Pawnship and the Business of Slaving

The Business of Slaving: Pawnship in Western Africa, c. 1600-1810 Author(s): Paul E. Lovejoy and David Richardson Source: The Journal of African History, Vol. 42, No. 1 (2001), pp. 67-89 Published by: Cambridge University Press Stable URL: http://www.jstor.org/stable/3647216 Accessed: 28/04/2009 14:24 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=cup. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected]. Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to The Journal of African History. http://www.jstor.org Journal of African History, 42 (2001), pp. 67-89. ? 2001 Cambridge University Press WESTERN PAWNSHIP SLAVING: OF BUSINESS THE AFRICA, BY PAUL IN c. 1600-1810* E. LOVEJOY York University, and DAVID 67 Printed in the United Kingdom Toronto RICHARDSON University of Hull The use of people as pawns to underpin credit was widespread in western Africa during the era of the trans-Atlantic slave trade. This study examines where and when pawns were used in commercial transactions involving It is shown that European European slave merchants in the period c. i6oo-i8io. merchants relied on pawnship as an instrument of credit protection in many places, though not everywhere. Europeans apparently did not hold pawns at Ouidah (after I727), at Bonny or on the Angolan coast. Nonetheless, the reliance on pawnship elsewhere highlights the influence of African institutions on the development of the slave trade. ABSTRACT: KEY WORDS: Western Africa, slavery, slave trade, economic. Common 'A pledge is a slave's brother' saying on Sierra Leone coast, c. I78os1 THE institution of pawnship, specifically the use of people as collateral for credit, helped to underpin the Atlantic slave trade.2 In an earlier article, we the role of pawnship in focused on the port of Old Calabar, demonstrating there.3 In this study we local with merchants British merchants of the trade * The research for this paper was funded in part by a grant from the Social Sciences and Humanities Research Council of Canada, for Lovejoy, and a University of Hull research grant, for Richardson. The authors wish to thank Robin Law, Jose Curto and Joseph C. Miller for their references and comments, and Silke Strickrodt for her assistance in the Public Record Office. An earlier version was presented at the African Studies Association Annual Meeting, Philadelphia, I4 Nov. 1999, at a session organized by Kristin Mann. 1 Testimony of Robert Hume, who was on eleven slaving voyages to western Africa in the late eighteenth century; see Minutes of Evidence taken on the Second Reading of the Bill entitled, 'An Act to prohibit the Trading for Slaves on the Coast of Africa, within certain limits', in F. William Torrington (ed.), House of Lords Sessional Papers I798-99 (Dobbs Ferry, NY, I974), III, I 12. 2 For a general discussion of human pawning in Africa, see Toyin Falola and Paul E. Lovejoy (eds.), Pawnship in Africa: Debt Bondage in Historical Perspective (Boulder, 1994) and the various studies therein. Also see E. A. Oroge, 'Iwofa: An historical survey of the Yoruba institution of indenture', African EconomicHistory, I4 (1985), 75-Io6; and Paul E. Lovejoy, Transformationsin Slavery: A History of Slavery in Africa (Cambridge, 2nd. ed., 2000), 13-14, 109-10, 117-I8, 143, I69, 176, I82. Paul E. Lovejoy and David Richardson, 'Trust, pawnship, and Atlantic history: The institutional foundations of the Old Calabar slave trade ', American Historical Review, 104 3 (i999), 333-55. 68 PAUL E. LOVEJOY AND DAVID RICHARDSON show that Old Calabar was not unique, that people were used as pawns in the export slave trade in the seventeenth and eighteenth centuries elsewhere in Western Africa. Hence this study contributes to our understanding of the 'trust' system, as the commercial arrangements involving credit were often known.4 We distinguish between individuals reportedly enslaved for debt, people used as pawns in the slave trade but not sold, and people seized or 'panyarred', usually for debts or for crimes. It is not always possible to distinguish among these categories in reference to sources of slaves for the transatlantic trade. A small minority of enslaved Africans crossed the Atlantic as a result of pawning, but relatively low numbers are not an indication of the importance of pawning in Western Africa or its significance in the Atlantic trade. As William Snelgrave reported in the early eighteenth century, those being held for debt on the Slave Coast were rarely sold, 'few of these come into the hands of Europeans, being kept by their Countrymen for their own use'.5 We argue that the presence of pawns in enslaved populations provides evidence about credit mechanisms in the domestic and international economies of those parts of Africa affected by the Atlantic slave trade that help to explain how Africa interacted with the larger Atlantic world. Our aim is to examine the available primary documentation on pawning and the slave trade. While we cannot explore all the issues this suggests, we would highlight several important distinctions. First, pawning was not important in the trans-Atlantic slave trade everywhere or in all periods of the trade. Second, the factors that seem to have determined whether or not pawning was important include the local political and judicial context in which slave trading operated. Specifically, pawnship was less important where there was a strong centralized authority, such as the port of Ouidah during the Dahomey period after 1727 and at Bonny throughout its history. Third, in Senegambia, where Islamic law was in force and Muslim merchants dominated trade, pawning was not important in the export trade, except as a source of individuals from non-Muslim areas who had been entrapped through debt to Muslims. The institution was of dubious legal status in Islamic law and hence could not easily form the basis of credit arrangements.6 In Senegambia as well as other places in Upper Guinea and 4 For a discussion of credit and 'trust', see for example Alan Ryder, Benin and the Martin Lynn, Commerce (London, I969), 78, I30-3, 140, 322-4; and Economic Change in West Africa: The Palm Oil Trade in the Nineteenth Century (Cambridge, I998), 67-8; Philip D. Curtin, Economic Change in Precolonial Africa: Senegambia in the Era of the Slave Trade (Madison, 1975), 302-8; and Joseph C. Miller, Way of Death: Merchant Capitalism and the Angolan Slave Trade, I730-I830 (Madison, Europeans, I485-1897 I988), I75-89, 284-313, 537-41, 570-97, 598-604. 5 William Snelgrave, A New Account of Some Parts of Guinea and the Slave Trade (London, I734), I59. 6 It is perhaps significant that pawnship was prevalent in those places where Islam was relatively unimportant or not present at all. Muslim merchants employed other mechanisms to secure credit, and pawnship had a questionable status in Islamic law anyway. Evidence from the nineteenth and twentieth centuries indicates that Muslim merchants often treated non-Muslim pawns as slaves who could be sold. Muslims purchasing pawns seldom had personal relationships with those deposing of pawns, and hence there was little to prevent sale into slavery. The institution of pawnship exposed individuals to the risks of enslavement in the Muslim world, just as it did in the Atlantic, THE BUSINESS OF SLAVING 69 in Angola, credit was tied to marriage arrangements between local women and foreign merchants. Analysis of the institution has ultimately to explain why pawnship was adapted to the export slave trade in seemingly similar ways in many parts of Western African but not throughout the coast. In considering the role of pawnship in the slave trade, we address the thesis of Walter Rodney that on the Upper Guinea coast, and by extension elsewhere, servile institutions were not significant early in the Atlantic slave trade but became more so under the influence of trans-Atlantic slavery. Rodney claims there was increasing reliance on enslavement as a form of punishment, as commercial and political institutions adjusted in the context of the slave trade; the incidence of slavery and 'other forms of social oppression' became more pronounced.7 According to Rodney, 'to borrow and fail to repay the most trifling item in the seventeenth century was sufficient warrant for arrest, conviction, and execution by sale.... [P]eople [were] sold for debt'.8 Rodney failed, however, to examine enslavement for debt more fully, which we contend increased in ways that are directly traceable to the Atlantic slave trade. Pawnship seems to have been in use in the seventeenth century and enforcing debt repayment through the reduction of pawns to slavery seems to have become common as the export slave trade grew. The role of pawnship in the slave trade suggests that Rodney's thesis can be extended to include changes in the practice of this institution. We note that pawnship was not universally adopted in transactions along the African coast; variations in its use are a subject for further study. Moreover, our approach differs from Rodney's in that he emphasized the relationship of European credit to African oppression, while we highlight European adaptability to African practices. We argue that the relationship between capitalism and slavery, as formulated initially by Eric Williams, is revealed in its complexities through the ways in which European slave traders adjusted to the existence of local credit arrangements by relying on servile relationships other than slavery.9 Williams did not explore credit mechanisms in operation in Africa and did not therefore appreciate how European capital penetrated local markets there. Pawnship tied local credit arrangements to the slave trade, thereby exposing people to the risk of enslavement as well as underpinning the trade itself. It and such risks ultimatelyunderminedthe practiceof using people to secure credit; but not as long as the slave trade was in operation.For a discussion of Islamic commercial practicesin West Africa, see John Hunwick, 'Islamic financialinstitutions: theoretical structures and some of their practicalapplicationsin sub-SaharanAfrica', in Endre Stiansen and Jane Guyer (eds.), Currencies,Credit and Culture: African Financial Institutionsin HistoricalPerspective(Uppsala, I999). Also see Paul E. Lovejoy, Caravans of Kola: The Hausa Kola Trade, I70oo-I900 (Zaria and Ibadan, I980), and Lovejoy, Salt of the Desert Sun: A History of Salt Productionand Trade in the Central Sudan (Cambridge, I986). On the question of pawning in Islam, see Falola and Lovejoy, Pawnship in Africa, 14-15. 7 WalterRodney, 'Africanslaveryand other forms of social oppressionon the Upper Guinea Coast in the context of the Atlantic slave trade', Journalof African History, 7 (1966), 431-43; and A History of the Upper Guinea Coast, I545 to I8oo (Oxford, 1970). 8 Accountof de Anguiano,apparentlyfrom Buenaventurade Carrocera(ed.), Misiones Capuchinasen Africa, II, I45 [no place or date of publicationgiven], cited in Rodney, Upper Guinea Coast, I09. 9 Eric Williams, Capitalism and Slavery (London, 1944). PAUL 70 E. LOVEJOY AND DAVID RICHARDSON is not always clear when and where pawning became important. The historical record establishes, however, that pawning was common at the Gold Coast and Bight of Benin in the seventeenth century and at Upper Gabon and the Loango coast by the Guinea, Old Calabar, Cameroons, second half of the eighteenth century. Of the various types of servility that underpinned the Atlantic world, slavery has rightly attracted most attention, but the existence of indenture and convict labour, impressed sailors and child labour reveals that servile relationships more generally were fundamental to the development of the early modern world. Pawnship, enslavement for debt and panyarring were other forms of servility in Western Africa closely with slavery that sometimes associated resulted in transport to the Americas.l? PAWNSHIP IN WESTERN AFRICA James Phipps, an agent for the Royal African Company on the Gold Coast in i710-22, described pawning among local merchants as 'the Custom among themselves for letting out money or money worth'."1 Similarly in Fraser of who in traded Africa for Bristol, twenty years, 1790 Captain James claimed that pawning was the 'custom of the country'.12 In both cases, the 10 According to Ebiegberi J. Alagoa and Atei M. Okorobia ('Pawnship in Nembe, Niger delta', in Falola and Lovejoy, Pawnship in Africa, 79), pawnship was unknown in the Niger delta before the era of the transatlantic slave trade, although there is no empirical evidence to support their claim. Ray Kea has also postulated a late development for pawnship, again linking the origins of the institution to the slave trade. According to Kea, pawning was an 'institutionalized practice ... associated with credit-debt relations... [and was] well established in the coastal and inland towns in the early seventeenth century; [its] origins ... perhaps to be sought in the early sixteenth century or the later fifteenth century', that is when the Atlantic slave trade began (see Settlements, Trade and Polities in the Seventeenth-Century Gold Coast [Baltimore, I982], 244). The absence of information on pawning before the seventeenth century makes such a discussion premature, if not impossible. For example, in his study of the upper Guinea coast, George Brooks does not mention pawning (see Landlords and Strangers: Ecology, Society, and Trade in Western Africa, I000-I630 [Boulder, 1993], 143-96). Moreover, pawning is not evident in the external trade of Senegambia, according to Curtin, Senegambia in the Era of the Slave Trade, 247, 302-8; Boubacar Barry, Senegambia and the Atlantic Slave Trade (Cambridge, I997), 26-50; and James F. Searing, West African Slavery and Atlantic Commerce: The Senegal River Valley, I700-I860 (Cambridge, I993), 93-Io8. In the interior of Angola, where pawnship was widely practised, the linkages between pawning and the Atlantic trade are far from clear, but it does not seem that pawning was used to underpin the slave trade, except on the Loango coast; see Miller, Way of Death, 48-70, 94-103, 124, 179-80, 305-6; and Isabel de Castro Henriques, Commerceet changementen Angola au XIXe siecle: Imbangala et Tshokweface a modernite (Paris, 1995), I, 97-139. John Thornton Surprisingly, does not mention pawnship in his study of the slave trade. The omission in the first edition of his book is perhaps understandable because it focuses on the period before the eighteenth century, but the second edition includes the eighteenth century, and there is no mention of pawnship (Africa and Africans in the Making of the Atlantic World, I400-I800 [Cambridge, rev. ed., I998]). 11 John Phipps, undated, but probably before 1720, in reference to trade at Ketu, Public Record Office (PRO), C 113/274. 12 Testimony of James Fraser, 29 Jan. 1790, in Sheila Lambert (ed.), House of Commons Sessional Papers of the Eighteenth Century, Minutes of Evidence on the Slave Trade I790, Part i (Wilmington, 1975), LXXI, 15. THE BUSINESS OF SLAVING 7I is that was and of implication pawning indigenous long-standing. implicitly According to William Bosman, on the basis of his experience in the I690s: when a Negroe finds he cannot subsist he pawns himself for a certain Sum of Money, or his Friends do it for him; and he is maintained by the one to whom he has been pawned, on condition that he performs such work as he is ordered to do, which is not in the least slavish, being chiefly to defend his Master on occasion, and in sowing time to work as much as he himself pleases.13 Bosman thought the institution was 'commendable amongst the Negroes' because it served as a safety valve: 'we find no poor amongst them who beg; For no matter how little they possess, they never beg', the reason being that individuals were pawned. The practices described refer to transactions that had little if any connection with the external slave trade, although it is difficult to isolate the latter from the local economy. Pawning of people as collateral for credit was considered to be distinct from slavery. The saying quoted at the beginning of this paper - that 'a pawn is a slave's brother' - suggests that slavery and pawnship were related but not the same.'4 Antera Duke, a leading Old Calabar merchant in the I78os, used the English word 'pawn' ('pown', 'paun', or 'prown') in a diary that he kept and distinguishes 'pawn' from 'slave'."5 This distinction recurs in virtually all West African languages."6 Pawns and slaves were also separate social and legal categories, though in the past debt bondage was sometimes confused with slavery, just as it is today. Because of this confusion, it is not always clear that references to 'debt slavery' are to pawnship. Indebtedness sometimes led to actual enslavement and sale, since the failure to fulfill the terms of loans resulted in seizure for debts. In many cases those collectively held accountable for debts were also seized and even sold. One account of Whydah in the early eighteenth century suggested that failure by the head of family to repay debts 'within a fixed time, which happens very often' obliges 'him and all his family to do what the man who made the loan to him orders him like slaves .17 Similarly, Richard Miles, an agent for the Company of Merchants Trading to Africa on the Gold Coast in 1765-84, reported that he had known 'thousands of instances' of debt in which 'the person [was] sold as a Slave for the benefit of the creditor'.18 Whether or not these refer to pawning is difficult to assess, but the fact that individuals ended up as slaves seems clear. After the Dahomian conquest of 13 William Bosman, A New and Accurate Description of the Coast of Guinea [1705] (London, I967), I76. Also see Albert van Dantzig, 'English Bosman and Dutch Bosman: a comparison of texts', History in Africa, 4 (1977), I22, and for a discussion Kea, Gold Coast, 303. Seventeenth-Century 14 Testimony of Robert Hume, in Torrington,LordsSessionalPapers, III, 1 I2. 15 'Diary of AnteraDuke [Nteiro Edem Efiom]', in Daryll Forde (ed.), Efik Traders of Old Calabar (London, 1954), 86, 95, 109, I Io. 16 See Falola and Lovejoy, Pawnship in Africa, 13-15. 17 'Relationdu Royaumede Judasen Guinee ', Ms in Archivesd'Outre-Mer,Aix-en- Provence: Dep6t des Fortifications des Colonies, Cotes d'Afrique, Ms 104, 48, as cited in Robin Law, 'On pawningand enslavementfor debt in the pre-colonialSlave Coast', in Falola and Lovejoy, Pawnshipin Africa, 6i. 18 Testimony of RichardMiles, in Lambert, SessionalPapers, LXVIII, 37, 38, 40, 42. It should be noted that Miles had considerableknowledge of local practices on the Gold Coast and claimedto have been fluent in Twi. PAUL 72 E. LOVEJOY AND DAVID RICHARDSON Whydah in I727, refugees fled westward along the lagoon, settling near Grand Popo, but as a result of famine, some of them were apparently pawned and perhaps ultimately sold. In 1730, Snelgrave reported that, while the refugees 'were being supplied with what they wanted by their Neighbours', this 'constantly decreased their Numbers, they being obliged to sell their Wives, Children and Servants for Provisions, and other necessaries, because they had no Money left'.19 Again, the practice being described seems to be pawning, not actual sale into slavery, at least not initially. The term for pawn varies considerably in the languages along the West African coast. It is awowa in Akan,20 awubame in Ewe, awoba in Ga,21 iwofa in Yoruba,22 iyoha in Edo (Benin),23 and abrofa in Akwamu,24 all of which appear to be similar, and ubion in Efik/Ibibio,25 igba ibe in Igbo,26 pagi in Ijaw,27 and gbanu in Fon,28 where similarity is less obvious. This diversity of terms offers few clues about how pawnship functioned through time, but suggests that the institution was old and dispersed. The similarity of terms used in Akan, Ga, Yoruba and Edo raises the possibility that historical connections existed in the development of the institution in the Bight of Benin and the Gold Coast, or at least suggests that the term derives from an older, common term of great antiquity corresponding to the linguistic origins of these related languages. The documentary evidence for the use of pawns before the nineteenth century is nonetheless scanty, although circumstantial evidence suggests that pawning was indigenous to Western Africa. As a credit system, the pawning of individuals relied on social relationships, often kinship, to protect those being held in pawn, wherein membership in a kin group implied some insurance in situations of indebtedness. This attribute of pawnship suggests that the institution was indigenous. Kinship strategies certainly affected the practice of pawning. Richard Edwards, who was at the Gold Coast between 1752 and I755, found that 'Fathers pawn their Children, for they rely on the Descendants of the Female Branches of the Family for Heirs, as in that Case they are sure of the Blood; and they have much less Consideration for their own Children'.29 19 20 Snelgrave, Guinea and the Slave Trade, 113. Raymond Dumett and Marion Johnson, 'Britain and the suppression of slavery in the Gold Coast colony, Ashanti and the northern territories', in Suzanne Miers and Richard Roberts (eds.), The End of Slavery in Africa (Madison, 1988), 76. 21 Akosua Adoma Perbi, 'A history of indigenous slavery in Ghana from the 15th to the i9th centuries' (Ph.D. thesis, University of Ghana, I997), 23. 22 According to Robin Law, iwofa is derived from the verb fa, to pledge, and was not applied to non-human forms of collateral, such as farms, althoughfa was so used (Falola and Lovejoy, Pawnship in Africa, 25, n). Also see Samuel Johnson, The History of the Yorubas (London, 1973), 126. Uyilawa Usuanlele, 'Pawnship in Edo society: from Benin kingdom to Benin province under colonial rule', in Falola and Lovejoy, Pawnship in Africa, 105. 24 According to Kea (Seventeenth-Century Gold Coast, 105), the term also meant free peasant, so may not have specifically meant 'pawn'. 23 25 Hugh Goldie, 26 Felix K. Dictionary of the Efik Language (Farnborough, I964), 300. Ekechi, 'Pawnship in Igbo society', in Falola and Lovejoy, Pawnship in Africa, 83. David Northrup (Trade without Rulers: Pre-Colonial EconomicDevelopment in South-Eastern Nigeria [Oxford, 1978], 73) claims that the terms for pawn and slave are 27 Alagoa and Okorobia, 'Pawnship in Nembe', 72. the same - ohu. 28 Law, 'Pawning and enslavement for debt', 55. 29 Testimony of Richard Edwards, in Lambert, Sessional Papers, LXIX, 44. THE BUSINESS OF SLAVING 73 These comments refer to the matrilineality of Akan society, although whether or not fathers actually wanted to pawn their children is open to question. In effect, Edwards was indicating an order of priority of who would be pawned if necessary, the matrilineal line being protected, thereby suggesting an indigenous origin.30 Those held for debt could normally expect some degree of protection from sale, but once sale took place, people were slaves, whether the sale was to European traders for export or otherwise. At Whydah in the I69os, Bosman found that the only time that 'parents here sell their children, men their wives, and one brother the other', were under conditions of 'necessity, or some great crime'.31 Bosman implicitly concedes that some of those enslaved in this way ended up being sold for export, but he clearly regarded this as rare, suggesting that there may have been restrictions on the export of pawns, not that pawning was unusual. and the logic of kinship structures suggest pawnship was of Terminology considerable antiquity. Available documentation thus presents glimpses into an institution that was intertwined with secret societies, religious systems and regional market demands. Pawning did not always involve people. Gold, jewelry, cloth and even trees and land could be pawned, although pawning of the latter is not documented before the nineteenth century.32 It follows that the use of goods as collateral for debt was not tied only to human beings but was clearly understood as a mechanism for securing debts more generally. This suggests that human collateral was a variant of a larger practice whose antiquity is difficult to document but whose origins were not necessarily connected with slavery.33 Moreover, the practice was different from enslavement for debt. 30 Similarly, Snelgrave (Guinea and the Slave Trade, 159) learned at Whydah that 'it is common for some inland People, to sell their Children for Slaves, tho' they are under no Necessity for so doing; which I am inclined to believe. But I never observed that the People near the Sea Coast practice this, unless compelled thereto by extreme Want and Famine'. Snelgrave was referring to pawnship in the context of the recent conquest of Whydah by Dahomey and the flight of refugees from the conquered town. The further alienation through sale into slavery was a danger arising from the resulting deprivation. 31 Bosman, Coast of Guinea, 364. As Law notes, the reference to 'necessity' should probably be understood to allude to pawning to secure debts, rather than outright sale; see 'Pawning and enslavement for debt', 64. 32 For the pawning of trees, see Falola and Lovejoy, Pawnship in Africa, 8. For the pawning of land, see Kristin Mann, 'The political economy of credit and the culture of obligation in late nineteenth-century Lagos', unpublished paper presented at the annual meeting of the African Studies Association, Philadelphia, 1999. 33 In the Kingdom of Benin, a unit of account known in its anglicized form as ' pawns' was employed in trade. In the I79os, John Adams learned that 'The medium of exchange is salt, and calculations are made in pawns, one of which is equal to a bar in Bonny, or 2S. 6d. sterling'; see John Adams, Remarks on the Country Extendingfrom Cape Palmas to River Congo (London, i823), i 16. Christopher Hasell, a Liverpool trader active in the Benin trade in 1770, kept his accounts in 'pawns', including purchases of provisions, with details of goods exchanged for yams, goats and other foodstuffs. In a letter dated Benin Gato, 9 Aug. 1770, Hasell addresses an unknown person, but almost certainly the master of the True Blue, indicating that he had 'broke trade' on 30 July: 'The dashees and Customs are a deal larger than ever was known[.] Trade is but Dead and slaves but indifferent & I broke Trade at Ioo pawns pr Man, & 90ofor A Woman, Webster [another trader] giving 95 for A Woman & 115 for A man'; see ' Accounts of the vessel, True Blue, which traded to Benin in 1770', 25 pages, Christopher Hasell Papers, 1763-1773. (The PAUL 74 PAWNING E. LOVEJOY AND DAVID AS A DETERRENT RICHARDSON AGAINST PANYARRING Pawnship was distinct from 'panyarring', a term used in Africa to refer to the arbitrary seizure of people or goods for debt or an alleged injury. In contrast to pawning, where security was provided beforehand, panyarring involved the seizure of goods or people to force payment or their sale to compensate the creditor. As Robin Law has noted, pawning was an institutionalized practice that attempted to prevent panyarring.34 Those panyarred were hostages who had to be redeemed to settle the debt or otherwise risked being sold into slavery. Samuel Brun, who traded at the Gold Coast in the i6ios, learned that where 'an individual of one town was in debt to a person in another town', the creditor might seize an individual from the debtor's town and hold him 'until the real debtor arrives '. If the latter 'does not come, then the innocent person is sold, [and] can no longer be redeemed as [he] is taken very far inland from whence he never returns'.5 The reference to 'inland' highlights the indigenous nature of the transaction and enforcement of a contract without reference to trans-Atlantic slavery. Brun's comment also linked panyarring to collective responsibility for debt, a point also noted by Bosman who traded on the Gold Coast in the I690s. According to Bosman: A distinguished Person in one Country hath Money owing him from a Person of an adjacent Country, which is not so speedily paid as he desires; on which he causes as many Goods, Freemen, or Slaves to be seized by violence and rapine in the Country where his Debtor lives, as will richly pay him; the People so seized he claps in Irons, and if not redeemed has them sold, in order to raise Money for the payment of the Debt. If the Debtor be an honest man and the Debt just, he immediately endeavours by the satisfaction of his Creditors to free his Countrymen.36 The point was further reinforced in the early eighteenth century by Phipps who observed that at the Gold Coast creditors could 'take... Satisfaction of the Borrower or any of his friends or Townspeople all of which are liable to answer the Debt'.37 Moreover, panyarring was also the recourse open to papersareheld privatelyin Cumbria.)As Alan Ryderhas demonstrated,however,these 'pawns' were actuallypieces of local cloth, and the original Portuguesepano or pagne renderedin English as 'pawn', see Beninand the Europeans,207, 230. We wish to thank Robin Law for drawingRyder'sanalysisto our attention. 34 35 Law, 'Pawning and enslavement for debt', 62-4. Samuel Brun, Schiffarten(Basel,I624), 76-7, as discussed in Kea, SeventeenthCenturyGoldCoast,244-5. Brooks(Landlordsand Strangers,205) also mentionsthe use of hostages on the upper Guinea coast, apparentlyin the late sixteenth century, but without references. 36 For Bosman's description,see Coastof Guinea,176. Also see van Dantzig, 'English Bosmanand Dutch Bosman';and Kea, Seventeenth-Century GoldCoast,245. There are numerousexamplesof panyarringin the Rawlinsonpapersdating to theI68os; see the various entries in Robin Law (ed.), The Englishin West AfricaI681-1683: The Local Correspondenceof the Royal African Company of England,I68I-I699, Part I (Oxford, 1997), Letters 28, 39, 40, 51, 59, II0, 122, 157, 69, 172, 177, 2 I, 226, 264, 280, 288,322, 330, 358, 383, 394, 400, 420, 480, 484, 485, 487, 504, 518, 525, 526, 533, 639. 37 In Report of James Phipps, Fetu, undated, c. 1720 or earlier, PRO CII3/274. FebruaryI699, the ruler of Abora tried to seize a caravan of Akyem traders because he claimed he was owed 7 bendas (one benda was worth two ounces troy in gold), not by THE BUSINESS OF SLAVING 75 creditors who wished to put pressure on defaulters in the Bight of Benin. In the late seventeenth century, the King of Allada threatened merchants in default that 'all their wives would be taken', presumably as pawns but saleable as forfeiture on the debts.38 Similarly, des Marchais reported that at Whydah 'the custom... is hard on debtors, because a creditor is permitted to seize his debtor in lieu of payment and to sell him'.39 Panyarring caused people to be held as if they were hostages, not pawns, but like pawns, they could be redeemed. In the case of hostages there was often a time limitation on redemption, while in pawning this was not always the case. In the early eighteenth century, Snelgrave found that 'Debtors who refuse to pay their Debts [at Whydah], or are insolvent, are... liable to be made Slaves; but their Friends may redeem them'. Snelgrave notes explicitly that debtors might in the end suffer sale if their 'friends' were unable or unwilling to pay their redemption, in which case the debtors were 'generally sold for the benefit of their creditors'.40 On the Gold Coast, agreements might specifically preclude panyarring if the contracted debt was in default. One such case involved several Twifo traders who owed John Kabes of Little Komenda a substantial sum of gold.41 Pawns had greater protection against sale into slavery than hostages, but both practices could result in enslavement.42 Pawning theoretically secured the loan in advance, thereby forestalling arbitrary seizure and sale into slavery, but in fact pawns were sometimes sold. In both cases, individuals were exposed to the possibility of enslavement if the debt was not settled and the persons being held were not redeemed. Holding of pawns appears to have been an institutional response to avoid arbitrary practices associated with panyarring, but this did not always work. USE OF PAWNS AS SECURITY IN TRADE WITH EUROPEANS Pawns came to the attention of contemporary European observers, whose accounts provide most of the precolonial evidence on pawnship in Western Africa, because they were involved in the institution. Europeans recognized pawnship and indeed participated in its enforcement and thereby benefited from its provisions, although, once again, distinctions between pawning, panyarring and enslavement for debt were not always clear. European merchants knowingly bought debtors, pawns and hostages who had been seized for debts. William James learned at Old Calabar that 'some of the Slaves sold to the Europeans are such as have become so in consequence of members of the caravan but by other Akyem merchants. In I703, Okofomu, an Akani merchant resident in Eguafo seized two Akani traders because Apodi, the king, was in debt for a large sum of gold. See Kea, Seventeenth-Century Gold Coast, 245. For a discussion of gold values, see ibid. I90-I. 'Voyage du Delbee', as cited in Law, 'Pawning and enslavement for debt', 61. 39 Des Marchais, 'Journal du Voiage', f.5ov, as cited in Law, 'Pawning and enslavement for debt', 61 40 Snelgrave, Guinea and the Slave Trade, 158-59. Law considers Snelgrave's reference to 'slavery' 'an inexact allusion to pawning'; see 'Pawning and enslavement for debt', 41 Kea, Seventeenth-Century Gold Coast, 245. 6I-2. 42 See, for example, the discussion in Per 0. Hernaes, Slaves, Danes, and African Coast 38 Society (Trondheim, I998), I9. 76 PAUL E. LOVEJOY AND DAVID RICHARDSON Debt' without otherwise distinguishing among the victims.43 Likewise, Luis Antonio de Oliveira Mendes mentions indebtedness as a major reason for enslavement in west-central Africa in the late eighteenth century, though his observations apparently do not refer to pawns.44 In I827, the governor of Benguela ordered the freeing of 'many Africans imprisoned by European traders on the pretext that their forefathers had failed to liquidate debts', suggesting enslavement for debt but probably not pawning.45 Cases where enslavement was associated directly with personal debt always seem to have constituted a small portion of the slave trade.46 Many European merchants nevertheless used pawnship to underpin credit, specifically in the form of advances of goods against slaves or other goods.47 Safeguarding their interests therefore involved Europeans in credit arrangements that depended upon local mediation and enforcement and inevitably drew them into pawning arrangements. The abolitionist Thomas Clarkson noted in 1788 that African merchants received goods on credit from the slave ships 'for the purpose of slaving those ships, on whose account they travel'. He went on to note that the merchants were 'obliged to leave a pledge or security for their return', the pledge consisting of 'their own relations, who are detained till they come back'.48 The extent to which Europeans relied on pawnship is difficult to assess, but reflecting on twenty years' experience, James Fraser observed in 1790 that 'by the custom of most parts of the coast on which I have traded, we are not at liberty to refuse a pawn'.49 Although Europeans relied on pawns as collateral for credit advances on 43 Testimony of William James, in Lambert, Sessional Papers, LXIX, 48-49. Robert Hume was alleged to have taken 30 pawns to the Americas on one ship that he commanded on the Windward Coast in the I780s; see ibid. III, 99, 109-II. One ship sailed from Old Calabar in the I78os with 120 pawns on board; see PRO C 107/I2, Richard Rogers to James Rogers, Old Calabar, Apr. 1788 [exact date not recorded]. 44 Luis Ant6nio de Oliveira Mendes, Memdria a Respeito dos Escravos e Trafico da Escravatura entre a Costa d'Africa e o Brazil [1793] (Porto, I977), 40-I. 45 Ralph Delgado, A Famosa e Histdrica Benguela: Catdlogo dos Governadores(1779 a I940) (Lisbon, 1945), 99. Also see Maria Emilia Madeira Santos, Nos caminhosde Africa: serventia e posse (Angola seculo XIX) (Lisbon, I998). 46 Among the liberated slaves interviewed by linguist S. Koelle in Sierra Leone in 1849-50, approximately 7 per cent claimed to have been enslaved for debt in some form. Many of these 'recaptives' came from the interior of the Bights of Benin and Biafra and the region of the upper Guinea coast and Sierra Leone. See P. E. H. Hair, 'The enslavement of Koelle's informants', Journal of African History, 4 (1965), I99. For an analysis of the same data for the Bight of Biafra, see Northrup, Trade Without Rulers, 80. 47 The evidence for the use of pawns in the slave trade comes largely from English, Dutch and Danish sources, which may reflect the incomplete research into the sources for French and Portuguese trading practices. While there is some information on French involvement in pawning, there is virtually nothing on Portuguese use of pawns. It may be that the Portuguese trade in west-central Africa did not rely on pawning because Portuguese merchants and their Luso-African and African partners actually transported merchandise inland themselves or had the possibility of doing so; see Miller, Way of Death, 48-70, 115-26, 179-89, 305-7. Thomas Clarkson, An Essay on the Slavery and Commerceof the Human Species, & particularly the African [1785] (London, 2nd edn., 1788), 27-8. The second edition of Essay on Slavery and Commerce included Clarkson's observations arising from his research in Liverpool and Bristol in 1788, as well as letters and other materials sent to him. 49 Testimony of James Fraser, 29 Jan. 1790; in Lambert, Sessional Papers, LXXI, 15. 48 THE BUSINESS OF SLAVING 77 many parts of the African coast, people were only one type of collateral. As Ray Kea has demonstrated, gold, jewelry and other valuables, not just be humans, might pawned. According to Kea, van Groenestein, governor of the Danish fort at Fredriksborg, relied on various types of pawned goods to secure credit.50 In 1694, pawned gold was accepted on board ships at Cape Coast as collateral for goods advanced on credit. One report noted that 'some of the best traders will come and declare us to give them credit for the value of two or three marks of gold' and went on to indicate that '[c]ommonly they will leave some pledge in our hands till payment, as great collars of gold, which they will be sure to redeem'. 5 Fifty years later, in 1744, parcels of gold dust that had been pawned and taken to Barbados were expected to be returned to the Gold Coast unopened to ensure that there had been no tampering.52 At least on the Gold Coast, therefore, merchants accepted both goods and people as collateral for debts.53 On the Gold Coast, where gold was 50 According to Kea, 'the entries show that a number of these [local] persons had current accounts with van Groenestein and that these accounts enabled them to make payments to receive commodities and/or money (gold) by way of written orders, that is by credit transfers from one account to another. Loans were advanced against the current accounts and against deposited securities (pawns), which took the form of gold dust, gold ornaments, brass basins, textiles, people and so on'. In one transaction, for example, Fitero Ahen purchased goods to the amount of 1,296 dambas; his cousin conducted the trade on his behalf, leaving gold as security on the balance of 144 dambas (8 dambas = I ounce troy). Ahen's cousin paid the balance and redeemed the gold in the presence of a witness, Obiri ('Abrij'), one of the trading servants at Fredriksborg. See Kea, Seventeenth-Century Gold Coast, 227, 228, based on the accounts journal of van Groenestein at Fredriksborg. The entry is dated 2 Dec. 1668, and reads: 'Received from his cousin, and returned to him the gold pawn of 2 marks i ounce 9 angles in the presence of Abrij'. Kea, Seventeenth-Century Gold Coast, 237. Most advances were to people with political status, including ahenfo, asanfo, and batafo. 51 Ibid. 238. The report also noted that the traders promised in a certain number of days when their servants or boys return from the inland countries to deliver the produce of such merchandise. 52 The pawned items were 'wrought and festish gold for sundry people on the coast of Africa... for Debts due ye concerned and in time must be delivered'; see PRO Chancery Masters Exhibits, C 103/I30, Papers of Thomas Hall, George Hamilton to Hall, Barbados, 23 May I744. The ruler of Afutu pawned gold valued at 27,168 dambas at Fredriksborg in Feb. I674. According to Kea (Seventeenth-Century Gold Coast, 238, some of this gold had probably served as surety for an earlier loan in Aug. 1669. 241-2), In Nov. I701, the 'king and country of Saboe' left 33 bendas worth of gold pawns with the English at Cape Coast, on which 40 'fuzees', 6 barrels of gunpowder, 63 half-firkins of tallow, 48 blue perpetuanas, and 51 sheets were credited. Between I689 and I701, the pawns that the Asebuhene deposited at various times at Cape Coast Castle had a combined value in excess of Ioo,ooo dambas. In Aug. I693, the ruler of Tantumkwerri pawned gold jewelry 'weighing nine ounces' (3,456 dambas) at the Royal African Company factory, against an advance of 35 English carpets. 53 Hence Jan Claessen Cutta of Afutu borrowed 6,144 dambas in gold from van Groenestein in Aug. I669; 'several pawns were kept at the fort' as surety, including a trading servant of the company, pledged for 768 dambas. What else was pawned is not mentioned but may have included gold ornaments and jewelry. See Kea, SeventeenthCentury Gold Coast, 231. Similarly, in late Aug. 1676, the debts of a deceased merchant at Cape Coast who owed van Groenestein '25 benda 5 english' were secured through the pawning of 'a large Negro sword embellished with gold and Corfu Braffo's son named Obing'. In July I713, the ruler of Twifo received an advance of oo00 bendas in gold, apparently for state purposes. His securities consisted of gold jewelry, aggrey beads and 78 PAUL E. LOVEJOY AND DAVID RICHARDSON panned and mined and was in demand among Europeans, this was, of course, possible, but it was not so at Old Calabar and other places where there were few items other than people that could be pawned. Theoretically, pawned items could be interchanged or sold if not redeemed, but when pawns were people, sale meant enslavement. Human pawns were therefore not only acceptable as collateral; unredeemed pawns also became a source of slaves for the trans-Atlantic traffic. The Bristol captain, James Fraser, noted in I790 that 'Every pawn that is received is considered a Slave, until he is redeemed - if their friends refuse, or are not able to redeem them, they are carried off and sold'.54 Similarly, when asked if he knew that at the time that 'Pledges were delivered [to] you... they were Slaves or Freeman', Robert Hume replied that 'We always conceive that a Pledge if not redeemed is a Slave; and [that] they consider themselves as such when they come on board'.55 Hume went on to suggest, however, that while the export of unredeemed pawns was legitimate, concern over the consequences of such action cautioned against it. Specifically, he claimed: I would not carry them off for fear of injuring my future Voyages, or my Interest with the Natives... [I]t is a general Practice with all Ships to send these Pledges on Shore, should they even lose by them, for fear of hurting their Interest with the Natives.56 Disruption of trade was precisely the outcome of one such occurrence in 1788 when the shipment of 30 pawns by a vessel from Bimbia, Cameroon, caused local merchants to seize the captains and crew of two British ships in retaliation. The two captains involved later claimed that their voyages were 'entirely ruined, the Natives being determined to make no further Trade with either of us, nor pay the above Debts [of 30 pawns] until their Sons, Daughters, and &c. are returned'.57 The extent to which Europeans adopted local African practices in relation to debt security is also indicated by the fact that on occasion Europeans themselves were held as pawns. When African merchants extended credit to ships or feared arbitrary seizure by ships, Europeans might be required to provide collateral in a form that was locally acceptable. This might mean receiving European crew as pawns. Thus, while trading at Alampo on the Gold Coast in March i682, Charles Towgood reported sending 'one whiteman ashoare [sic] to stay till I sailed, for they sent off word they would 'his friends and wives'. For these and other examples, see ibid. 231, 235, 241. Also see George Metcalf, 'Gold, assortmentsand the trade ounce: Fante merchants and the problem of supply and demand in the I770s', Journal of African History, 28 (1987), 32; and letterof RichardThelwall,Anamabu,25 Sept. I683, in Law, Englishin WestAfrica, I40. 54 Fraser (Lambert, SessionalPapers, LXXI, 15) operated on the African coast for twenty years, beginning in I772, and his comments applied to 'Angola', by which he probablymeant the Loangocoast, but also to the upper Guinea coast and 'other parts'. 55 Evidenceof Robert Hume, in Torrington,LordsSessionalPapers, III, 112. It was in this context that Hume suggested that 'To use the African language,a Pledge is a Slave's Brother'. 56 Ibid. III, IIo. Hume also noted that 'all Pledges that are carriedoff in generalfret, get sickly, and often cause a great Mortalityin Ships'. 57 Letter of James McGauty, William Willoughby and Isaac Nixen, Lambert,SessionalPapers,LXIX, 318. I Mar. 1788, in THE BUSINESS OF SLAVING 79 not bring off one slave before I sent one, for fear of being panyard'.58 Pawning as a form of insurance against panyarring or as security for credit clearly transcended, therefore, the Afro-European commercial nexus. Its adoption by both Africans and Europeans was essential in the regularization of trade on the Western African coast. When Europeans advanced goods to local merchants, they required the pawns as security, but in cases where European ships were in debt to local merchants, the latter were known to demand pawns from among ships' crew. In the export trade, pawnship worked both ways, though given that credit largely went into Africa, it was African merchants who provided most pawns. PREVALENCE OF PAWNING IN THE BUSINESS SLAVE TRADE OF THE ATLANTIC The incidence of pawning in Western Africa probably grew in tandem with the growth of slave exports to the Americas, although pawnship was not adopted as a mechanism for underpinning credit relationship in all parts of the coast. On the Gold Coast, the Danes, Dutch and English relied, as we have seen, on pawning throughout the seventeenth and eighteenth centuries.59 The English held pawns in special enclosures outside their trade castles. In the middle of the eighteenth century, they were reportedly held on board a special ship, which Edwards described as 'the Floating Factory'. By 1789 when he testified before Parliament, he claimed that pawns were 'kept in the Forts'.60 We have demonstrated elsewhere that pawnship helped to underpin British trade at Old Calabar by the middle of the eighteenth century, though it was not apparently significant at nearby Bonny.61 The Old Calabar case is particularly revealing because there is documentary evidence provided by local merchants to complement the more extensive accounts and testimonies of Liverpool and Bristol merchants. In 1763, the master of a Liverpool brig reported getting 'pledges' from the 'Kings Town [i.e. Old Town], Dukes and Tom Henshaws [towns]'.62 John Ashley Hall, a London captain who 58 Charles Towgood, Alampo, 19 Mar. I682, in Law, English in West Africa, 271-2. Similarly,Hugh Shearsreportedin Sept., laterin the sameyear,that '3 slavesmore I was forced to pay to Capt Ware'saccount, for the Negroes kept Capt Nur's man and mine which wee gave for pawnes, and would not permit them to come aboard, for Capt Ware did panyar 3 freemen and slaves, and wee had much adoe to gett them to trade with us' (Hugh Shears, Alampo, 30 Sept. 1682; ibid. 276). 59 In addition to the various citations above, see also the numerous entries in Law, English in West Africa, especially letters 15, i6, 33, 40, 112, i22, 226, 379, 387, 388, 397, 400, 414, 440, 464, 465, 477, 479, 533, 584, and 586, which demonstrate that the Royal African Company regularly engaged in pawning in the early i68os at its various factories on the Gold Coast. 60 Testimony of Richard Edwards, in Lambert, Sessional Papers, LXIX, 44. 61 Lovejoy and Richardson, 'Trust, pawnship and Atlantic history'. 62 Letter of James Berry of Liverpool, 3 Apr. 1763, in Gomer Williams, History of the Liverpool Privateers and Letters of Marque, with an Account of the Liverpool Slave Trade (London, 1897), 533. Also see evidence of James Morley, in Lambert, Sessional Papers, LXXIII, I49-54. Morley told the Parliamentary Enquiry that his ship, the Amelia, was at Old Calabar in I763 or I764, but from other evidence, it appears that the voyage in question was made in I761-2; see David Richardson (ed.), Bristol, Africa and the Eighteenth-Century Slave Trade to America (Bristol, 1991), III, I39. 8o PAUL E. LOVEJOY AND DAVID RICHARDSON traded at Old Calabar in 1772-3, also accepted pawns as collateral, claiming before Parliament in I790 that it was 'the way the trade is carried on' in the Calabar and Del Rey rivers, that is, the Cross river estuary. There are, indeed, many references to pawning in Antera Duke's diary in I785-8.63 At Gabon, where ships acquired ebony and redwood as well as slaves, pawnship has also underpinned trade, as Samuel Swan observed in I8I0. Noting that one could not 'avoid trusting' the local traders 'more or less', Swan observed that '[s]lave ships, and sometimes produce ships, take pawns when they trust money out', the pawns 'generally [being] the children of the head men who are trusted'. He went on to observe that '[w]hen there are several ships in the river it is best to get these pawns for Ivory, & Ivory pawns for wood, which keep till paid', a method said to 'save time & loss .64 The manipulation of the items to be held as pawns suggests that captains were well acquainted with local pawning practices and adjusted their business accordingly. Pawning was also common on the Loango coast, in areas where the British, French and Dutch operated, and perhaps elsewhere in the region as well.65 As early as 1617-18, Dutch traders were accepting pawns as collateral at Mpinda in the Kingdom of Kongo.66 Nearly two centuries later, in relation to British trade, it was said that 'the custom at Angola [i.e. the Loango coast]' was 'for traders and inhabitants to pawn their Slaves, their children or relations, to procure goods for different purposes'. It was also reported to be sometimes 'the custom with some great men in the country to order some of their friends, their relations, of families, to be sold: and... this being done in passion or resentment, the people who receive this commission chuse to deliver them as pawns, taking the value of a Slave in return', the master being free to choose 'to redeem the person so pledged'.67 The British ships were not alone in trading at Loango and it is assumed that pawnship was commonly underpinned the trade of others, including the French and Dutch, though there is little evidence of this in published accounts. Pawns were used as security for goods advanced on credit in upper Guinea as well, not only on the coast but also among those involved in long-distance 63 Antera Duke, 'Diary', 86, 95, 109, I0o. 64 Captain Samuel Swan, Memoranda on Africa, Ms, Library of Congress, c. 181o, in George E. Brooks, Jr., Yankee Traders, Old Coasters & African Middlemen: A History of American Legitimate Trade with West Africa in the Nineteenth Century (Boston, 1970), 338. 65 See Phyllis M. Martin, The External Trade of the Loango Coast, I576-1870 (Oxford, 1972), I03, 167, citing N. Uring, The Voyage and Travels of Captain Nathaniel Uring [1726] (London, I928), 34, 38; L. B. Proyart, Histoire de Loango, Kakongo et autres royaumes d'Afrique (Paris, I776), I36-7; J. Cuvelier (ed.), Documents sur une mission franfaise au Kakongo, I766-1776 (Brussels, 1953), 52; L. Degrandpre, Voyage a la cote occidentale d'Afriquefaits dans les annees 1786-I787 (Paris, I80o), 211-2. Also see Miller, Way of Death, 50, 52, 95, I79-80. 66 According to the report of Manuel Baptista, the Bishop of Angola and Kongo, Dutch 'factories' operating at Mpinda in I617-I8 accepted both male and female pawns; see Alfredo de Albuquerque Felner, Angola: apontamentos sobre a ocupafao e inz'cio do estabelecimento dos Portugueses no Congo, Angola e Benguela (extraidos de documentos histdricos) (Coimbra, 1933), 477. At the time, the Dutch were more interested in ivory than slaves. According to Jean Cuvelier, 'L'ancien Congo d'apris Pierre van den Broecke (i608-I6I2)', Bulletin des Sceances de l'Academie Royale des Sciences Coloniales, I (1955), 179, these pawns were called nsimbi. We wish to thank Jose Curto for this information. 67 Fraser's testimony, in Lambert, Sessional Papers, LXXI, 3. THE BUSINESS OF SLAVING 8i trade into the interior.68 John Matthews, who visited the coast in I785-7, claimed it 'is customary, indeed, for people of all ranks to put their children out as pledges'. He went on to note that 'they are careful either to redeem them in time or to pawn them to the resident traders or established factories', the pawns being 'generally considered as a protection for your property'.69 Consistent with this, Captain James Fryer of Liverpool was instructed in January 1790 to proceed to the Isles de Los, Sierra Leone, and the Banana Islands but 'not to trust any Goods to the Natives on forfeiture of your Commission & priviledges [sic] hereafter mention'd except you get pawns'.70 Merchants residing on the coast, in turn, used pawns to protect credit given in trade inland. For example, in 1757 Nicholas Owen, who lived at Sherbro, was informed by his brother, Blayney, then upriver trading, that he had 'rece'd by the man Fonga 4 brass pans and 2 half pieces of baft'. He also reported buying 'one pawn from Bbong, so that I am quite out of sortments on account of lending money to Bereyboso for 2 slaves'.71 Pawning no doubt underpinned the substantial expansion in slave trading that occurred in upper Guinea after I750. RATIOS AND NUMBERS OF PAWNS The number of pawns deposited by slave suppliers with shipmasters appears to have been subject to individual negotiation, the outcome of which depended, among other things, on trading conditions at the time of negotiation and the skills and probably the status of those involved. Robert Hume testified before Parliament in 1789 that he had been involved in the slave trade for 'near eleven' years and related one instance where the wife of a local king had been held as pledge for the delivery of 80 slaves at Cape Mount.72 Including relatives of traders among pawns was seen by masters of ships as improving the likelihood of contracts being fulfilled, though given asymmetries in knowledge among negotiators about the relationship of pawns to traders, some deception by local merchants in negotiating such matters was possible.73 Edwards, who visited the African coast in I752-5, 68 Pawnshipis not mentionedby David Hancock in his otherwiseexcellent study of Britishslavingon the SierraLeone coast, and specificallyat Bance Island; see Citizensof the World: LondonMerchantsand the Integrationof the British Atlantic Community, I735-I785 69 (Cambridge, I995), 172-220. John Matthews, A Voyage to the River Sierra Leone (London, I788), 155-6. 70 Robert Bostock to James Fryer, January 1790, Bostock Papers, Liverpool Record Office, 387 MD 55. Also see references to transactions at Junk in 1783; Fraser's testimony in Lambert, Sessional Papers, LXXII, 621-2. 71 Nicholas Owen, Journal of a Slave Trader (London, 1930), 58-9. 72 Testimony of RobertHume, in Minutes of Evidencetaken on the Second Reading of the Bill entitled, 'An Act to prohibit the Trading for Slaves on the Coast of Africa, within certain Limits', in Torrington, Lords Sessional Papers, III, 99, I2. 73 Pawns were often related to merchants. In 1762, for example, it appears that Archibong Robin John of Old Calabar was a pawn on the Amelia of Bristol; see the testimony of James Morley, in Lambert, Sessional Papers, LXXIII, 154. For another example, see the account of James Berry, 3 Apr. 1763, in Williams, Liverpool Privateers, 534. For pawns held by Lace in 1776, see the deposition of King George [Robin John Ephraim], Jno. Robin John, Otto Ephraim [Robin John Otto Ephraim], and Orrock PAUL 82 E. LOVEJOY AND DAVID RICHARDSON understood that pawns sometimes became slaves, but did not think that 'Europeans, to whom Children are so pawned, are guilty of Fraud and Abuses in this Respect, and sell them before the Time for which they are pawned is expired'. Edwards claimed that 'He never knew an Instance of their being so'.74 The 1788 crisis at Bimbia suggests otherwise, however, as the king and traders of Cameroons protested that one British trader had carried off 'thirty of the King and Traders Sons and Daughters and Relations, which were put on board the said Ship as Pawns'. They claimed that 'the greatest Part of the Slaves and Ivory was paid for their Release' but that the offending captain 'absolutely refused their Release when redeemed'.75 Problems such as these no doubt encouraged some African merchants not to allow their kin to be held as pawns, as a Liverpool brig discovered at Old Calabar in 1763, when Robin John Town refused 'a son for pledge'.76 Other things being equal, European slavers tried to obtain a high ratio of pawns to slaves. In April 1788, Richard Rogers reported trying at Old Calabar to achieve a ratio of two pawns for every three slaves to be delivered and was not alone in expecting such a ratio.77 Excepting cases of close relationships to important local merchants or officials, ship captains naturally tried to maximize the ratio of pawns to slaves, thereby increasing their leverage over local dealers. The latter equally preferred a low ratio. If possible, they preferred to dispense with pawns altogether since their retention by traders represented a loss of labour and the costs of their maintenance seems to have fallen on the debtor. Such costs provided powerful incentives to redeem pawns on schedule or even earlier.78 There are also signs that, on the Gold Coast at least, self-pawning was common, though only with individuals who were regularly employed by European firms and thus had prospects of regular earnings. Such earnings were effectively used as collateral, as Kea has shown, and because canoe men had little else to pledge, they were often 'obliged to place themselves in pawn'.79 One such Robin John, in ibid. 541. In 1787 King Ambo was required to 'give one of his sons to [Captain]Tatam' of Liverpoolin place of a pawn who had absconded;see AnteraDuke, 'Diary', 59. 74 Testimony of Richard Edwards, in Lambert, Sessional Papers, LXIX, 44. 75 Letter of William Quarrier and James Bean, 27 Feb. 1788, in Lambert, Sessional Papers, LXIX, 317. James Berry, 3 Apr. 1763, in Williams, Liverpool Privateers, 533. PRO C I07/I2, Richard Rogers to James Rogers, Apr. 1788. James Arnold, in evidence to Parliament in 1790, reported that the master of the Bristol ship, Ruby, which traded at Bimbia in Cameroons in 1787-88 had released 1 slaves and some ivory to three local merchants in order to allow them to redeem pawns from another ship, the master of which was threatening to sail away with pawns. Arnold reported that the 1 slaves secured the release of 'Six or Seven' of the merchants' pawns (Lambert, Sessional Papers, 78 Antera Duke, 'Diary', 35. LXXII, 52). 79 Hence on 21 Dec. i668, Ando Wassa of Amanfro was advanced goods, using his monthly allowance owed to him by the fort as security. See Kea, Seventeenth-Century Gold Coast, 229. In Oct. i68i, three canoe men of Soko, or 'English' Accra, were listed as pawns of the Royal African Company for the sum of I,044 dambas: Agya ('Agga') for 288 dambas, Okyere ('Aquerry') for 576 dambas, and Asa ('Asha') for I80 dambas (ibid. 243). Also see letter of John Thorne, Offra (Kingdom of Allada), 4 Dec. 1681, in Law, 76 77 English in West Africa, 223. THE BUSINESS OF SLAVING 83 case occurred in 1704 when Cawera, a canoe man of Cape Coast, 'Engaged himself to ye Company for a Pawn' on account of three ounces of Gold which was paid for him to the king of Seboo [Asebu] to whom he was indebted so much'.80 We have no evidence that self-pawning by Africans to other traders was common, but in resorting to such practices canoe men perhaps hoped to invest in trade and rise to become independent traders in their own right. TIME LIMITATIONS ON DEBTS Although Europeans adopted indigenous systems of pawnship to underpin coastal credit relationships in the slave trade, there was one crucial distinction in the way in which Europeans sought to enforce debt repayment. This was reliance on specified time limits on credit tied to movements of slave ships. Edwards claimed that at the Gold Coast time was a factor in pawning contracts, noting that people 'are usually pawned for a certain Number of Months'.8" Referring to Sierra Leone, Matthews in 1788 also assumed that a time limit applied in pawning arrangements, noting that pawns were 'liable to be sent off, if not redeemed in due time'. He reinforced the point by arguing that 'a person, whether a slave or the son of a freeman, if not redeemed at the expiration of the time limited for his redemption, becomes so much the absolute property of the person to whom he was pawned, that... it is intirely [sic] at the option of his master whether he will ever after let him be redeemed, though they should offer twenty for one, or he should be a son of the most powerful person in the country'.82 Whatever the rights of creditors over pawns were, because of the potential damage to future commercial dealings that might arise from their enslavement, ship captains were often reluctant to carry them away, and if they were taken off, they were not always sold but occasionally returned to West Africa.83 Because of the complications arising from holding pawns, some captains in fact transferred pawns to other ships in exchange for slaves in order to avoid sailing with them and thereby facing the difficult decision of selling them or not in the Americas. In 1788, Richard Rogers had 60 pawns on board his ship, most of whom had been 'redeem'd from other ships'.84 Similarly, James Fraser noted in 1790 a practice existed whereby 'the owner or friend' of a pawn 'commonly borrows Slaves from a ship which is not so far forward in completing her cargo, and takes the pawn out of the ship that is ready to 80 The transaction was dated 19 June 1704. Kea (Seventeenth-Century Gold Coast, 243). According to Kea's calcuation, 3 oz. of gold were equal to 1,052 dambas, and a fat sheep cost between 120-150 dambasin gold, so that the total amount of indebtedness was dambas. Similarly, Abo ('Abboo'), a Cape Coast canoe man, pawned himself 1,172-1,202 in Sept. 1704 for 960 dambas of perpetuanas (textiles), while in Nov. 1704, Cape Coast canoe man, Kwawu ('Quow') placed himself in pawn for 388 dambas, although the advanced goods were not specified (ibid. 243). 81 Testimony of Richard Edwards, in Lambert, Sessional Papers, LXIX, 44. 82 Matthews, Voyage to Sierra Leone, I55-6. 83 Thus the African Queen, trading from Old Calabar to Jamaica in 1793, had on board 'a Negro Boy of about i6[.] He was put on Board on the Coast as a Pawn and we thought it advisable not to sell him' (John Cunningham and John Perry to James Rogers, o Mar. 1793, 84 PRO C 107/59. PRO C 107/12, Apr. 1788, James Rogers to Richard Rogers. 84 PAUL E. LOVEJOY AND DAVID RICHARDSON depart, and puts it [sic] on board the other'.85 Whether transfers of pawns between ships needed permission from local merchants is unclear, although in local practice, as far as known, such alterations in arrangements required the approval of close kin. 'Should they not pay in...time', the master of a ship was nevertheless justified, according to one trader, in carrying away pawns 'to sell in the West Indies'.86 There are indications that some masters followed this advice. In 1773, Robin John Ephraim of Old Calabar observed that one ship had left with his pawns, including four of his sons.87 Fifteen years later, the Gascoigne of Liverpool left the same port with 540 slaves and pawns, the latter amounting to ' 20 Pledges'. Even then, it was said, the debts left behind at Old Calabar represented a further hundred.88 Enslavement of pawns on board ships effectively brought pawnship and panyarring into line, essentially completing a contractually based process of debt-enforcement that rested on the threat of enslavement.89 The process assumed, of course, that when giving pawns as collateral for credit African merchants themselves recognized they were engaging in a time-limited contract. This seems to have been acknowledged by at least one Old Calabar merchant who, in a deposition witnessed by three others in August 1776, recognized his failure to redeem a boy being held as a pawn and thus the legitimacy of the boy's enslavement.90 Others complained about abuse of contracts by Europeans, but the fact that pawns, including kin, could be and indeed were carried away provided African merchants with a powerful incentive to fulfil their contracts. As a result, human pawnship became one of the most efficient or cost effective mechanisms available to European merchants to protect credit in the slave trade. Whether or not time-limitations on the repayment of debts in Africa was an innovation of the export slave trade, enslavement of pawns after a specified repayment period had expired was more likely to be imposed in situations involving ship-based trade, with consequences that were difficult to reverse. In some cases, people were enslaved if debts were not paid within a certain time and in others people were panyarred to recover debts, but when ships set sail with pawns and hostages, it proved difficult to redeem them. The prompt delivery of contracted amounts of slaves, gold, logwood or other goods for which merchandise had been advanced was ultimately the surest way of preventing such acts of enslavement. 85 Fraser'stestimony in Lambert,SessionalPapers,LXXI, 15. PRO C 107/12, 20 July 1788, RichardRogers to James Rogers. The idea that one might 'borrow'pawnsimpliesthat, even afterbeing shippedfromthe coast, pawnsmight still be redeemed,at least in theory. 86 87 Letter to Thomas Jones, Old Calabar, i Nov. 1773, in Williams, Liverpool Privateers, 543-5. 88 PRO C 107/12, Richard Rogers to James Rogers, Old Calabar, Apr. 1788 [exact date not recorded]. 89 As Clarksonlearned to his chagrin, a prominent merchant, whom he would not name, was involved in a case that was 'an horrid instanceof cruelty, practicedonly last year [1787] by an English captain,on the body of an innocent pledge, whose father had not returnedin time', and hence the boy was taken as a slave across the Atlantic. See Slavery and Commerce, 27. 90 As cited in Williams, Liverpool Privateers, 541. THE BUSINESS RETRIBUTION OF SLAVING AND RECOURSE IN PAWNING 85 CASES Although it was accepted that unredeemed pawns could legitimately be carried away to America, not all shipments of pawns were seen as legitimate. Pawns were usually entrusted to ship captains pending redemption, and their security depended on captains meeting the terms of contracts. In some cases this proved difficult. In the early I770s, for example, the seizure of a Liverpool ship on the Windward Coast by a French naval vessel created a danger that the pawns on board would be shipped to America as slaves. When told that this indeed was likely to be the case, James Fraser, master of a nearby Bristol ship, reportedly 'remonstrated against the impropriety and injustice of it', and threatened the French officers with legal recourse: I assured them that the Chamber of Commerce in France, and the African Company in England, would prosecute him and his Officers for so doing, unless he gave an opportunity to the friends of the pawns to redeem them, by paying the Rice or Slaves for which they were pawned.91 The threat to seek legal intervention in France may not seem to be very credible, but the fact that such recourse was even considered suggests that the practice of pawning was sometimes perceived in Atlantic-wide terms.92 In many cases, captains themselves appear deliberately to have broken contracts, carrying away pawns without allowing appropriate time for their redemption. In some instances, kidnapping of 'free' Africans compounded the situation. African merchants sometimes responded to such misdemeanours by seizing or panyarring sailors to try to force the return of pawns or freemen,93 but as the crisis at Bimbia in I788 shows, retribution could be exacted against ships and crews unconnected with the original offence. On this occasion, a Dutch ship successfully chased the British slaver that had illegally taken away the pawns and thereby secured the release from captivity of the crews of the other British ships seized by Cameroons merchants in response to the offence.94 On other occasions, such events 91 Fraser'stestimony in Lambert,SessionalPapers,LXXI, 15. In anothercase, the Bristolship, Swift, was seized by a Frenchprivateer,La Liberte, in 1793, and some of the seized goods were taken on board the French ship, with the remainderransomed,includingsome slaves, for I1,ooo.The paymentwas made in the form of a draft, drawnupon the accountof the Swift's captain,William Roper, 'giving 92 his mate as a Hostage' (see Phil Lycott to John Anderson, 13 June I793, PRO C 107/59). The incident took place between Cape Palmas and Rio Pongo in May 1793 and came before the AttorneyGeneralwho, on 2 July I793, declaredthe contractunlawfulon the groundsthat ransomingcapturedvessels was prohibitedby Britishlaw. Incidentally,the mate concernedwas James Humphreys,who was allowed under the contract full pay duringhis captivityand 3 shillingsper day while held in prisonin Franceuntil his return to Britain. We do not know what became of him. The case leaves open the issue of pawnshipunderBritishlaw becausethe judge did not passjudgementon it and therefore there is still room to doubt whether the threat was credible, but also underlines the importanceof legal sanctionof contracts. 93 Grandy King George [Robin John Ephraim] to Ambrose Lace, 13 Jan. 1773, in Williams, Liverpool Privateers, 543. 94 See the testimonyof JamesArnold before the Lords of Trade on the Slave Trade, I789; letter of William Quarrier and James Bean, Quans Town, Cameroons, 27 Feb. 1788; and letter of JamesMcGauty,WilliamWilloughbyand IsaacNixen, on boardthe Othello, River Cameroons, I Mar. 1788 in Lambert, SessionalPapers, LXIX, 50-55, 86 PAUL E. LOVEJOY AND DAVID RICHARDSON triggered outright attacks on ships and their crews that resulted in ships being 'cut off' at the coast. Illegal taking of pawns was usually met by often of a collective summary and violent retribution, nature, and not the it was one reason that necessarily against offending ship.95 Consequently, some captains at least were reluctant to sell pawns or otherwise take them off the coast. Illegal confiscation was not, however, the only source of disputes between Europeans and Africans relating to pawns. For example, pawns sometimes absconded before time and had to be replaced. A common feature of to such acts was a sense of collective responsibility on the part of responses African merchants. In many areas where pawnship existed, secret societies of adult males, dominated and most powerful in their by the wealthiest community, regulated credit, drawing on religion and the power of secret councils to do so. One function of the Poro society, an association of males in Sierra Leone that dates from the sixteenth century, was to enforce debt repayment as well as to protect individuals pawned as surety for credit.96 According to Matthews, who visited Sierra Leone in 1785-7, were a pawn 'sent off' before his time for redemption had expired, or even after, 'without giving notice to the person who pawned him, a palavar, or action, would be brought against the person offending'.97 At Old Calabar, the Ekpe society fulfilled a similar role.98 In his diary, Antera Duke describes occasions when 'Grand Egbo [Ekpe]' was collectively 'blown' against ships 317-I8. For a discussion of the political implications of the crisis, see Ralph A. Austen and Jonathan Derrick, Middlemen of the Cameroons Rivers: The Duala and their Hinterland, c. I6oo-c. I9oo (Cambridge, I999), 38-4I. 95 There are instances where pawns or dependents of debtors who were carried away from the coast and reached America were pursued by both European and African merchants, and in some cases were returned to Africa. John Olderman, engaged in the African trade for over twenty years, making thirteen voyages to the upper Guinea coast and Sierra Leone before I797, claimed that he had held pawns in 'few Instances only' but recounted the story of a 'free boy' who had been to sold as a slave. According to Olderman's testimony, 'He was carried to the West Indies, and brought to London to be restored to his Father by the next Ship in the Employ that went out; he was put on board a ship called The Lively, in April 1794; the Ship was taken by the Sanspareil, a French Ship of Eighty-four Guns; that Ship was afterwards captured by our Fleet on the First Day of June of that Year; he was taken on board the Majestic, and from that Period he has been on board of different Ships, especially the Blanche. I have traced him upon the Navy Books, and the last that I heard of him, he was discharged from the Blanche at Portsmouth' (Evidence of John Olderman, in Torrington, Lords Sessional Papers, III, 148,15 I). Whether the boy ever made it back to Africa is unknown. One of the most widely publicized cases of pawns being identified in the Americas and then returned to Africa involved two boys from Old Calabar, who wete taken to Barbados, Maryland and then Britain, where they were the subject of a court case concerning their freedom in I773 (see Ruth Paley, 'After Somerset: Mansfield, slavery and the law of England 1772-1830', in Norma Landau and Donna Andrews (eds.), Crime, Law and Society (Cambridge, forthcoming). 96 For the introduction of Poro into the region, see Brooks, Landlords and Strangers, 304-6; Rodney, Upper Guinea Coast, 65-9; F. W. Butt-Thompson, West African Secret Societies (London, I929); Kenneth Little, 'The political function of the Poro', Africa, 35 (I965), 98 350-64; 36 (1966), 62-72. 97 Matthews, Voyage to Sierra Leone, 155-6. Lovejoy and Richardson, 'Old Calabar slave trade', 333-55, and the literature discussed therein. THE BUSINESS OF SLAVING 87 close to completing their trade and perhaps contemplating taking away pawns still held on board.99 Sometimes, Ekpe was 'blown' to force an Old Calabar merchant to replace a recalcitrant pawn, perhaps in an attempt to deter other ships from leaving with pawns and thereby protecting trade as well as those in pawn. In these instances, Ekpe functioned as a 'merchant guild', co-ordinating local action to monitor the treatment of pawns and to safeguard the interests of the leading merchants of Old Calabar who used them as collateral. On the Gold Coast, according to Richard Miles writing in the I770s, pawns were only 'condemned to sale for debt' after due procedure. Miles also noted that in most towns there was a 'Palaver House, which strictly speaking, is a court of justice', where 'the Judges or Elders of the town (few of which are under the age of sixty or seventy) assemble to hear the parties'. The 'trials', noted Miles, 'are in open day, and free for any person to hear that pleases'.00? As elsewhere, therefore, on the Gold Coast pawnship and credit arrangements became institutionalized, regulated by 'palavar houses' that undertook in relation to debt enforcement and protection of pawns functions similar to the secret societies of Upper Guinea and the Bight of Biafra. CONCLUSION Pawning arrangements shifted some of the risk associated with credit advances by Europeans to African merchants because, in the event of default, unredeemed pawns could be shipped to America for sale. Occasionally Europeans were held as pawns, but generally credit flowed from ship to shore and pawns were held in European custody, whether on shore or on board ship. Pawnship gave shipmasters some insurance against default by slave suppliers. Credit arrangements on the African coast were thus consistent with economizing on transaction costs, but placed at risk the kin of local traders. Hence, in places where pawns were used as collateral in trade, transactions became linked to social relations. In this respect, the study of pawnship has implications for understanding the cultural, political and economic interface between different systems of valuation and exchange in the Atlantic world.101 'Commercial' pawning involving European ships was in principal governed by a more precise time schedule than its domestic equivalent. By 99 In 1787, one British ship threatened to sail with pawns because slaves had not been delivered. Accordingly, Ekpe intervened. Antera Duke reports that the Ekpe council invoked a decree expressly forbidding 'any Captain to send any Calabar pawn, which was given for my slave goods, away in his tender'. Also see Kannan K. Nair, Politics and (London, 1972), 15. Society in South-Eastern Nigeria, 184I-I906 100 Testimony of Richard Miles, in Lambert, Sessional Papers, LXVIII, 41. 101 See, for example, Jane I. Guyer (ed.), Money Matters: Instability, Values and Social Payments in the Modern History of West African Communities (London, 1995), although in her introduction on the currency 'interface' and its dynamics, Guyer does not discuss pawnship in her 'people's view' of the economies of the precolonial period (ibid. i-Io, I4). Joseph Miller (Way of Death, conclusion) considers that there were different systems of valuation, at least in west-central Africa, ones that were 'European' or 'African'. Miller is interested in how these meshed in the abstract world of market behaviour and economic rationality, but his discussion of pawns is confined to the victims of the slave trade and not the ways in which credit flowed through the use of pawns. 88 PAUL E. LOVEJOY AND DAVID RICHARDSON most accounts the latter might continue indefinitely, even across generations, though it was often terminated through marriage in the case of women, a feature of pawning in domestic economies that has been long recognized.102 Sometimes in such situations pawns became slaves, the justification for such action usually relating to foreclosure for debt. The use of pawns as commercial hostages in the export trade with Europeans further blurred distinctions between slaves and pawns. Time limitations on redemption may well have represented a significant modification of domestic pawnship. The ability to foreclose on debts by sailing away helped to reshape mechanisms to enforce contracts, suggesting parallels with the time restrictions in enforcement practices associated with panyarring. The number of 'pledges' or pawns deposited as security for credit was subject to negotiation and was implicitly intended to prevent the arbitrary seizure and sale of people through panyarring. Hence there had to be some understanding of what protection, if any, pawns might expect, and perhaps most importantly of all, under what circumstances Europeans might enforce contracts by sailing away with pawns, so that foreclosure was not confused with panyarring. Pawning was not universal in the slave trade and varied in importance in the local societies and economies of western Africa. Its use in Atlantic slaving seems to have increased as the export slave trade reached its peak in the late eighteenth century.103 As a system of credit security, pawnship appears to have become vital to the growth of the Atlantic slave trade, underpinning the business operations of many European merchants and their African suppliers in several parts of Africa. Some transferring of pawns between European ships occurred, but pawnship essentially concerned Afro-European relations, not relations between Europeans stationed at the coast or even relations between European and mulatto traders. This restricted use of pawns by Europeans may explain why Hancock failed to discover their use at the London-owned factory at Bance Island, Sierra Leone, even though pawns were commonly used nearby.l04 Variations in the incidence of pawning, geographically and temporally, and the presence of panyarring, appear to have been related to political See the variousstudies in Falola and Lovejoy, Pawnshipin Africa. David Eltis suggests that pawning was not that important in trade between Europeanand African merchantson the Gold Coast in the late seventeenth century, suggestingthatpawningbecamemorecommonin the eighteenthcenturyon variousparts of the Africancoast; see The Riseof AfricanSlavery in the Americas(Cambridge,I999), 102 103 155, I8o, I87. 104 Hancock,Citizensof the World,172-220, but see, for example,in I790, a Lancaster ship trading at Cape Mesurado trying 'to get pawns redeem'd & Collect in Debts'; see RichardMartinto JamesRogersand Co., 9 September1790 (Rogerspapers,PRO C 107/I2). Similarly, when the Sierra Leone Company established its headquarters on land granted by 'King' Naimbana,the regent of the Koya Temne, in 1787, the land was purchasedfor 'a paltryconsiderationof about thirty pounds; and for the good faith and true performanceof the contract,the King said he would pledge his second son, John Frederic,whom [Alexander]Falconbridgemight take with him to England'; see Anna MariaFalconbridge,Narrativeof Two Voyagesto theRiverSierraLeoneduringthe Years I79I-I793 (London, 2nd ed., I802), 60. Falconbridge agreed to take the boy to be educatedbut not as a hostage,althoughif this distinctionwas recognizedis not clear. On the position of Naimbana,see ChristopherFyfe, A History of Sierra Leone (London, I962), 19-20. THE BUSINESS OF SLAVING 89 in Africa. Sometimes within which credit worked structures arrangements such structures seem to have provided security for credit without resort to pawning, even though pawnship perhaps existed in local economies supplying the export trade. This was apparently the case at Whydah after its conquest by Dahomey in I727. Similarly, pawnship does not seem to have been a factor in the rise of Bonny as the leading slaving port in the Bight of Biafra, though it appears to have been fundamental at other ports in the region. Furthermore, pawnship arrangements involving Europeans are not reported in areas where Muslim merchants were dominant, such as Senegambia, nor, with the possible exception of Benguela, did they feature in the export trade from Portuguese-controlled areas south of Kongo. Pawnship was important, however, in areas where secret societies and agegraded 'palavar' houses were involved in debt enforcement. In these areas, too, panyarring was to be found. The restricted use of pawns in relations between African and European merchants to certain places on the African coast modifies Rodney's focus on the increasing incidence of social oppression in the context of the slave trade. According to Rodney, 'many of the forms of slavery and subjection present in Africa in the nineteenth and twentieth centuries and considered indigenous to that continent were in reality engendered by the Atlantic slave-trade'.105 Our analysis suggests that pawning at least was indigenous to many, if not most, parts of western Africa but was adapted as an institution underpinning credit in trans-Atlantic trade, apparently in the seventeenth century on the Gold Coast, and elsewhere by the eighteenth century.106 It also appears that this process of historical adaptation depended on local conditions as much as on Atlantic influences. Without mechanisms such as pawnship to secure credit, the number of slaves purchased by Europeans in Western Africa would have been considerably less.'07 The relationship between 'capitalism and slavery', as articulated by Williams, has to be modified to allow for the complex ways in which credit penetrated Africa, thereby enabling trade to take place across different cultural interfaces.'08 Reliance on pawnship in Western Africa highlights the flexibility of European capital in adjusting to local conditions by exploiting servile relationships other than slavery, but subordinating those relationships to the requirements of the export trade in slaves. The interface between Europe and Africa resulted in institutional adjustments on both sides of the cultural frontier. As in other areas of Atlantic history, the intensified use of pawns in commercial relations between European and African merchants in the eighteenth century corresponded to key periods of development of slavery in the Americas; pawning helped therefore to underpin the expansion of the Atlantic economy. 105 Rodney, 'Social Oppression', 443. 106Also see Robin Law, The Slave Coast of West Africa, I550-I750: Atlantic Slave Trade on an African Society (Oxford, 1991), 68-9. The Impact of the 107 For an elaborationof this point, see Lovejoy and Richardson,'Trust, pawnship, 108Williams, Capitalismand Slavery. and Atlantichistory', 333-6.