Free People, Slaves and Pawns in the Western Gold Coast: The Demography of Dependency in a Mid-Nineteenth-Century British Archival Source

In 1849, the Acting Governor and Judicial Assessor of the British settlements on the Gold Coast, James Coleman Fitzpatrick,¹ forwarded to the Colonial Office two tables providing an estimate, or rough census, of the population of Dixcove and Appolonia, two polities in the western Gold Coast, which, at the time, were under some form of British influence. The tables included details about territorial subdivisions, men, women, children and slaves and pawns (listed under the same headings). The document can be found in the Colonial Office papers, Original correspondence, in the National Archives of the United Kingdom.² Considering the period and the area it deals with, the document is indeed a very rare, source both in terms of historical demography and the history of personal dependency and slavery in this section of the Gold Coast.

The table concerning Dixcove was quoted extensively in an article published in 1956 by Henry Swanzy in the “Transactions of the Gold Coast and Togoland Historical Society”. Swanzy was writing about his family’s history and the citation of the data concerning Dixcove’s population served his own purpose of discussing the crucial role played by his ancestor Francis (Frank) Swanzy³ in that town (Swanzy 1956).

The background

Fitzpatrick’s tables cover the two sections of the Windward Coast (west of the British post at Sekondi), which had an alliance with the British, and over which the British claimed a form of “jurisdiction,” though it would be an exaggeration to describe it as coherent political control. Fitzpatrick’s attempt to take a census of Dixcove, and to make at least a rough estimate of the population of Appolonia, was consistent with the proposal he articulated in his accompanying dispatch to the Colonial Office, i.e., to restructure the system of revenue raising for the Administration of the settlements by doing away with the cumbersome and inefficient system of customs duties (highly unpopular especially with the merchants), and instituting a poll tax levy.⁴

Dixcove, built around a natural harbor on the coast of Ahanta, was the main British station in the western Gold Coast, during this period. The town possessed a substantial fort that the English built in the 1690s. Political authority from Dixcove extended only a short distance along the coast, but covered much more substantial territory inland. The town was an autonomous polity, though its territory was entirely surrounded by the so-called Dutch Ahanta, a district over which the Dutch had established a sort of protectorate following a war against the king of Busua in 1837-38.⁵

Appolonia was an extensive polity running from the Ankobra River to the Tano River and Lagoons (in the areas currently crossed by the Ghana-Côte d’Ivoire frontier). In 1765, authorities in Appolonia had established an alliance with the British, who built a fort in the town of Beyin. However, from the fort at Cape Coast, prior to 1821 when they abandoned it, the British did not gain a position from which to exercise effective control over local rulers. King Kaku Aka (ca. 1832-1848) successfully acquired commercial and military power in the region. He exercised internal rule in full independence and maintained relations with major powers in the region: Asante, the Dutch, and the French, among others. In 1835, George Maclean tried to rein in Kaku Aka by entering his domains with a show of military power to force him to sign a treaty of protection. The British, however, could not sustain occupation of Appolonia, so Maclean’s tactics produced no lasting results with Kaku Aka, who continued to operate independent of British oversight and influence.

By 1847, Kaku Aka’s aggressiveness had alienated all of his African and European neighbours, who openly demanded that British officials take action against him. In March and April 1848, Lt.-Governor William Winniett led a punitive expedition against Kaku Aka that included a coalition of forces from a number of polities in the Gold Coast. Kaku Aka was captured thanks to the defection of his main subordinates who surrendered him to the invaders. Coalition leaders tried him and imprisoned him in Cape Coast Castle. He died there on December 28, 1851.⁶ Despite their military success, the British did not permanently occupy Appolonia. After the capture of the king, leaders in Appolonia did not choose a successor to occupy the royal stool. The British mediated a power-sharing agreement among the main generals (safohyenle) of the deposed king,⁷ a policy that spelled the end of the strongest and most centralized polity of the littoral west of Elmina.

For the societies of the Gold Coast, the first half of the nineteenth century had been a time of dramatic economic and social transformation. The abolition and gradual suppression of the transatlantic slave trade⁸ resulted in important changes in the domestic labour market. The development of cash crop agriculture, boosted by the growth of “legitimate commerce” was just one aspect of a more general expansion of farming for local markets: primarily urban areas and ports. The result was an increased internal demand for slaves and generally bonded labour for employment in agriculture (Lovejoy and Richardson 1995: 32-56; Kea 1995: 119-43; Law 1995: 6-9). In the eighteenth century, the ruling military aristocracies exercised more or less total, direct control over the recruitment of bonded labour (through war, the justice system, and debt), its exploitation (in the military and on rulers’ estates, for example), and its exportation (as merchants in the transatlantic Slave Trade).

During the first half of the nineteenth century, the socio-demographic characteristics of slave owners markedly changed, as the expansion of agricultural production and commerce increased the purchasing power of social groups other than the military aristocracy, making bonded labour a highly requested – and to some extent affordable – investment for broader sections of society (Manning, 1985: 853; Adu-Boahen, 2010: 124-5). The internal market for slavery expanded, while the extension of credit mechanisms (linked also to the expansion of petty trade) led to a growth in human pawns. As far as the British residents were concerned, the laws designed to limit and discourage slave-holding by British subjects who resided on the coast were weak and ineffective.⁹ Virtually all British subjects in the country had bonded labourers under their control, even, to some extent, the Wesleyan missionaries (Metcalfe 1962: 311-13). Generally, British residents circumvented anti-slavery laws by entrusting their dependents of slave status to their African or Euro-African women and relations (Adu-Boahen, 2010: 128), and keeping direct control over individuals held in pawnship or other forms of bondage.

The official position against slavery and other forms of bonded status was reflected in the language Europeans employed with regard to these different types of personal dependency. On one hand, British officials and merchants generally avoided the term “slave” when referring to a Briton’s bonded dependents, but reserved it for bonded people held by Africans, Euro-Africans or Europeans of nationalities other than British.¹⁰ On the other hand, European and Euro-African residents tended to include slavery and pawnship within the same category of dependency. This was a product of two opposing historical trends: the weakening of fundamental distinctions between the two categories in the heydays of the slave trade,¹¹ and a diluted European perception of the typical features of slavery following abolition and anti-slavery legislation. In the eighteenth century those fundamental features were largely shared by Europeans and Africans, including aspects of severe constriction and violence, which were intrinsic to forms of bonded status. In the nineteenth century post-abolition period, British residents who were still largely benefitting from control over bonded people tended to defuse accusations of slave-holding by portraying these forms of subjection as patently distinct from slavery. At the same time, some prominent European residents reinterpreted features of Gold Coast slavery--in which they were accused of taking part--by ambiguously extrapolating elements of integration and paternalism. Accordingly, they described local slavery--unlike American plantation slavery-- as fundamentally benign and inclusive, except for aspects of cruelty and abuse, which were imputed to the primitive state of African society in general, and not to the institution of slavery in itself. It was the formulation of a lasting Gold Coast “Domestic Slavery” paradigm.

“African domestic slavery” as an anthropological model is a twentieth-century invention, but it was characterized by a representation of relationships of dependency that had a much longer history, and a genealogy that is recognisable to those familiar with the writings of nineteenth-century Europeans and Euro-Africans who were fully integrated into the coastal society of the Gold Coast.¹² Brodie Cruickshank,¹³ a Briton who resided in the Gold Coast, provided the most authoritative early formulation of the “Domestic slavery” paradigm (indeed he made specific reference to the Fante region):

The condition of the slaves in the countries under our protection is by no means one of unmitigated hardship. In ordinary cases, the slave is considered as a member of his master’s family, and often succeeds to his property, in default of a natural heir. He eats with him from the same dish and has an equal share in all his simple enjoyments. He intermarries with his children, and is allowed to acquire property of his own, over which, unless under very extraordinary circumstances, his master exercises no control. He sometimes even acquires wealth and consideration far superior to his master, who may occasionally be seen swelling his importance by following in his train. They address each other as ‘my father’ and ‘my son’, and differ in little in their relations from the respect and obedience implied in these endearing epithets (Cruickshank 1853, 2: 240).

Cruickshank was a seasoned resident in the country and part of a large network of local interests, alliances, and clientele. He had slaves and individuals held as security, in addition to accoutrements that accentuated a characteristic paternalistic idyll. One must not however dismiss his somewhat idyllic—and highly questionable—description of Gold Coast slavery as mere self-justification. Cruickshank’s analysis of the relationship between slave and master reflects a theme that is well represented in accounts of local society. It illustrates a key element in the formulation of the historical discourse on West Africa where the language of kinship and alliance operated as a fundamental tool for inclusion and integration.

These dynamics counsel careful consideration of the context in which the two tables were allegedly produced. J. C. Fitzpatrick was the acting governor of the Gold Coast during Lt.-Governor W. Winniett’s sick-leave in Britain from January 1849 to February 1850. In April 1849 he visited the British fort located in the town of Dixcove.. For many years the fort had been under the supervision of Frank Swanzy,¹⁴ whom the Colonial Office, against the advice of Cruickshank and other influential Gold Coast administrators, had preferred over Fitzpatrick for Judicial Assessor in 1844. The acting governor made unfavorable remarks to the Colonial Office about the state of British administration in Dixcove, pointing out abuses in the exercise of justice and creeping discontent among the local population (Swanzy 1962: 104). He implicitly blamed Frank Swanzy for the problems. From Dixcove, Fitzpatrick moved to Appolonia where he very effectively obtained the payment of part of the war reparations and fines that the British had imposed on the chiefs of Appolonia following the 1848 expedition. In the process, the acting governor repeatedly exercised severe physical constrictions on ëbanyenle, the main British ally in Appolonia and a close acquaintance of Swanzy, in order to induce him to hand over large quantities of gold dust. ëbanyenle,¹⁵ installed in 1848 as chief of the former royal capital of Atuabo, was tied up, beaten and severely humiliated in front of his subjects. Swanzy responded with a complaint to the Colonial Office in which he denounced the acting governor’s abuses. However, the issue between Fitzpatrick and ëbanyenle worsened during the acting governor’s visit to Atuabo to collect the remaining reparation money. The people of the town mounted an attack for which Fitzpatrick ordered the town set ablaze and Ɛbanyenle and four other chiefs arrested, taken to Cape Coast, lashed in public and condemned to forced labour. A serious case against Fitzpatrick was made in which Frank Swanzy was the main plaintiff. The two men had become sworn enemies.¹⁶

Fitzpatrick’s animosity towards Swanzy is enshrined in his population table for Dixcove. The town and its territory are portrayed as a concentration of social relationships rooted in slavery, affecting a massive proportion of the population. Without any open declaration, Fitzpatrick points to Swanzy as the real mastermind of this system of dire subjection.

The state of documentary evidence does not detail how Fitzpatrick gathered his population figures about Dixcove and Appolonia. In the letter accompanying the statistical returns, he recognizes that his figures “are not of course strictly accurate,” adding immediately “but I apprehend they are very nearly so as I took great pains in consulting persons of the greatest knowledge and experience in the two localities.”¹⁷ Regrettably, he does not provide the names of his sources, fuelling the reader’s doubt as to the reliability of the figures and urging the exercise of critical prudence in using them. Despite this limitation and the inherent prejudice against Swanzy evident in the data, these data do offer some unique insights into the basic relationships of dependency in these two important sections of the Western Gold Coast.

Demography and personal dependency

The two tables list the main subdivisions: in the case of Dixcove, these are the quarters of the town along with their subordinate settlements in the countryside (the number of dependent villages – 32 – is provided in detail); in the case of Appolonia, it is the two royal residencies, Atuabo and Beyin, and the head-towns of the main territorial partitions of the country. Both tables show the name of the chief in charge of the subdivision, and the number of “houses” of each subdivision.¹⁸ The tables also include the number of men capable of bearing arms, the number of women of mature age, of children, of “pawns” and “slaves” gathered under the same heading, the number of free people, and finally the total number of inhabitants.

There is a noticeable difference in the degree of accuracy between the figures reported for Dixcove and those for Appolonia. The data for Dixcove are detailed to the very unit, while those reported for Appolonia are rough estimates rounded off to the nearest hundred. Indeed the British had the advantage of an incomparably higher degree of knowledge and control of the local situation in Dixcove than in the remote former kingdom of Appolonia, which they had invaded less than a year earlier, without having established any form of direct supervision whatever, let alone any real control.

Figure 1.

Dixcove

Dixcove has a population of 2857 individuals:¹⁹ 1570 of the inhabitants are classified as “free”, and 1287 as “pawns or slaves,” i.e 45.04%. However, the footnote states that from 600 to 800 individuals are to be added to the Swanzy Kweku subdivision, raising the grand total to between 3457 and 3657. It appears that these were “men able to bear arms.” It is best to interpret Fitzpatrick’s footnote to suggest that there were additional “pawns and slaves.” Indeed he wrote in his letter to Lord Grey that “nearly two thirds of the population are slaves and pawns.”²⁰

Figure 2.

Appolonia

In the Swanzy Kweku subdivision, 660 of the 1278 inhabitants are labeled as “pawns and slaves.” This subdivision, which was headed by the native merchant Richard Brace, was the location of the residence, stores, assets and staff of Frank Swanzy. The British trader and magistrate was a crucial node in the social and economic setup of Dixcove and its territory, and the apex of a hierarchy of patron-client relationships, which included a huge number of dependents, pawns and slaves whom he controlled directly or indirectly through his trade-partner and ally, Richard Brace. This is exactly what Fitzpatrick wanted to assert through his census table, namely to point to Frank Swanzy as the prototype of the European resident merchant who pays lip-service to emancipation and freedom while actually functioning as a stumbling block in the fight against unacceptable social habits and practices by fostering forms of personal and group subjection legally prohibited for British subjects (Swanzy 1956: 101). Fitzpatrick’s table on Dixcove suggests that “pawns and slaves” form between 45% and 55% of the total population, figures in line with the data we have for other parts of Atlantic West Africa of that day.

The Appolonia table tells a different story. To be precise, the document covers just a portion of the ex-kingdom, more or less the two capitals and the towns which headed the leading subdivisions of the army. They are all located on the littoral, with the exception of Nkroful (the birth-place of Kwame Nkrumah) and Adusuazo (or Edukrom). Curiously, some other important and well-known towns on the coast, such as Anokyi, Bonyere, Ezinlibo, Kabenlasuazo, etc. are not mentioned. It was only from 1857 that the British acquired some awareness of the demographic relevance of the towns in the interior of Appolonia, and only out of the need to extend the levy taxes. Fitzpatricks’ calculation of 9160 inhabitants contradicts other reliable estimates, which place the population of Appolonia at over 20000, a figure more in keeping with later assessments.²¹ Fitzpatick’s figure of slightly over 9000 is more likely a very credible idea of the population of coastal Appolonia: the area which is called locally Anwianu - the name means literally “on the sand/beach (anwia)” -, and which is perceived as sharing a strong common identity in opposition to Namulenu (namule: farming settlement or area), the interior countryside surrounded by the forest.

Atuabo and Beyin, the two neighbouring royal residences, held 4,000 people (over 43% of the total population). The rate of “houses”/” men capable of bearing arms” was 1/1 in all towns with the exception of Atuabo and Beyin, where it grew exponentially to 300 houses against 1000 men. Likewise, the “mature females” were 150% of the men in all the subordinate towns, while their number rose to 200% of adult males in the two capitals.

Evidently, the populations of Atuabo and Beyin had a different composition compared with the rest of the country, comprised as it was of the king’s direct dependents: his family, his notable followers, and especially the members of the king’s gyaase, who guarded and attended to the royal residence.²²

Brodie Cruickshank provided some details of the dramatic turmoil that accompanied and followed the collapse of the old order. He described a wave of riots, vendettas and marauding that struck Kaku Aka’s family, his wives, friends, supporters, and dependents. (Cruickshank 1849: Atuambo, Appolonia, April 30^th 1848). It is likely that a consequence of this disruption was a permanent exodus of some segments of the population. Others, such as dissidents and refugees of the ancien régime did eventually return from exile.. Fitzpatrick listed the town of Baku, which had a population of 1,400, as the third largest settlement in Appolonia. It was the hometown and private residence of ëbanyenle, the fugitive ex-general of Kaku Aka, who the British had installed as the new ruler of Atuabo. However, the exact demographic changes that resulted from these developments remain unclear.

A striking detail from Fitzpatrick’s table on Appolonia falls under the heading: “slaves and pawns.” It has no figures, just a sentence stating “all free. The parents having a right to pawn their children.” In this way, Fitzpatrick indicated that, in his view, at the time he made his estimate there were neither slaves nor pawns in the ex-kingdom of Appolonia. He was stating this about a section of the Gold Coast that in the past had been notorious in the British mind for the autocratic nature of its monarchical power and the peculiar situation where, according to an eighteenth-century European source, the whole population of the core of the kingdom was made up of personal slaves or direct dependents of the rulers.²³

The statement in Fitpatrick’s table was based on his assumption that the ancient hierarchy of personal dependency had been dismantled by the events of 1848, bringing about the king’s removal and the virtual suspension of the monarchy of Appolonia. The acting governor’s table was a statement to the effect that the end of the tyranny had resulted in the direct and immediate dissolution of all bonded status in the country: if all had been slaves of the king, his removal had set all slaves free. No more master, no more slaves.

While local power relationships had been shaken, they had not been totally subverted by the removal of their previous guarantor. This was indirectly substantiated in the population table by the sentence stating that rights which were inherent in structures of social leadership were still recognized: parents had the right to pawn their children, family-heads could pawn their fellow members, and chiefs could pawn their subjects. Thus the reproduction of forms of bonded status was assured. Slavery and pawnship, which Fitzpatrick proclaimed extinct in 1849, was soon restored by the new rulers, who built their own personal networks of slaves and pawns.²⁴

Fitzpatrick juxtaposed the massive presence of slavery and pawnship in Dixcove, Frank Swanzy’s fief, to the new situation of Appolonia, where the elimination of the old tyranny manu militari had a liberating effect that could be consolidated through the support of responsible British officers with more enlightened views than Swanzy’s, to extirpate such still deeply-rooted native customs as the right of families to reduce their fellow members into bonded status.

The two population tables are evidence of a perception and representation of slavery and dependency among Europeans and Euroafricans in mid-nineteenth century West Africa, well beyond the contingency of Fitzpatrick’s dislike for the habits and practices of British Gold Coasters. “Intolerable” slavery, in the sense of a direct form of total control over the life of the enslaved persons, almost necessarily conducive to harshness and arbitrariness, and as such no longer accepted by British law, was associated with the presence of an independent native society and political power: precisely what Appolonia had been like before the capture of its king. A milder, reformed, and to some extent “tolerable” native version of slavery might exist where native institutions were to some extent supervised and moderated by European presence and influence, as was the case in Dixcove and post-expedition Appolonia. The underlying assumption was that, wherever native institutions persisted as the basic framework of society, resident Europeans could not but accept a high degree of compromise with entrenched local habits and culture.

To return to our case, the bone of contention between Fitzpatrick on the one hand, and Frank Swanzy, Brodie Cruickshank, and other local merchants on the other, concerned the line between what was acceptable and what was unacceptable, between shared social languages and intrinsically incompatible ones, a line which was to be drawn through the extensive gray area created by an ancient history of African/European interaction on the Gold Coast.

Changing hierarchies of personal dependency

The decades following the abolition of the Atlantic slave trade witnessed accelerated changes in forms of subjection, such as slavery and pawnship. A new geography of personal dependence emerged within Gold Coast society, and basic social institutions were transformed by these developments, particularly family and residential patterns. The available data are fragmentary, and it is difficult to systematize them in a coherent picture. One may, however, attempt to outline a rough picture.

Compared to the eighteenth century, the mid-nineteenth century western Gold Coast was characterized by important differences in the modalities of settlement. The expansion of agriculture with the development of cash crops and increased production of food crops²⁵ brought about a redistribution of the population from towns to permanent farming settlements in forest areas. This development was associated with rising challenges to established hierarchies, especially political ones. One of Frank Swanzy’s statements before the 1842 Parliamentary Committee on the British bases along the West African Coast illustrates this process in the interior of Ahanta. All attempts by the British Magistrate court in Dixcove to provide arbitrations to African residents in the undefined area subject to the influence of the fort relied on the role of chiefs in enforcing the summonses issued by the Magistracy. Swanzy emphasized his notation that any effective exercise of British judiciary power or influence over any native individual could take place only “if he lives under one [chief].”²⁶ Swanzy’s words do not refer only to cases of disputed authority, but rather to the fact that many individuals and communities were not under the real authority of any chief whatsoever.

In his statement, Swanzy alluded to the extensive forest areas of Dixcove, Ahanta and Wassa: a large “frontier” of expansion for gold-mining and new agricultural production,²⁷ including cash crops like palm oil and groundnuts, in which it was not just the uniqueness or prevalence of political powers that were called into question, but rather the efficacy of any exercise of authority by any power, whether African or European. Swanzy outlined what was a relatively recent relationship between Dixcove and the communities of its hinterland. This had certainly not been the case in the eighteenth century, when widespread insecurity and protracted conflict in the region had determined both a substantial concentration of the population in fortified settlements and a distinctive hold on society by political and military hierarchies.

A critical reading of our documents gives an impression of growing mobility within the social landscape of the Western Gold Coast, which led to substantial changes in the condition of the bonded sections of the population. Although the hierarchies of status and power of the previous century were still in place and the power balance within society had not really changed, social transformations extended from distinct processes, such as the effective contraction of the transatlantic slave trade, a marked reduction in warfare, improvement in general security, a moderate extension of European jurisdiction over some coastal areas, the growth of so-called “legitimate” commerce, and new agricultural developments. Indeed, these phenomena produced an “opening up” effect on the Gold Coast society, paving the way for remarkable changes in crucial aspects of local life and culture, such as those entailed by religious change and the expansion of Western education.

However, the first and most important aspect influenced by this “opening up” of society was the fundamental structure of the family. The lower social classes and marginal groups were increasingly successful in asserting forms of autonomy from the power of the military/commercial aristocracy. In the eighteenth century, the ruler’s household and allied groups--in a word the gyaase of rulers and big men--had very often constituted a large percentage of the total population in the main towns, but during the nineteenth century the population was less concentrated in and around the chiefs’ residences, with communities formed by slaves and dependents tending to spread out to forest areas which they played a key role in colonizing. At the same time, this process contributed to radically reducing the direct hold and extraction capacity of rulers and big-men over large sections of the population of these areas, over which they had nominal control.

A crucial development was the substantial growth in the number of independent households, together with the increase in social relevance of matrilineages, different from the ruling ones. Such independent households and matrilineages were eager to augment their importance within the communities through acquisition of new members. These developments clearly favoured the integration and upward mobility of slaves, pawns, refugees, immigrants and, generally speaking, all, above all women. Dependents took part to an unprecedented degree in the efforts to build the new order, and in so doing they gained unprecedented rights to a share in its fruits.

These processes pervaded nineteenth century Gold Coast society, transforming the landscape of social relationships and setting the pace for radical social change decades ahead of the formal establishment of colonial rule. As far as the Gold Coast is concerned (not Asante, not the North), the forms of “domestic slavery” and pawnship which were eventually suppressed by colonial legislation, hastening the transition towards forms of wage labour, were already involved in local adaptive responses to the tumultuous changes of the early nineteenth century, which made those relationships a far cry from the slavery, panyarring²⁸ and pawnship that had characterized Gold Coast society in the second half of the eighteenth century.

In conclusion, Fitzpatrick’s population estimates say a lot about the thorny issue of bonded status and slavery among Europeans and Euroafricans in the mid-nineteenth-century Gold Coast. The “non-saids” in the two tables are even more meaningful than the many explicit declarations of crucial insights. The scant notations about the disproportionate number of “slaves and pawns” in the Swanzy Kweku subdivision of Dixcove and the gender ratio in Beyin and Atuabo--where everyone was ,in theory, free--outline a social institution which is the most effective agent for preserving and reproducing personal dependency. This institution is the gyaase of the leader: no matter if this leader is the king of Appolonia or the British magistrate of Dixcove.