were all open, one to the next. Through the sequence of the final four, the packed earthen floor sloped from either side to- ward a single gutter that ran down the spine of the rooms.

These five chambers were designed to intro- duce more daylight and air to the Africans con- tained therein than were the earlier quarry cells. The last three chambers had large openings near the crest of the roof vault—presumably with iron bars limiting human passage—that would have once allowed for ventilation and light from above. These openings are now blocked up, creating an even darker experience of the interior. The first

chamber has a curious shaft leading from the entrance passage into the chamber, presumably intended to filter some natural light from the en- trance tunnel into this space. In all five, small ventilators on the walls opposite the windows were intended to allow some airflow through the chambers. Much larger than the previous cham- bers, far better lighted and with far greater ca- pacity for airflow, more easily accessed and more easily cleaned than the former site, the new pris- ons were—according to contemporary descrip- tions—capable of holding upwards of two thou- sand people, doubling the capacity of the earlier chambers.103

Yet the claimed capacity of the prison was far higher than the reality of persons held in these spaces at any moment in time. The launch of the massive NEH-funded Voyages: The Trans-Atlantic Slave Trade Database by David Eltis and Martin Halbert has made available an extraordinary body of numeric information about the slave trade. A recent query of the database uncovered data about every slaver voyage that departed from Cape Coast Castle from 1685 to 1807.104 Data from the Voyages records make clear that the re- construction of the prisons by the Company of Merchants was in clear response to the increased demand on the castle. Under the Royal African

Figure 11. The Castle at Anomabo, Ghana. Photograph by Louis P. Nelson, July 2012.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 105

Company, from 1685 to 1749, Cape Coast Castle provided an average of 517 enslaved Africans per year. Those shipments increased rapidly under the Company of Merchants, who supplied from Cape Coast an average of 842 slaves per year in the 1750s and then a sharp increase to 1,514 slaves per year in the 1760s, an almost threefold increase over the average annual rate under the Royal African Company. Clearly in the face of that sharp increase in the 1760s, the Company of Merchants felt the need to outfit the castle by building enlarged prisons in the late 1760s in response to the heavier demand and to better in- sure the health (i.e., value) of their investments. And they were right. The average number of en- slaved Africans departing from Cape Coast each year jumped to 1,847 in the 1780s and peaked in the 1790s at 1,998 per year, a fourfold increase over the rates of exportation under the Royal Af- rican Company. It is important to note, however, that these rates are averages over the decade, and

as such they mask the remarkable variation by year; trade in one individual year could be as much as double the average rate.

Even so, thousands of people were never held in these prisons at any one point in time. The castle’s busiest year for export was 1785, in which Cape Coast provided an astonishing 4,118 Afri- cans to twelve slaver ships—nine of which were scheduled for Jamaica. But these 4,118 individu- als were loaded onto ships that left in seasonal clusters, with 790, 580, and 1,095 in March, April, and May, respectively, and 650, 400, and 470 in August, September, and October, respec- tively. May 1785 was the heaviest month of supply, possibly the heaviest month in the fort’s entire 250 years of slave trading, and only in that month did the total number of enslaved crest 1,000. But that does not mean that on any given day there were actually 1,000 Africans in captivity in the prison. These 1,095 departed in two waves over the course of the month: 320 departed on a ship

Figure 12. Interior of the slave hole, Anomabo Castle, Ghana. Photograph by Louis P. Nelson, July 2012.

106 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

on May 2, and the other 775 were on two ships that both departed on May 9. And enslaved Af- ricans arrived in the castle in small lots of ten and twenty. Ship captains were usually eager to purchase slaves as quickly as possible, especially if there were competitors. Ship logs make clear that captains moved recently purchased slaves onto their ships soon after purchase to avoid hav- ing to pay the castle to house and feed slaves now owned by the captain. As a result, in the heavi- est month of supply, the prison of Cape Coast Castle contained a few hundred individuals at any one point in time—more than could fit in a single chamber, but far fewer than would fill all five—never coming close to the proclaimed 2,000-person capacity.

Even so, these roomier chambers, with better

light and air, were still miserable. Accounts sug- gest that upon arrival at the fort Africans were “struck into chains,” meaning bound by chains into gangs of ten to twenty. Individuals so struck remained in their gang until sold to a ship.105 Cugoano’s account of his own capture in 1770 is the sole known account of the experience of con- tainment in Cape Coast’s prison. He was kept in the hole “for three days, where I heard the groans and cries of many, and saw some of my fellow captives.” The separation from and then reunion with his friends betrays the reality of multiple African traders and multiple routes to this single prison space. Describing a still, dark, and likely very damp cavernous space, his account reports more on what he heard than what he saw: “When a vessel arrived to conduct us away to the ship, it was a most horrible scene; there was nothing to be heard but the rattling of chains, smacking of whips, and the groans and cries of our fel- low men. Some would not stir from the ground, when they were lashed and beat in the most hor- rible manner.”106 The sharp sounds of whips and the low groans or loud cries certainly echoed through the cavernous chambers of Cape Coast’s slave prison.

Contrary to Cugoano’s account, which reports that his coterie of slaves remained in the prisons for the three days of his containment, other re- cords suggest that by the second half of the eigh- teenth century enslaved Africans were spending the majority of their days not in the chambers but in the courtyard or parade ground of Cape Coast and other castles.107 Like the drive to build

Figure 13. Still image from a three-dimensional model of the 1768 dungeons at Cape Coast Castle, Cape Coast, Ghana. See the online version of this article at JSTOR (www.jstor.org) for the three-dimensional model from which this is taken. Model by Jason Truesdale, 2013.

Figure 14. Entrance tunnel leading down into the 1768 dungeons of Cape Coast Castle. Photograph by Louis P. Nelson, July 2012.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 107

healthier slave prisons, this decision was likely motivated by the drive to preserve the health of the enslaved as a means of sustaining their re- sale value. A 1756 plan of Cape Coast demarcates the newly established yard for enslaved women, just above the large bastion called Greenhill’s Point. Records show that the large parade ground dominating the rest of the castle’s open interior was used as a yard for male slaves, although they remained bound in lots of ten or twenty. These yards created an outdoor space for slaves to take in some fresh air and exercise. Accounts also suggest that the enslaved were driven down to the ocean once or twice daily for washing.108

The sights and sounds of everyday life occa- sionally interrupted long days in the courtyard. The records of Cape Coast Castle report an as- sortment of tropical animals kept in the castle as pets, from chimpanzees to parrots and, once, even a leopard called Sai.109 This menagerie only added variety to the sounds of violence that

marked the soundscape. Vultures, circling to col- lect scraps from butchering, frequented the cas- tle, as did voracious biting red ants.110 The regular ringing of the castle bell marked the hours.111 In the evenings, after the return of the enslaved to the subterranean chambers, the rattle of chains was a constant sound that echoed enough to be heard consistently throughout the apartments above.112 The hourly bells were occasionally bro- ken by the fife-and-drum corps, who played for newly arrived visitors, at special events in the governor’s hall, and at the many funerals for sail- ors and officers alike.113 Screams also punctuated the day—from newly arrived coffles, from castle slaves, regularly flogged, and from the slaves of merchants in town, frequently sent to the castle for whippings.114

The announcement of the arrival of a ship’s captain by the fife-and-drum corps meant the likelihood of a sale. In preparation the castle slaves would wash and oil the bodies of Africans

Figure 15. Interior of chamber 4, Cape Coast Castle. The end wall window to the courtyard has been closed in. The sewage gutter runs left to right from chamber to chamber down the spine of the prisons. Photograph by Louis P. Nelson, July 2012.

108 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

to be sold. This event took place in the courtyard, with slaves “either chained or loose, so that they form a circle.” The buyers then undertook a thor- ough examination of the bodies of those offered for sale: “The slave must open his mouth wide, show his teeth; they smell in his mouth, and look very carefully into his eyes; he must perform all manner of movements with his arms and legs; the secret parts are examined, during which I have seen many, especially young, Negresses crying.” Not all submitted to inspections. A 1767 account from the Dutch castle of Elmina reports that when the slaves were ordered from the pris- ons into the yards, they “cut their own throats: one Negroe even cut the throat of his wife and then his own; the yard of the noble company’s chief castle was thus turned into a bloodbath.”115

Barring such violent resistance, slaves se- lected for sale were physically marked as human property. After the inspection the buyer initially “marks those slaves he has chosen with a piece

of white chalk right across the head.”116 After the sale victims were frequently branded “as the seller often sees a chance to exchange the sold slave for one of inferior quality.”117 Those so se- lected and then branded were then sent out of the castle to canoes waiting on the shores.

The process of moving those recently sold to the canoes was entirely restructured by a second major construction in the later eighteenth cen- tury, a new set of warehouses completed in 1777 (Figure 16).118 Letters from the governor to the Company of Merchants in London make it clear that the company prepared designs for the new construction but that the governor rejected them, preferring to build them to his own design. “I am now erecting a fortification in the form of a horn work,” he wrote to the company, describing a type of fortification with two adjacent bastions extend- ing from the body of a castle. He hoped that they would “not blame me for deviating from Mr. Hip- perley’s design, and be enabled at the same time

Figure 16. Plan of Cape Coast Castle. The hatching indicates the extent of late eighteenth- century additions. 1 indicates the air vents into the seventeenth- century dungeon; 2 points to the 1768 dungeons; 3 indicates the 1777 hornwork and warehouses. Drawn by Jason Truesdale based on 1944 drawings produced by the University of Ghana and fieldwork undertaken by the author in July 2012.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 109

to see the impropriety of Mr. Oxbridge’s plan.” He was quite pleased with the new construction, given that the “old warehouses are likely to fall soon.” He continued that the new warehouses within the hornwork bastions could “contain all that used to be put in the old warehouse,” includ- ing “all the liquors, provisions, and store that will ever be in the castle at one time.”119 These new warehouses, in the form of a hornwork, were in fact also a new channel by which the recently sold could be safely marched out of  the castle and toward the canoes waiting outside the castle walls. The new construction included a taper- ing (and thereby increasingly constrained) pas- sage that gently sloped down from the height of the parade ground to an exterior platform only a few steps above the beach. In addition to its gradual constriction—limiting the damage pos- sible from the press of panicked Africans—there was a sentry box immediately above with a clear view over the passageway (Figure 17). From this position a guard had easy surveillance over the stream of enslaved and could shoot if necessary. The investment of so much attention to a pas- sage out to the sea clearly suggests that these few minutes were fraught with anxiety for both cap- tor and captive.

The prisons added to Cape Coast Castle in the late eighteenth century were part of the broader discourse on contemporary European institu- tions of containment and reform so eloquently discussed by Michel Foucault. In Discipline and Punish: The Birth of the Prison, Foucault contrasts these institutions as distinct from sites of slav- ery.120 I disagree. The commodification process of slave making diminished acts of punishment in an attempt to preserve the value of the slave— flogging in the castle was usually reserved for laboring slaves not intended for resale—but many dimensions of the discipline that shaped prisons, military installations, and schools were readily apparent in these slave forts. Emergent mechanisms of time discipline and the mak- ing of docile bodies were central to the work- ings of both African prisons and European in- stitutions of reform. Surveillance and organized/ specialized containment, if not simultaneously universal and individual as in the Panopticon,

were both in evidence in the castle in nascent form. Furthermore, the processes of inspection by potential purchasers certainly engaged in the classification and objectification processes so clearly articulated by Foucault.121 Those spaces called “slave holes” in the middle decades of the eighteenth century were by the 1790s commonly called slave “prisons,” mapping directly onto the institutional use of this term in Britain in these same years.122

As imagined by their builders, late eighteenth- century British slave castles in West Africa were

Figure 17. View from the sentry box over the passage to the ocean gate, Cape Coast Castle. Photograph by Louis P. Nelson, July 2012.

110 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

prisons, commercial machines dedicated to the production of slaves, just as their counterparts in Britain were dedicated to the social (re)making of the inmate. One space within the late eighteenth- century castle, a single cell labeled the “Black Hole” on early plans, more than any other dem- onstrates the discipline uniting the two. Its stone floor is still today inscribed with deep circular scars cut by shackles, anger, and long days of iso- lation (Figure 18).123 Whether for soldiers or slaves, or both, this space manifests most acutely the emerging mechanism for correction: isolation.124 Through time discipline, compartmentali zation, isolation, and the systematic production of doc- ile bodies, African castles were not distinct from prisons and schools in Europe. In fact, African prisons were loci of experimentation, at the very least contingent with the development of English and European institutions of reform and correc- tion. In this way, the castles of Africa’s coast stand at the front edge of modernity.

Ship Trade The slave castles along the coast loom large in the historical imagination of the transatlantic slave trade. Yet the vast majority of Africans traded by English and American slavers never saw a fort. Modern histories of the transatlantic African slave trade break the west coast of Africa into six regions running from northwest to southeast: Senegambia; Sierra Leone, also called the Wind- ward Coast; the Gold Coast; the Bight of Benin; the Bight of Biafra; and West-Central Africa. In the broadest terms, the European trade slowly moved down the coast from Senegambia through the Gold Coast in the seventeenth century to West-Central Africa by the early nineteenth cen- tury. The English and American slave trade fol- lowed suit, strongly centered in the Windward and Gold Coasts in the seventeenth and early eighteenth centuries and in the Bight of Benin and West-Central Africa from the middle of the eighteenth century into the nineteenth century. Although the English built substantial slaving castles all along the Gold Coast, only one in five Africans sold into slavery—approximately 22 percent—passed through the castles of the Gold Coast, most now in modern Ghana. The records

of the Royal African Company make clear that even in the earliest decades a substantial percent- age of Africans taken into the slave trade were driven by African traders along the coast and sold to English slavers, bypassing the European castles altogether.125 The greatest concentration, approximately 860,000, or 29 percent, came from the Bight of Biafra, most of those during the second half of the eighteenth century. At the heart of the Bight of Biafra was Old Calabar, the greatest supplier of enslaved labor to the British Caribbean and the primary source of Africans for Jamaica in the 1790s, its decade of greatest consumption.126 The remaining came from the coastal factories of African and European trad- ers scattered all along the west coast. This dis- tribution meant that the majority of purchases by slaver captains were not from “fort-trade”— conducted in the massive slaving castles along the Gold Coast—but from “ship-trade.” This is where slaver captains negotiated with traders right from the decks of their own ships, leading one sailor to describe his ship as “a floating fac- tory.” Slaver ships were simply an extension of the coastal castles, spaces framed around con- tainment, correction, and discipline.127

Nicholas Owen was an Irish sailor working in Sierra Leone in 1754 as a small-scale slave trader, first in the employ of an Englishman and even- tually on his own. His surviving journal cap- tures a wealth of information about his day-to- day life. Near the end of his record, he describes his house, similar to the houses of many other smaller European slave traders in the region: “As our houses are not built for durance or strength, we can soon erect one on long sticks muded over and whitewash’d, having the inside lined with mats and well thatched aloft.” He appended to this description his rough sketch of the house, which shows an oblong house with rounded ends supported by earthfast posts and sheltered by a thatch roof. Behind the building extends an en- closed pen, a space surely designed to keep ani- mals out but also able to keep people in (Figure 19). Built almost entirely of found natural ma- terials, Owen’s house differed little from those of the Africans with whom he frequently social- ized and traded, which were “built of mud and

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 111

Figure 18. Floor of the “Black Hole” isolation chamber, Cape Coast Castle. The curved gouges in the stone floor are the evidence of hours of isolation as captives scraped their shackles back and forth around a central fixed point. The bouquet is a tribute left behind by one of the thousands of visitors to these spaces every year. Photograph by Louis P. Nelson, March 2011.

112 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

stakes drove into the ground and covered with long grass.” Like most traders, his settlement comprised his own house and that of his African laborers, both free and enslaved: “About twenty yards from my house stands that of my people, likewise my cookroom.” Other entries also men- tion a storeroom, which was surely well secured, as it housed all the necessaries for his trade, the bulk of which was iron and lead bars.128

Owen never mentioned a dedicated space for the containment of his slaves, but it is likely that one existed. Late in his journal, he described being both sickly and low staffed, and in these instances he mentioned three or four enslaved Africans “I have now by me in the house,” sug- gesting that as something worthy of note because it was not his usual practice.129 The presumption is that his purchased Africans—never more than about five in number—were usually housed in the rear building occupied by those Africans who worked for him or in a small-scaled barracoon, a temporary pen used to house captives while trad- ers waited to negotiate with passing ships.130

Sometimes, African communities on the coast became slaving centers. Robin Law’s excellent study of slaving in Benin suggests that the coastal town of Ouidah boasted a number of secure com-

munal buildings, called “trunks,” where Africans owned by many local merchants could be stored.131 One Dutch trader reported in 1705 that all traders bringing slaves to the coastal market had to pay a local chief for the use of the trunk, or barracoon, “with which they guarantee that the slaves will not escape.”132 Whether individual or communal, these barracoons differed only in scale. A Brit- ish naval captain described them in the 1840s as “sheds made of heavy piles, driven deep into the earth, and lashed together with bamboos, thatched with palm leaves.” The walls usually rose four to six feet, with an open space of four feet be- tween the wall and the roof “for the circulation of air.” A central range of piles that carried the ridge of the thatched roof supported larger barracoons. Along each line of piles hung a chain with “neck links” for securing slaves at two-foot intervals, and the floor was often boarded to protect slaves from skin-boring insects.133 An image of one such barracoon was published in the Illustrated London News in the late 1840s (Figure 20). Constructed of vertical posts close enough to prevent human passage but wide enough to allow air circulation, this barracoon appears to hold between twenty and thirty people. Barracoons could be found all along the African coast but were most commonly concentrated in Senegambia and the Windward Coast.

Unlike their counterparts elsewhere along the African coast, chiefs in the Bight of Biafra never granted Europeans the right to build trading castles or settlements on their soil. This policy resulted in the development of African-operated trading centers along various rivers, centers that were connected to vast trading networks into the African interior. These trading centers were the seats of African merchants who engaged in regu- lar correspondence with merchants in England, especially in London through the seventeenth century, Bristol in the early eighteenth century, and Liverpool at the height of the trade in the late eighteenth century.134 The lack of a castle and its infrastructure as a land-based interme- diary space meant that much more negotiation took place on board the deck of the ship. Even so, by the end of the eighteenth century this mode of purchase came to dominate the traffick-

Figure 19. Drawing of Owen’s house. Reprinted from Nicholas Owen, Journal of a Slave-Dealer (1759; repr., London: Routledge and Sons, 1930).

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 113

ing in slaves. These ships, waiting sometimes for months in the safe haven of the Niger and Cross River Deltas, collected almost one-third of all those slaves taken from Africa in British and American slaver ships, a higher percentage than any other region.

One of the most intensive slaving centers was Old Calabar, a series of Efik trading towns in the estuaries of the Cross River Delta, now in modern Nigeria.135 The Efik had been trad- ing with English slaver ships since the first sus- tained contact in Old Calabar in the 1660s, but this economy grew rapidly over the course of the eighteenth century.136 The leading town of Old Calabar was Duke Town, a community of about two thousand residents by the early nineteenth century.137 Over the course of the eighteenth cen- tury, this region was transformed by the transat- lantic slave trade. By the later eighteenth century, Efik traders were selling between three thousand and five thousand Africans annually, 85 percent to English and American slavers. The region had become the nexus of the British slave trade.138 A large trading network in this region was respon- sible for the capture of Olaudah Equiano and his sister in 1765 (see Figure 3).139

The vast majority of slaves made the final leg of their journey to Duke Town in the bottom of

one of Calabar’s famous war canoes. Probably the most remarkable description was of Efik chief Eyamba’s war canoe:

His great canoe was gaily decked out with several ensigns streaming in the wind, British ensigns, with his name thereon in large letters. The little house amidships was brilliantly painted red and yellow. Astride the roof thereof sat two men beat- ing drums with might and main. Before it stood Eyamba, shaded by his grand umbrella, dressed as usual, except in having a gold laved cocked hat under his arm, and a splendid sword, a present from the Dutch Government, at his side. In the bows a large gun pointed forward, and before it stood a man with a bundle of reeds, which he kept shaking at arms length to warn every obstacle and danger out of the way. On each side sat fifteen men with paddles, and between them down the center stood a row of men armed with cutlasses and guns. The king’s body-guard were immediately around him. A train of inferior canoes, ornamented and arranged in the same style, belonging to the lesser gentry, were in his wake.140

As this description suggests, these great canoes were usually among the best of a much larger collection that all belonged to a canoe house,

Figure 20. “Slave Barracoon,” Illustrated London News 14 (April 14, 1849): 237.

114 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

the major organizational framework for trading among the Efik. Usually headed by a man of great charisma and talent in trade, these canoe houses functioned like trading companies connecting numerous households, often related by kinship, in different locations along a river system.141

After journeys that sometimes lasted weeks or even months, the Calabar war canoes returned to Duke Town, where their captives were then dis- tributed among the houses of a town.142 Most of the Efik and other Africans in this region lived in wattle-and-daub houses covered in thatch. These buildings were usually built in clusters around a yard; typically, an Efik man had his own house and separate houses for his wives, his freemen, and his slaves.143 Duke Town was described in 1805 as “composed of a number of low houses, supported by mangrove sticks, and covered over with bamboo, laid across and afterwards thatched with bamboo leaves.”144 A few decades later, the town was similarly described as filled with houses, “low, mud-plastered, and palm-thatched, without windows, but each with a capacious door, leading into a small court-yard. There was hardly anything like a street, and the passages between the houses were narrow, crooked, rugged.”145 An early twentieth-century photograph of Duke Town, surviving on a postcard, suggests that the townscape had changed little over the past cen- tury (Figure 21). Densely constructed of one-story mud-walled houses, Duke Town and other Old Calabar villages would have seemed generally fa- miliar to most captive Africans.

Like all towns in Calabar, the center of Duke Town was dominated by the palaver house, which functioned simultaneously as a religious and a governmental center. The palaver house in nearby Creek Town was described in the 1840s as

a large, low shed, with its end to the street, and quite open in front. Several immense posts of solid mangrove supported the ridge-pole. A broad seat of hard-beaten clay ran down the two sides; the further end was closed by a recess for Egbo mys- teries; in front was the great Egbo drum, fixed on a frame, to be beaten only on occasions of public importance; and before it were two upright pen- tagonal stones, “pillars of remembrance” of basal-

tic appearance, which had been brought originally from the Camaroon country. On both these and the drum was the blood of sacrifices. In the center of the street, before the Palaver House, stood the figure of a man, rudely carved out of a great post, which also formed its pedestal. Up the front of the base a serpent seemed crawling, and up the back an alligator.

Upon inquiry, the figure was identified as the devil.146 The palaver house in Duke Town, not visible in the postcard image, was fairly similar in form except that it stood on a rise in town, as- cended to from the road by a set of steps.147

The palaver house functioned as the site for entertaining by the local chief, both on the oc- casion of visitors and for regular feasting and celebration.148 But as the site for all executions, the palaver house was for many a place of hor- ror. One early nineteenth-century visitor to the town noted that a man condemned to death was “chained to the palaver-house, with a large timber chain round his neck.”149 Efik practice insisted on executions by beheading as judgment for those found guilty of murder, but also as ritual: the decease of an important man occasioned the be- heading of many of his slaves so that they might accompany him in death.150 The diary of Antera Duke is replete with accounts of executions of enslaved Africans on such occasions. On the death of Duke, a local leader, in 1786, the family decapitated fifty enslaved Africans in one day in the palaver house.151 As a result, the interior of the palaver house was regularly stained with the blood of sacrifices.

If the palaver house was a place of sociabil- ity but also judgment and fear, the houses of the major traders were surely objects of wonder. Be- ginning at least in the later decades of the eigh- teenth century, leading Efik traders began to import whole house frames from their English trading partners, most commonly those from Liverpool.152 A visitor to Old Calabar, one Mr. Nicholls, described a typical trader’s house dur- ing his visit in 1805:

The principal traders’ houses are built of wood, brought out by the different captains from Liver-

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 115

pool, oblong, and thatched with bamboo leaves, which last very well two years. The house I reside in was brought out by Mr. Patrick Fairweather; was built in the year 1785, and still remains very good. A description of mine will suffice for all the rest, as they are all built upon the same principle; the house is about twenty yards long and thirty feet high, with a ground floor, a first floor, and a kind of cock-loft: the first floor contains two rooms, one I occupy, and the other my attendants, and two small rooms in each wing for bedrooms. My room is about forty feet long, twenty-five feet wide, and fifteen feet high, and has been very handsomely finished. A covered gallery surrounds the house.

As mentioned by Nicholls, these houses took the common form of a raised two-story house, with a gallery enclosing the whole on all sides. Below the main living floor were the ground-floor ser- vice spaces, sometimes walled, sometimes de- fined only by pillars carrying the upper floor. Such a building in Duke Town survived into the 1880s to be represented in a view from that decade (Figure 22). By the late eighteenth cen- tury, of course, the Efik had been trading with the English for generations and had managed to accumulate a stockpile of English material goods and social practices—so much so, in fact, that Nicholls noted that the interior of his guest- house put him “in mind of a drawing room in England.” He continued:

I have two large pier-glasses, seven feet by four, elegantly gilt and ornamented; twenty five ditto, from two and a half to four feet; three large sophas, twelve chairs, two handsome escrutoire desks, six tables, two large garde vines, one handsome marble side-board, and an immense quantity of glasses, china, and earthen ware; six paintings, and twenty large engravings, five clocks, and two musical ditto; and a pretty jumble of furniture it is.153

Duke and other Efik traders typically dressed in English clothing when entertained on a slaver ship by its captain.154

Adopting select Anglo material practices was not unique to the Efik. As early as the 1750s, Nicho- las Owen reported a mulatto African trader on

the Windward Coast who “lives after the manner of the English, having his house well furnish’d with English goods and his table tolerably well furnish’d with the country produce. He dresses gayley and commonly makes use of silver at his table, having a good side board of plate.”155 In 1773 Robin John, African merchant in the Bight of Biafra, ordered from Ambrose Lace, merchant in Liverpool, a large mirror, one table, six chairs, two armchairs, two small writing desks, twelve pewter plates, four dishes, twelve knives, twelve forks, two large table spoons, one pair of balances, one case of razors for shaving, one hundred yards of chintz, and two coats with gold lace, one red and one blue all “for my Salf.”156 For newly ar- rived Africans, these traders’ houses, their fur- nishings, and certainly their dress would have seemed very far from familiar.

The structure of the slave trade in Old Calabar eliminated the need for large barracoons. The greatest reason for this was the simple fact that slaving expeditions were usually launched once a ship had safely anchored in the river and an agree- ment had been reached between the ship’s cap- tain and the African merchant.157 As expeditions arrived, captives were stored in various houses in town for only a few days at most while they were washed and fed in preparation for sale to the captain. For these few days those captive Afri- cans were certainly in chains while they awaited purchase; English trading records provide ample evidence for the purchase by Efik traders of iron

Figure 21. Postcard, “Government Hill from Duketown, Old Calabar,” Bob to J. B. Scott, Old Calabar, February 2, 1903. Image courtesy Museum Victoria, Victoria, Australia.

116 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

manillas (shackles), often inscribed with the Af- rican merchant’s name.158 In 1773, for example, Efik trader Robin John ordered from Liverpool merchant Ambrose Lace “large leg manillas with locks, and large iron manillas” for his “Room of irons,” suggesting that John had a small prison in his courtyard house compound.159 Once sold, most Africans would be housed on board the ship—usually below deck—until the vessel was prepared to depart for the Caribbean.

Shore to Ship The view of the ocean and the slave ship surely incited a range of emotions. Writing about his experience in the 1760s, Equiano reported: “The first object which saluted my eyes was the sea, and a slave ship, which was then riding at an- chor.  .  .  . These filled me with astonishment,

which was soon converted to terror, when I was carried on board.”160 As the majority of people sold into slavery came from Africa’s interior, few had ever seen the ocean, a European, a sailing ship, or the ever-present sharks that favored fac- tory sites and followed slave ships.

The first leg of the middle passage was the voyage from shore to ship.161 In many cases this passage was undertaken via canoe provided by African merchants and manned by free African “canoemen,” but sometimes slaves were trans- ported in the yawl or longboat carried on the ship.162 For the enslaved headed to those ships an- chored in the calmer waters of a river delta, this voyage was brief and generally uneventful. For the majority, however, the passage was harrow- ing and occasionally fatal. Without the protection of a harbor or river, ships along the Gold Coast

Figure 22. “Old Calabar Factories, Near Duke Town.” The large galleried house is likely a manufactured house imported from Liverpool. From H. M. Stanley, The Congo and the Founding of Its Free State: A Story of Work and Exploration (New York: Harper and Brothers, 1885).

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 117

and the Windward Coast typically anchored well past the surf in a band of the coastal waters com- monly called “the roads,” which at times could be as much as a mile off shore.163 Like the canoes used on Old Calabar, those used to pass through the surf were usually fashioned from a single large tree and could shuttle as many as twenty captives, who usually remained chained in pairs for the short voyage to the ship (Figure 23).164 For the majority the ocean was both unfamiliar and terrifying, and reports of resistance, small- scale insurrection, and suicide of those boarded on canoes are numerous. For some their fear of the ocean’s mighty waves was justified, espe- cially in the rainy season from April to July.165

The overloading of canoes or the violence of the surf sometimes resulted in an overturned canoe. Chained in pairs and often unable to swim, Af- ricans thrown from a canoe perished by drown- ing or were attacked by the sharks.166 Sometimes, they threw themselves overboard willingly. One English captain reported: “The Negroes are so willful and loth to leave their own country, that they have often leap’d out of canoes . . . into the sea, and kept underwater till they are drowned.”167

Once on board the deck of the ship, the ar- riving Africans were stripped naked, if they had not been already, and then were inspected by the captain and the ship’s surgeon if the ship re- tained one.168 This inspection took place under a

Figure 23. Image of canoe under construction, Elmina, Ghana. Ruins of a pier rise from the sand of the beach behind. Photograph by Louis P. Nelson, March 2011.

118 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

thatched roof that was built over a majority of the deck to provide shade during the stay along the African coast.169 The inspection was humiliating and terrifying. Equiano’s description of the expe- rience is a rare report:

I was immediately handled and tossed up, to see if I were sound, by some of the crew, and I was not persuaded that I had got into a world of bad spirits and they were going to kill me. Their com- plexions, too, differing so much from ours, their long hair, and the language they spoke united to confirm in me this belief. Indeed, such were the horrors of views and fears at that moment that if ten thousand worlds had been my own, I would freely have parted with them all to have exchanged my condition with that of the meanest slave in my own country.170

Those remaining on board were then usually shaved and sent into the secure chambers below deck; those rejected by the captain or the surgeon boarded the canoe once again for the return jour- ney to the shore.171

Since Africans were usually loaded on ships soon after their purchase by the captain, those Africans purchased early in a ship’s season along the coast of Africa often had a long and miser- able waiting period before the ship departed. Surviving accounts suggest that ships often re- mained along the coast for at least three months, sometimes as long as a year, waiting to collect a full cargo of captives.172 Ship’s captains or Brit- ish merchants trading in Africans often wrote to the governor of Cape Coast Castle in advance of their departure for Africa requesting the gov- ernor begin collecting and reserving prime Af- ricans for their voyage in an attempt to reduce the wait along the coast.173 Because captains widely believed the newly enslaved more likely to engage in insurrection or suicide while the coastline was still in sight, captives were often kept below deck from the time of purchase to the time of departure from the coast. Most historians estimate that as a result of this long wait, often entirely below deck, just as many if not more Africans died while waiting to depart as dur-

ing the middle passage itself.174 After months of harboring in river deltas or cruising “the roads,” the ship would eventually fill to capacity, and the captain would depart the coast. Under horrific conditions approximately three million enslaved Africans disembarked from the coast of Africa on English and American slavers and began the months-long voyage of the middle passage to the New World.175

Real Spaces What can we learn from such a history of archi- tectural experiences? Understanding the archi- tecture of the slave trade helps to ground the horror and the tragedy of these events. We need reminders that this history did not happen in the abstract, that it took place in real spaces and times. We must also come to recognize that these spaces—monumental or missing—were compo- nents of a massive international trading scheme that generated enormous wealth for a few, mur- dered millions, and laid the foundations for early modern and modern international trade and capitalism. This is especially important because so many of the spaces associated with “the most magnificent drama in the last thousand years of human history,” to quote W. E. B. Du Bois, are now gone. The castles remain, but they are not the whole story.

The transatlantic slave trade was an extra- ordinary machine that displaced approximately nine million people, incidentally killing an addi- tional five million. Approximately one in three of those enslaved in Africa perished in the process of capture, containment, and passage or in the first year of life in the Americas. Yet there are no marked battlefields and few sites of memory. The strategies of resistance survive most clearly in the songs and dances of village elders. Coffles con- structed temporary spaces—between the yoke and bodies that it conjoined by force. Thankfully, all of these spaces are now entirely gone, with the sole reminder being the one known surviving yoke in Liverpool, likely brought back as part of the abolitionist cause. The castles along the coast of Ghana have become increasingly important sites of heritage tourism, and that is to be cele-

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 119

brated. But their actual functions and interrela- tionships remain poorly understood. Most castles were not major holding stations for thousands or even hundreds of enslaved Africans at a time but subsidiary providers sending small lots of five or eight, often by canoe, to the headquarters at Cape Coast. The majority of enslaved Africans never even saw a castle; they were held in the small African-built mud-walled cells of a factory or in open-air barracoons near the coast, waiting for the arrival of the next slaver ship. If some came from castles and others from factories, the one experience shared by all was the middle passage. Marcus Rediker’s extraordinary book on the sub- ject, The Slave Ship, is an excellent corrective to the unfeeling distance offered by too many histo- rians of the subject.176 In his history people have names, feelings. In many ways, his wonderful and horrible book inspired this article. Lastly, for architectural historians this essay reminds us of the importance of understanding spaces through the processes that generated them. Historians of architecture can no longer be contented solely with histories of making: the grand drama of real life, of everyday life, is just too important.

au thor biogr aphy Louis P. Nelson is associate professor of archi- tectural history and associate dean of Research and International Programs in the School of Ar- chitecture at the University of Virginia. He is a former editor of Buildings & Landscapes.

notes The research and writing of this article has depended on the support of many people. The Dean’s Office of the School of Architecture provided a travel grant to support one of the two field research trips that under- gird this article. Nicholas Ivor, director of the Cape Coast Castle, gave me permission to spend days at a time studying the fabric of the article’s central build- ing. He was also enormously helpful in providing sug- gestions for further research. An early version of this article was presented at the Shadows of Empire confer- ence hosted at the University of Ghana, Legon, by Kofi Baku and John Kwadwo Osei-Tutu. I am immensely

grateful to them and the many other participants at the conference; this article incorporates many of their comments and suggestions. Max Edelson and Maurie McInnis both read early drafts of this article and pro- vided essential critique. I received very helpful com- mentary from the two blind reviewers; whoever you are, thank you. I am indebted to Jason Truesdale, who produced the plans that illustrate this printed version of the article and the three-dimensional model avail- able in the digital supplement. And last, I owe a debt of gratitude to Marta Gutman and Cindy Falk, B&L’s spectacular editors, who pushed me toward a far bet- ter product through two rounds of revisions.

1. Scholarship on the internal African slave trade includes David Northrup, Trade without Rulers: Pre- colonial Economic Development in South Eastern Nige-

ria (Oxford, U.K.: Clarendon Press, 1978); B. G. Der, The Slave Trade in Northern Ghana (Accra, Ghana: Woeli Publishing, 1998); A. M. Howell, The Slave Trade and Reconciliation: A Northern Ghanaian Per-

spective (Accra, Ghana: Bible Church of Africa, 1998); Ann Bailey, African Voices of the Atlantic Slave Trade: Beyond the Silence and the Shame (Boston: Beacon Press, 2006); A. B. Stahl, “Historical Process and the Impact of the Atlantic Slave Trade in Banda, Ghana c.  1800–1920,” in West Africa during the Atlantic Slave Trade: Archaeological Perspectives, ed. C. R. De Corse, 38–58 (London: Leicester University Press, 2001); J. Boachie-Ansah, “Archaeological Research at Kasana: A Search for Evidence on the Historic Slave Trade in the Upper West Region of Ghana,” Journal of Environment and Culture 2, no. 1 (2005): 35–57; Joe Miller, “The Slave Trade in Congo and Angola,” in The African Diaspora: Interpretive Essays, ed. Martin Kilson and Robert Rotberg, 75–113 (Cambridge, Mass.: Har- vard University Press, 1976). The best sources on the slave castles of West Africa include William St. Clair, The Door of No Return: The History of the Cape Coast

Castle and the Atlantic Slave Trade (New York: Blue- bridge, 2006); Arnold Walker Lawrence, Trade Castles and Forts of West Africa (Stanford, Calif.: Stanford Uni- versity Press, 1964); Arnold Walker Lawrence, Forti- fied Trade-Posts: The English in West Africa, 1645–1822 (London: Cape, 1969); Albert Van Danzig, Forts and Castles of Ghana (Accra, Ghana: Sedco Publishing, 1980). See also Ray Kea, Settlements, Trade, and Poli- ties in the Seventeenth-Century Gold Coast (Baltimore,

120 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

Md.: Johns Hopkins University Press, 1982). The con- nections between slaving in Africa and plantations in the Caribbean are hard to ignore. The English settled St. Kitts in 1623 and Barbados two years later. In 1630 the English acquired their first permanent castle site in Africa at Cormantin. Similarly, the Danes built Fort Fredensborg only three years after their purchase of the island of St. Croix.

2. The term slave hole was used in the middle of the eighteenth century to refer specifically to a cham- ber within the strong walls of a bastion used to con- tain slaves. These were often ventilated by holes in the paved surface of the bastion. By the 1790s these spaces were commonly referred to as prisons. See Law- rence, Fortified Trade-Posts, 181, 223 (for slave hole); 69, 70, 148, 168 (for slave prison).

3. Cugoano tells his story in Thoughts and Senti- ments on the Evil and Wicked Traffic of the Slavery and

Commerce of the Human Species (London, 1791), 7–9. 4. Henri Lefebvre, The Production of Space, trans.

Donald Nicholson-Smith (Malden, Mass.: Blackwell, 1991).

5. I undertook fieldwork in Ghana in the spring of 2011 and again in the summer of 2012. Over the course of those trips, I examined eight different cas- tles and completed measured drawings of one, Cape Coast Castle.

6. Many historical issues do not concern me here. The economic implications of changes in markets or the preferences of planters for Africans from different regions, for example, are both important components of the historiography, but as they do not directly im- plicate the experience of the enslaved in space, they play no role in this telling. For discussion, see David Galenson, Traders, Planters, and Slaves: Market Behav- ior in Early English America (New York: Cambridge University Press, 1986); and Daniel Littlefield, Rice and Slaves: Ethnicity and Slave Trade in Colonial South

Carolina (Baton Rouge: Louisiana State University Press, 1981), chap. 2.

7. K. G. Davies, The Emergence of International Business, 1200–1800: The Royal African Company, vol. 5 (London: Taylor and Francis, 1957).

8. Michel Foucault, Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan (New York: Pantheon Books/Random House, 1977).

9. On the recognition of humanity in criminals, see Foucault, Discipline and Punish, 74, 125.

10. Barbot quoted in Akosua Perbi, A History of In- digenous Slavery in Ghana (Accra, Ghana: Sub-Saharan Publishers, 2004), 29. See J. Barbot, Description of the Coasts of North and South Guinea, 5 vols. (London, 1732).

11. Northrup, Trade without Rulers, 68–80. 12. Both Bowman and Owen quotes cited in

Thorkild Hansen, Coast of Slaves, trans. Kari Dako (Accra, Ghana: Sub-Saharan Publishers, 2002), 48.

13. Another account was written by Louis Asa-Asa, quoted in Mark Rediker, The Slave Ship (London: John Murray, 2007), 102.

14. Cugoano, Thoughts and Sentiments, 7–9. 15. Transcribed from Asa-Asa’s telling and printed

in The History of Mary Prince: A West Indian Slave (Ann Arbor: University of Michigan Press, 1997), 51.

16. Northrup, Trade without Rulers, 66, 78. 17. For recent work on the internal slave trade and

the impact of European involvement, see “African Paths to the Middle Passage,” chap. 3 in Rediker, The Slave Ship.

18. S. A. Diouf, ed., Fighting the Slave Trade: West Af- rican Strategies (Athens: Ohio University Press, 2003). See Hilary Beckles, “African Resistance to the Trans- atlantic Slave Trade,” in The Transatlantic Slave Trade: Landmarks, Legacies, Expectations, ed. James Kwesi An- quandah, 81–91 (Accra, Ghana: Sub-Saharan Publish- ers, 2007); Naana Jane Opoku-Agyemang, “Living the Experience of the Slave Trade in Sankana and Gwollu: Implications for Tourism,” in Transatlantic Slave Trade, 210–24.

19. A. M. Howell, “Showers of Arrows: The Reac- tions and Resistance of the Kasana to Slave Raids in the 18th and 19th Centuries,” in Transatlantic Slave Trade, 189–209.

20. Opoku-Agyemang, “The Living Experience of the Slave Trade,” 214.

21. The use of caves as refuge appears elsewhere in West Africa. See Thiero Mouctar Bah, “Slave-Raiding and Defensive Systems South of Lake Chad from the Sixteenth to the Nineteenth Centuries,” in Fighting the Slave Trade: West African Strategies, ed. Sylviane A. Diouf, 15–30 (Athens: Ohio University Press, 2003).

22. Opoku-Agyemang, “The Living Experience of the Slave Trade,” 216–17.

23. Opoku-Agyemang, “The Living Experience of the Slave Trade,” 218–20. Very similar strategies were implemented farther east. See Bah, “Slave-Raiding and Defensive Systems South of Lake Chad,” 15–30; Martin

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 121

Klein, “Defensive Strategies: Wasulu, Masina, and the Slave Trade,” in Fighting the Slave Trade, 62–78.

24. Elisee Soumonni, “Lacustrine Villages in South Benin as Refuges from the Slave Trade,” in Diouf, Fighting the Slave Trade, 3–14.

25. See Bruce L. Mouser, A Slaving Voyage to Africa and Jamaica: the Log of the Sandown, 1793–94 (Bloom- ington: Indiana University Press, 2002).

26. Thomas Clarkson, Letters on the Slave Trade (London, 1791), 34–37.

27. H. C. Monrad, Description of the Guinea Coast and Its Inhabitants (Accra, Ghana: Sub-Saharan Pub- lishers, 2010), 219.

28. Cited in Hugh Thomas, The Slave Trade: The Story of the Atlantic Slave Trade, 1440–1870 (New York: Simon and Schuster, 1999), 381.

29. Estimates of mortality before even boarding the slave ship range from highs of 25 percent to lows of 5 percent. For more, see Joseph Miller, Way of Death, Mer- chant Capitalism, and the Angolan Slave Trade, 1730–1830 (Madison: University of Wisconsin Press, 1988); Pat- rick Manning, The African Diaspora: A History through Culture (New York: Columbia University Press, 2008).

30. Thomas, The Slave Trade, 381. 31. For a useful discussion of the internal trade in

those areas, see Miller, “The Slave Trade in Congo and Angola,” in Kilson and Rotberg, The African Diaspora.

32. Perbi, History of Indigenous Slavery, 40. 33. Perbi, History of Indigenous Slavery, 37–38; See

also S. N. Nkumbaan, “Historical Archaeology of Slave Route Case Study of Kasana and Sankana” (mas- ter’s thesis, University of Ghana, Legon, 2003).

34. Mary Prince, The History of Mary Prince (1831; repr., New York: Pandora Press, 1987), 52.

35. On Salaga, see Perbi, History of Indigenous Slav- ery, 44–47.

36. On the importance of the Asante to the inter- nal slave trade, see Perbi, History of Indigenous Slavery, 20–23.

37. J. A. Okoro, “Indigenous Water Management, Slavery and Slave Trade in Salaga” (unpublished paper, Department of Archaeology, Legon, 2003).

38. See the 1705 account reprinted in James An- quandah, “Researching the Historic Slave Trade in Ghana: An Overview,” in Transatlantic Slave Trade, 23–56; quote on 28–29.

39. On slave camps, see Akosua Perbi and Yaw Bredwa-Mensah, “Slave Camps in Pre-colonial Ghana:

The Case of Jenini in the Brong Ahafo Region,” in Transatlantic Slave Trade, 138–47.

40. Stephen Behrendt, A. J. H. Latham, and David Northrup, The Diary of Antera Duke, an Eighteenth- Century African Slave Trader (New York: Oxford Uni- versity Press, with the assistance of the International African Institute, 2010), 110–13.

41. Northrup, Trade without Rulers, 64. 42. Northrup, Trade without Rulers, 101. 43. Randy Sparks, The Two Princes of Calabar:

An Eighteenth-Century Atlantic Odyssey (Cambridge, Mass.: Harvard University Press, 2004), 49. For a broad study of the African canoe, see Robert Smith, “The Canoe in West African History,” Journal of Afri- can History 11, no. 4 (1970): 515–33.

44. Account of Isaac Parker appears in Sheila Lambert, ed., House of Commons Sessional Papers of the Eighteenth Century, vol. 73 (London: Scholarly Re- sources Incorporated, 1975), 126.

45. Olaudah Equiano, The Interesting Narrative of Olaudah Equiano, ed. Vincent Carretta (1789; repr. Penguin, 2003), 32.

46. The account of David Henderson appears in Sheila Lambert, ed., House of Commons Sessional Pa- pers of the Eighteenth Century, vol. 69 (London: Schol- arly Resources Incorporated, 1975), 13.

47. Johannes Rask, A Brief and Truthful Descrip- tion of a Journey to and from Guinea (1754; repr., Accra, Ghana: Sub-Saharan Publishers, 2008), 156–57.

48. St. Clair, The Door of No Return, chap. 5. 49. This complexity is made clear in “Soldiers and

Workers,” chap. 5 in St. Clair, Door of No Return. 50. Monrad, A Description of the Guinea Coast, 258. 51. Monrad, A Description of the Guinea Coast, 216. 52. Monrad, A Description of the Guinea Coast, 220. 53. Monrad, A Description of the Guinea Coast, 221. 54. Davies, Emergence of International Business, 5:9. 55. A great summary of this history appears in

Davies, Emergence of International Business, 5:12–15, 41–44, 213.

56. Barbot quoted in Lawrence, Fortified Trade- Posts, 184.

57. The best assessment of the physical develop- ment of Cape Coast and other West African castles is in Lawrence, Trade Castles and Forts of West Africa.

58. St. Clair, The Door of No Return, 45–46. 59. Davies, Emergence of International Business,

5:241–43.

122 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

60. Eveline Martin, British West African Settle- ments, 1750–1820 (New York: Longmans, 1927), 7. For a full account of the financial management of the Brit- ish West African castles, see also Davies, Emergence of International Business.

61. For an excellent summary of the development of these early castles, see Davies, Emergence of Interna- tional Business, 5:243–49.

62. “James Nightingale to Cape Coast Castle, Feb 15 1681,” MS Rawl C. 745, “Copies of Letters sent by the Out Factors Of the Royal African Company of England to the Chief Agents at Cape Coast Castle, Jan 27, 1680 to June 8, 1681,” Letterbooks of the Royal African Com- pany, Bodelian, Special Collections Library, Oxford.

63. “James Nightingale, Feb 20 1681,” MS Rawl C. 745, “Copies of Letters sent by the Out Factors Of the Royal African Company of England to the Chief Agents at Cape Coast Castle, Jan 27, 1680 to June 8, 1681,” Letterbooks of the Royal African Company, Bodelian, Special Collections Library, Oxford.

64. “1682: from Ralph Hassell,” MS Rawl C. 745, “Copies of Letters sent by the Out Factors Of the Royal African Company of England to the Chief Agents at Cape Coast Castle, Jan 27, 1680 to June 8, 1681,” Let- terbooks of the Royal African Company, Bodelian, Special Collections Library, Oxford.

65. For more on this, see Davies, Emergence of In- ternational Business.

66. St. Clair, Door of No Return, esp. chap. 8. 67. Monrad, A Description of the Guinea Coast, 212.

The records of the Company of Merchants offer a very similar list. See Martin, British West African Settle- ments, 45.

68. As scholars of early West African history and of the slave trade have come to argue, the internal African slave trade was much exacerbated by European involve- ment, especially through the introduction of guns— the most highly desired trade good. On this, see Jo- seph Inikori, “Changing Commodity Composition of Imports into West Africa, 1650–1850: A Window into the Impact of the Transatlantic Slave Trade on African Societies,” in Transatlantic Slave Trade, 57–80.

69. Martin, British West African Settlements, 50. 70. See Lawrence, Fortified Trade-Posts, 157. See

also Rask, A Brief and Truthful Description, 58. 71. Davies, Emergence of International Business,

5:226. 72. Davies, Emergence of International Business,

5:188; see also Door of No Return, chap. 8. For numer- ous examples of this, see John Newton, The Journal of a Slave Trader, 1750–1754, ed. Bernard Martin and Mark Spurrell (London: Epworth Press, 1962).

73. Davies, Emergence of International Business, 5:222, 262.

74. Lawrence, Fortified Trade-Posts, 292. 75. Lawrence, Fortified Trade-Posts, 187. 76. Lawrence, Fortified Trade-Posts, 298. 77. Lawrence, Trade Castles and Forts of West Africa,

190. 78. Davies, Emergence of International Business,

5:226. 79. Cape Coast is described in 1710 as having “re-

positories for one thousand slaves.” Davies, Emergence of International Business, 5:241.

80. St. Clair, Door of No Return, 77. 81. Barbot quoted in Lawrence, Trade Castles and

Forts of West Africa, 166. 82. St. Clair, Door of No Return, 223. 83. Davies, Emergence of International Business,

5:241. 84. Davies, Emergence of International Business,

5:242. 85. For a full account, see David Richardson, Bristol,

Africa, and the Eighteenth-Century Slave Trade to Amer-

ica, 3 vols. (Bristol: Bristol Record Society, 1986–90). 86. “London, May 28, 1756,” “Letterbook of the

Company of Merchants Trading to Africa, 1751–1769,” GB 0402 SSC/22, 2 vols., vol. 1, “Letters from the Com- pany to the Governor and Council of Cape Coast Castle, 1751–68,” The Royal Geographical Society, London.

87. “Letterbook of the Company of Merchants Trad- ing to Africa, 1751–1769,” GB 0402 SSC/22, 2 vols., vol. 1, “Letters from the Company to the Governor and Council of Cape Coast Castle, 1751–68,” p. 24, n.d., The Royal Geographical Society, London.

88. “London, June 24, 1761,” “Letterbook of the Company of Merchants Trading to Africa, 1751–1769,” GB 0402 SSC/22, 2 vols., vol. 1, “Letters from the Company to the Governor and Council of Cape Coast Castle, 1751–68,” p. 141, The Royal Geographical So- ciety, London.

89. “Sept 5, 1760,” “Letterbook of the Company of Merchants Trading to Africa, 1751–1769,” GB 0402 SSC/22, 2 vols., vol. 1, “Letters from the Company to the Governor and Council of Cape Coast Castle, 1751–68,” p. 135, The Royal Geographical Society, London.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 123

90. See Lawrence, Fortified Trade-Posts, 123. See also Thorkild Hansen, Coast of Slaves, trans. Kari Dako (Accra, Ghana: Sub-Saharan Publishers, 2002), 50.

91. See the Oxford English Dictionary, 2nd ed., s.v. “hole.”

92. St. Clair, Door of No Return, 220. 93. Lawrence, Fortified Trade-Posts, 302–4. 94. Monrad, A Description of the Guinea Coast,

221–22. 95. An excellent history of the castle at Anomabo

appears in St. Clair, Door of No Return, chap. 2. See also Lawrence, Trade Castles and Forts of West Africa, 349–55.

96. Stephen Behrendt, “‘Journal of an African Slaver,’ 1789–1792, and the Gold Coast Slave Trade of William Callow,” History in Africa 22 (1995): 61–71.

97. Martin, British West African Settlements, 17–18. 98. Martin, British West African Settlements, 49. 99. On the shipment of materials from England,

see “Letterbook of the Company of Merchants Trad- ing to Africa, 1751–1769,” GB 0402 SSC/22, 2 vols., vol. 1, “Letters from the Company to the Governor and Council of Cape Coast Castle, 1751–68,” The Royal Geographical Society, London.

100. Readers are encouraged to view the full three- dimensional model posted on Buildings & Landscape’s digital edition hosted by JSTOR (www.jstor.org).

101. D. Simmonds, “A Note on the Excavations in Cape Coast Castle,” Transactions of the Historical Soci- ety of Ghana 14, no. 2 (1973): 267–69.

102. St. Clair, Door of No Return, 212–13. 103. St. Clair, Door of No Return, 80. 104. This data was collected by the author from the

Voyages project for the specific purpose of interrogat- ing the volume of slaves departing from various cas- tles over the course of the period in question. These queries were undertaken in the spring of 2013.

105. Monrad, A Description of the Guinea Coast, 221. 106. Cugoano, Thoughts and Sentiments, 9. 107. Monrad, A Description of the Guinea Coast,

222. See also Hansen, Coast of Slaves, 82, 171. 108. St. Clair, Door of No Return, 81. Monrad sug-

gests only once a day; see Monrad, A Description of the Guinea Coast, 222.

109. St. Clair, Door of No Return, 72–73. 110. St. Clair, Door of No Return, 75. 111. St. Clair, Door of No Return, 63. 112. Monrad, A Description of the Guinea Coast, 266.

113. St. Clair, Door of No Return, 129. 114. Monrad, A Description of the Guinea Coast,

228. 115. Quoted in Beckles, “African Resistance to the

Transatlantic Slave Trade,” 81–91. 116. Monrad, A Description of the Guinea Coast,

222. A very similar description from the 1690s sug- gests that this was common practice throughout the period of the slave trade; see Hansen, Coast of Slaves, 29–30.

117. John Barbot in 1746, as cited in Hansen, Coast of Slaves, 31.

118. St. Clair, Door of No Return, 78; Lawrence, Trade Castles and Forts of West Africa, 190.

119. “Letterbook of the Company of Merchants Trading to Africa, 1751–1769,” GB 0402 SSC/22, 2 vols., vol. 2, “Letters from the Agent at Cape Coast Castle to the Royal African Company, 1767–69,” pp. 96–97, November 6, 1767, The Royal Geographical Society, London.

120. Foucault, Discipline and Punish, 137. In mak- ing this argument, I am aware of the objections by many to Stanley Elkin’s comparison in the late 1950s of plantation slavery and Nazi concentration camps. My argument is that West African slave castles and European prisons emerged contemporaneously and participate in a shared discourse on architectures of correction and (re)formation. See Stanley M. Elkins, Slavery: A Problem in American Institutional and In-

tellectual Life (Chicago: University of Chicago Press, 1959). For objections and debate, see Ann Lane, The Debate over Slavery: Stanley Elkins and His Critics (Urbana: University of Illinois Press, 1971); and John Blassingame, The Slave Community: Plantation Life in the Antebellum South (New York: Oxford University Press, 1972).

121. Foucault, Discipline and Punish, 184–87. 122. Lawrence, Fortified Trade-Posts, 69, 70, 148,

168. 123. Lawrence, Fortified Trade-Posts, 168. 124. Foucault, Discipline and Punish, 122. 125. See Davies, Emergence of International Busi-

ness, 5:242. 126. There are a number of excellent works on

the history, economics, and social structures of Old Calabar; for example, see Behrendt, Latham, and Northrup, The Diary of Antera Duke; and Sparks, The Two Princes of Calabar. On the importance of Calabar

124 |   BUILDINGS & LANDSCAPES  21,  no.  1 ,  SPRING 2014

to Jamaica, see Trevor Burnard and Kenneth Morgan, “The Dynamics of the Slave Market,” 208.

127. Rediker, The Slave Ship, 44. 128. Nicholas Owen, Journal of a Slave-Dealer (1759;

repr., London: Routledge and Sons, 1930), 33, 85, 105. 129. Owen, Journal of a Slave-Dealer, 100–101. 130. These barracoons may have been similar to

the places for the containment of slaves on Gorée Is- land called captiverie in the eighteenth century. See Lambert, House of Commons Sessional Papers, 73:25.

131. Robin Law, Ouidah: The Social History of a West African Slaving Port, 1727–1892 (Athens: Ohio Univer- sity Press, 2004), 137.

132. Account reprinted in James Anquandah, “Researching the Historic Slave Trade in Ghana: An Overview,” in Transatlantic Slave Trade, 23–56, quote on 28–29.

133. Frederick Forbes, Six Months’ Service in the Af- rican Blockade (London: R. Bentley, 1849), 113.

134. The complexity of these centers is discussed in detail in the introductory essays in Behrendt, Latham, and Northrup, The Diary of Antera Duke. On the im- portance of Liverpool, see 76–77.

135. On the Efik and Old Calabar, see Behrendt, Latham, and Northrup, The Diary of Antera Duke; Jo Anne Chandler, Efik Traders of Old Calabar (London: Dawsons of Pall Mall, 1963). See also Two Princes of Calabar; for an earlier eighteenth-century history, see Stephen Behrendt and Eric Graham, “African Mer- chants, Notables, and the Slave Trade at Old Calabar, 1720: Evidence from the National Archives in Scot- land,” History in Africa 30 (2003): 37–61. For a com- prehensive discussion of the structure of the slave trade in Angola, see Joseph Miller, “Some Aspects of the Commercial Organization of Slaving at Luanda, Angola, 1730–1830,” in The Uncommon Market: Essays in the Economic History of the Atlantic Slave Trade, ed. Henry Gemery and Jan Hogendorn (New York: Aca- demic Press, 1979). For a careful study of a slave port in the Bight of Benin, see Law, Ouidah.

136. Behrendt, Latham, and Northrup, The Diary of Antera Duke, 49.

137. Behrendt and Graham, “African Merchants, Notables, and the Slave Trade,” 50; Robin Hallett, ed., Records of the African Association, 1788–1831 (London: Thomas Nelson and Sons, 1964), 206.

138. Behrendt and Graham, “African Merchants, Notables and the Slave Trade,” 52.

139. I am aware of the critique of Equiano’s nar- rative as a partial fabrication. Nonetheless, I take the position that even if he was American-born, he would have known plenty of African-born slaves and their own stories of capture and shipment. Equiano, Inter- esting Narrative, 22.

140. Rev. Hope Masterton Waddell, Twenty-Nine Years in the West Indies and Central Africa, 1829–58 (London: Frank Cass, 1970), 247–48.

141. Northrup, Trade without Rulers, 89, 100. 142. Lambert, House of Commons Sessional Papers,

73:26. 143. Behrendt, Latham, and Northrup, The Diary

of Antera Duke, 16. 144. Hallett, Records of the African Association,

200. 145. Waddell, Twenty-Nine Years, 243. 146. Waddell, Twenty-Nine Years, 250. 147. Donald Simmons, “An Ethnographic Sketch

of the Efik People,” in Efik Traders of Old Calabar, ed. Daryll Forde (London: Oxford University Press, 1968), 16–17.

148. Hallett, Records of the African Association, 200, 203.

149. Hallett, Records of the African Association, 203. 150. Hallett, Records of the African Association, 203. 151. Reported by Antera Duke in his diary. See

Forde, Efik Traders of Old Calabar, 50. For a more re- cent discussion, see Behrent, Latham, and Northrup, The Diary of Antera Duke, 36–39. Human sacrifice was found among a number of West African com- munities; see Rask, A Brief and Truthful Description, 141–44.

152. The earliest example thus far is Egbo Young’s Liverpool Hall, so named by Antera Duke, in 1785. Forde, Efik Traders of Old Calabar, 28. See also Wad- dell, Twenty-Nine Years, 244; and Simmons, “An Eth- nographic Sketch of the Efik People,” in Forde, Efik Traders of Old Calabar, 9. Such exportation of house frames was also presumed among the English. In his discourse advocating for the English colonization of West Africa, Wadstrom recommends that all colonists bring a house frame with them. See Wadstrom, Essay on Colonization, vol. 1 (1794), 49.

153. Hallett, Records of the African Association, 207–8.

154. Behrendt, Latham, and Northrup, The Diary of Antera Duke, 57.

LOUIS P.  NELSON, ARCHITECTURES OF WEST AFRICAN ENSLAVEMENT  | 125

155. Owen, Journal of a Slave-Dealer, 76. 156. Behrendt, Latham, and Northrup, The Diary of

Antera Duke, 56–57. 157. Herbert Klein, Atlantic Slave Trade (Cam-

bridge: Cambridge University Press, 2010), 116–21. 158. Behrendt, Latham, and Northrup, The Diary

of Antera Duke, 57; Mungo Park, Travels in the Interior Districts of Africa (London, 1799; repr., Durham, N.C.: Duke University Press, 2000), 18.

159. Behrendt, Latham, and Northrup, The Diary of Antera Duke, 56–57.

160. Equiano, Interesting Narrative, 33. 161. The most important source for accurate in-

formation on the middle passage and shipboard ex- perience of the slave is Rediker, The Slave Ship. An interesting comparison to the transportation of slaves is the transportation of convicts. See Kenneth Mor- gan, “The Organization of the Convict Trade to Mary- land: Stevenson, Randolph, and Cheston, 1768–1775,” William and Mary Quarterly, 3rd ser., 42, no. 2 (April 1985): 201–27.

162. Rediker, The Slave Ship, 70. 163. St. Clair, The Door of No Return, 14. 164. Monrad, A Description of the Guinea Coast,

223; St. Clair, The Door of No Return, 25.

165. Law, Ouidah, 135. 166. Monrad, A Description of the Guinea Coast,

223; Rediker, The Slave Ship, 38. 167. Cited in Thomas, The Slave Trade, 404–05. 168. Law, Ouidah, 143. 169. Rediker, The Slave Ship, 143, 233–34. 170. Equiano, Interesting Narrative, 33. 171. St. Clair, The Door of No Return, 223; see also

Mouser, A Slaving Voyage, 104. 172. St. Clair, The Door of No Return, 11; Rediker,

The Slave Ship, 234; Davies, The Royal African Com- pany, 186.

173. St. Clair, The Door of No Return, 220. 174. Davies, The Royal African Company, 293. 175. For statistical information on the slave trade,

see Philip Curtin, The Atlantic Slave Trade: A Census (Madison: University of Wisconsin Press, 1969). For a detailed discussion of the conditions on slave ships, see Rediker, The Slave Ship; and Charles Gland and Herbert S. Klein, “The Allotment of Space for Slaves aboard Eighteenth-Century British Slave Ships,” Wil- liam and Mary Quarterly, 3rd ser., 42, no. 2 (April 1985): 238–48.

176. Rediker, The Slave Ship.