The Rise of the ‘Impenetrable Hedge’: African Intermediaries and the Legacy of Colonial Wars in Northern Ghana, 1896–1920

Abstract

This article sheds light on the effects that the unrest created by early twentieth-century colonial wars had on the increasing power of African colonial intermediaries in the Northern Territories of the Gold Coast. By managing interpretation processes as well as developing a monopoly on the use of violence, interpreters, soldiers, clerks, and chiefs created what the officers called an ‘impenetrable hedge’. Analyzing how these different figures deployed their power, the article argues that there was historical continuity in the formation of the intermediaries and the methods employed by them between the years of the colonial wars and the later British administration. Despite their institutional role, however, precariousness remained a central facet of these agents’ experiences. The article shows that they were continually challenged at a grassroots level, and suggests that it was only through the use of coercion that they were able to maintain their position.

In 1910, almost ten years after the military occupation of the northeastern part of the Northern Territories of the Gold Coast, Captain S. D. Nash, a British officer who had recently arrived in the region, lamented in a report that he was surrounded by an ‘impenetrable hedge’.¹ He complained that a handful of African intermediaries had significant sway in the interpretation and the enactment of commissioners’ orders. Interpreters and soldiers, as well as their wives, prevented complaints from reaching the colonial district court, especially if they were involved in the case. Intermediaries were effectively isolating the officers from what was happening in the district. As this article demonstrates, this was not an extraordinary occurrence, but rather a widespread approach among the Northern Territories’ colonial intermediaries in the first twenty years of the twentieth century. The peculiarity of this hedge consisted in the way intermediaries managed their power. Although it partly followed the verticality of the colonial hierarchy, it also functioned according to horizontal patterns. The people who made up this hedge collaborated to actively influence the enactment of colonial policies.

The role of intermediaries was significant in Africa, especially in the early years of colonial administration. The ‘impenetrable hedge’ of the Northern Territories is indeed strikingly similar to the ‘circle of iron’ analyzed by Emily Lynn Osborn in the French colony of Guinea, where an interpreter and a chief built a ‘wall of deceit and lies’ around a French commander.² In the early phases of colonial rule, intermediaries in French West Africa became political brokers between the population, the officers, and the local elites. Since European officers failed to establish control over the processes of interpretation, intermediaries created networks that connected them with chiefs. Analyzing the role of the intermediaries is thus crucial to assess the ‘highly localized permutations’ of colonialism, as it enlightens the ways colonial administration deployed and managed power.³ The intermediaries’ agency is a crucial element to help us understand these permutations, as stressed in a later collection of essays edited in the trail of Osborn's work.⁴ The role that intermediaries played in influencing, resisting, and cooperating with the colonial administrative machine, could add significant nuances to the analytical palette of long-standing debates on colonial domination, resistance, and collaboration.⁵

Rather than ‘collaboration’, however, recent studies suggest that the term ‘cooperation’ might be better suited to expressing the relations established in colonial encounters. Indeed, cooperation suggests a ‘positive interaction’ that considers not only the dimension of exploitation and the asymmetry of the encounter but also the room for maneuver that local agents had.⁶ It also remains crucial to emphasize the extent of awareness that indigenous actors had of their positions, as well as the fact that cooperation was built on existing precolonial political frameworks. Again, to understand how intermediaries influenced colonial rule, it is essential to study and understand their life stories, interests, and roles in the administration. In most cases, they were ‘bricoleurs’ who had developed complex characters and directed their actions toward different interests and purposes.⁷ They could embody patterns of accumulation of wealth and power that had characterized precolonial ‘big men’ or could act as missionaries of ‘modernity’ for their societies.⁸ Alternatively, they could pursue ‘sub-colonial’ rule projects in which precolonial practices and systems were imposed on other social groups, as Moses Ochonu suggests for the Hausa elite of the Nigerian Middle Belt.⁹ In the end, intermediaries were in the middle of two or more different cultural landscapes and had to find a bridge between them.

A famous example comes from Amadou Hampâté Bâ's novels.¹⁰ In L’Étrange Destin de Wangrin, Wangrin is a clerk and interpreter based in French West Africa, who inserts himself shrewdly into the colonial administration to exploit the officers’ ignorance for his profit. At the same time, he has to come to terms with local political figures and thus create a successful mediation between colonial and African authorities. Ralph Austen has written extensively on the life and works of Hampâté Bâ and his character, tracing the ‘real’ Wangrin and providing historical support for some of the events narrated by Hampâté Bâ.¹¹ Austen deromanticizes the ‘real’ Wangrin, demonstrating that he was far less corrupt than the novels suggest. In doing so, he restores the colonial intermediaries to their status of historical characters and not heroes/antiheroes of literary productions.¹²

Joël Glasman's work on the colonial police in Togo also highlights the historical complexity and agency of African intermediaries.¹³ He criticizes the common assumption that there existed a fixed ‘verticality’ in the administrative hierarchy.¹⁴ Indeed, an intermediary's position — between the people and the officers — was neither static nor unquestionable. Their possession of different kinds of capital (political, educational, military, or economic) triggered competition with other agents of colonial rule and often attracted contestation both from the local population and officers. This means that the colonial echelons were fluid, dynamic, and ultimately, more horizontal than one might expect.

The questions this article seeks to address are related to the role of African intermediaries and the nature of colonial rule. What were the conditions that enabled the erection of the ‘impenetrable hedge’? How did they affect the evolution of colonial administration? Who was choosing and monitoring these intermediaries? How did they, in turn, influence the colonial machinery? Answering these questions contributes to the study of intermediaries in two ways. Firstly, this research highlights the historical continuity in the intermediaries’ management of power, at the juncture between the precolonial and colonial periods. Secondly, it assesses that at this juncture the intermediaries had a central (albeit precarious) role in the colonial administration, showing how they actively influenced the enactment of colonial policies. Far from resembling literary characters whose successes depended entirely on individual shrewdness, intermediaries in the Northern Territories worked collaboratively, creating networks, while continually having their power and careers challenged.

This article demonstrates that the impenetrable hedge differed from the circle of iron in at least three important aspects: historical continuity, networking, and precariousness. Firstly, the impenetrable hedge developed in direct historical continuity with the colonial wars. The people who composed it had been formed during the colonial wars of the late nineteenth century. Many were soldiers, mercenaries, and slave raiders, and as such had assiduously practiced violence. When they ended up in the British colonial army, violence and coercion remained central to their approach to power, and they persisted even after the wars and without colonial endorsement. Whenever the administration selected its intermediaries from these mercenaries, a primary matter for the latter was merging their practices with colonial orders.

Secondly, in contrast with the circle of iron, which was made up of one chief and one interpreter, the impenetrable hedge was the result of the machinations of numerous actors. Although constabularies, interpreters, and chiefs all competed for power and wealth, at the same time, they depended on each other. In much of the Northern Territories, British-appointed chiefs had neither a coercive power over the population nor even — in many cases — publicly recognized legitimacy to access the chief's office. This was a widespread conundrum in much of colonial Africa. In late nineteenth-century Cameroon — to name but one example — chiefs found themselves in a precarious ‘in-between situation’.¹⁵ Their position resulted from the fusion of African polities and German bureaucracy, as well as the German administration's need to rely on African precolonial authorities. Chieftaincy represented a political arena where people negotiated (or acquired ex novo) the extent of their authority. In the early phases of colonial rule in the Northern Territories, soldiers and interpreters (who were often the same person) were ideal allies for aspiring chiefs, in that they could provide administrative and military backing. At the same time, soldiers and interpreters found it convenient to cooperate with chiefs, because they were the main political figures formally recognized by the officers. The impenetrable hedge was thus composed of soldiers, clerks, interpreters, and chiefs who shared the same interests.

The literature on intermediaries in the region, however, is mainly related to individuals who reached and maintained an influential position in the administration by virtue of their political skills. An interesting example is the figure of Bassana Moshi, the official interpreter of the district, stationed in Zuarungu since 1914.¹⁶ Bassana, an ex-soldier, carved a privileged political position by exploiting his role as a ‘one-man “circle of iron”’ and becoming a pivotal political broker in the region through the manipulation of the interpretation process.¹⁷ Though Bassana undoubtedly had exceptional political skills, he is far from an exception: he represents the tip of an iceberg that sinks into a complex and controversial historical past. Intermediaries in the Northern Territories did not operate in an entirely novel way. The impenetrable hedge, as this article demonstrates, reiterated patterns and networks formed during the precolonial past.

The third element which sets the impenetrable hedge apart from the circle of iron is the precariousness of the intermediaries, fostered by the feature of ‘accidentality’ acquired by early colonial rule in the region.¹⁸ Central to this accidentality was the divergence between the colonial officers’ aims and what they achieved.¹⁹ This resulted in an extremely fragile and unstable colonial apparatus, which in turn allowed for continuous contestation at a grassroots level against the agents of British colonialism. Although not opposed by the local rulers, intermediaries were often opposed by the local population. It was only through the monopoly of violence that the hedge was able to maintain power. Whenever this monopoly was broken, its power was challenged.

The article opens with the warfare dynamics in the region at the end of the nineteenth century. The analysis of British military intervention demonstrates that some African mercenaries took advantage of the unrest created by military operations. Thanks to their association with the colonial armies, they were able to operate undisturbed and cement their status as self-proclaimed authorities. The presence of these raiders set a significant precedent for the rising colonial administration. While some of these mercenaries were identified and convicted, others managed instead to enter the administration as intermediaries. The article then demonstrates the role that these intermediaries played in the implementation of colonial policies in the 1910s and 1920s. The selection of interpreters and chiefs, for example, was in part directed and mediated by these intermediaries. Subordinates did not simply thoroughly execute the commissioners’ orders, merely following top-down formal hierarchies. Rather, these orders underwent various stages of mediation and negotiation, passing through the impenetrable hedge. The last part of the article demonstrates the precariousness of the impenetrable hedge, analyzing two cases of violent contestation against colonial intermediaries. Grievances in some cases were not directed primarily against British officers but against the intermediaries. Eventually, however, internal changes in the administration, the emergence of a new class of interpreters and the implementation of indirect rule led to a gradual disappearance of the impenetrable hedge.

A failed ‘pacification’

In the last two decades of the nineteenth century, political unrest in northern Ghana grew, owing to the arrival of new powers on the scene. Groups of Zabarima horsemen (Muslim warriors and traders from the Songhay area) consolidated their position under the leadership of the warlord Babatu, who began a series of military campaigns to expand the area under his control.²⁰ In the 1890s, George Ekem Ferguson, a Fanti civil servant in the British administration, was sent to ascertain the commercial value of northern Ghana and negotiate treaties with local rulers.²¹ He undertook two missions, one in 1892 through Gonja and Dagomba, and the other in 1894 when he reached Ouagadougou. Here, he signed a treaty with Boukary Koutou, selected in 1889 as the mogho naba, the highest authority in those territories.²² By 1896 the turmoil was at its peak. As the French and British troops were approaching, a new actor entered the scene, Samory Turé, a Maninka warlord and ruler who fought for years against European armies.²³ In 1895, his army invaded western Gonja and Wa. Although he tried to form a coalition with the Zabarima's leaders, his troops were driven away by the French army.²⁴

By the end of the year, the British had already begun their military intervention in the region, bolstered by the support provided by Ferguson's diplomatic activities. In 1896, under the leadership of Major H. P. Northcott, the British troops occupied Gambaga, the Mamprusi commercial capital. The next year, Northcott was appointed the first commissioner and commandant in the Northern Territories of the Gold Coast.²⁵ His orders were to continue to pursue, wherever possible, the British advance. In the first half of 1897 two more expansionary missions were undertaken, one in the northwest, under the command of Lieutenant F. B. Henderson, and another towards the northeast, directed by Captain D. Stewart. Henderson was able to sign a treaty in Wa but was attacked and defeated by Samory's army shortly after. On this occasion, Ferguson, who also participated in the expedition, was killed, and Henderson was captured.²⁶ That same year, Babatu experienced his downfall and was compelled to leave British territories, while, in 1898, Samory was captured by the French.²⁷

Although the literature on colonial wars in northern Ghana tends somewhat to place the figure of Koutou, the mogho naba of Ouagadougou, in the background, his story is of central importance for this article, since some of his men became colonial intermediaries in the British administration.²⁸ In 1897, French troops occupied Ouagadougou, exiled Koutou, and installed his brother in his place.²⁹ Following several unsuccessful attempts to retake the throne, in 1898 Koutou was forced to retreat to the Northern Territories. The British administration, by virtue of their past treaty, decided to have the former mogho naba and his entourage settle in Zongoiri, a settlement not too far from Gambaga, where he stayed until his death in 1904. Their presence, however, contributed considerably to the political instability of those territories.

Raiding and looting were common practices endorsed by European officers during the war, and they did not stop afterwards. With the end of the war, the British military operations in northern Ghana took the form of raids intended to suppress caravan robbing and any form of local hostility against the colonial presence. The process of so-called ‘pacification’ lasted a decade among the noncentralized societies, especially the Dagarti and Gurensi, ostensibly because of their continued hostility and pursuit of banditry.³⁰ However, ignorance of the context and prejudices against the local population continued to be part of the colonial approach. These biases certainly helped to legitimize conquest and occupation, in the eyes of the officers, as a just and civilizing mission.³¹ The officers were convinced that the noncentralized societies were essentially devoted to banditry, even though this conclusion was not always backed by clear corroborating evidence.³² The end of the colonial wars resulted in a high presence of dismissed soldiers from the armies involved in the conflicts. They formed small bands to gain wealth through the raiding of caravans and settlements. Some of these raiders, ironically, operated in the region by collaborating with the same officers who formally professed to be advocates of peace and civilization. The settlement of Koutou in Zongoiri, for example, led to an increase in the caravan lootings on nearby routes, most likely performed by his army.³³

If most of Koutou's entourage settled in Zongoiri, twenty of his horsemen were conscripted in 1899 in the Gold Coast Regiment (GCR) under the command of Major A. Morris to be employed in the pacification raids against noncentralized societies.³⁴ Thanks to their prowess in raiding, they attracted the praise of many officers, who saw them as a powerful and helpful brigade.³⁵ This could explain the fact that some officers tolerated their activities of slave raiding and pillaging against the population, even though they recognized the dangers of using such a brigade.³⁶ During one of the raids under the British army, an officer complained in such a manner:

I am disinclined to employ Moshis as on the previous expedition they gave much trouble – looting friendly villages … these men are the scourge of the neighbourhood – I have punished 7 in the last 3 months for looting and slave dealing and I believe the hostility of the Fra fras is largely due to their unprovoked depredations.³⁷

However, this unit did not stop its activity. Thanks to their allegiance to the British army, some continued their pillaging undisturbed, convincing the local population that their presence and activities were endorsed by the Europeans. Others, instead, entered the colonial administration as interpreters.

The so-called ‘Moshi horsemen’, though, were not performing according to completely new scripts. Enid Schilkrout asserts that Koutou's horsemen were mostly nakomce.³⁸ To be nakomce in Mossi society meant claiming descent to royal lineage, namely having an aristocratic status that allowed access to the role of chiefs.³⁹ They were entitled to demand tribute from the population, which, if not delivered upon command, could be taken by force. The majority of the nakomce were also horsemen in the Mossi army, which means that they were accustomed to raiding, a frequent military tactic of the Mossi cavalry.⁴⁰ In particular, Koutou's horsemen had been accustomed for years to supporting themselves through looting, seen as a legitimate form of maintenance for these men of Mossi nobility who were deprived of an income along with their exiled leader.⁴¹ Another activity considered legitimate for the nakomce was slave raiding, carried out in particular against noncentralized societies, which Koutou's horsemen took part in while acting for the GCR.⁴²

The actions of Koutou's horsemen can be thus interpreted in a more nuanced way. While they certainly acted for profit, it would be naive to make them fit into simplistic dichotomies that take for granted the existence of an already fully functioning and hegemonic colonial legal framework. They were not simply bandits or outlaws, but aristocrats who, having lost their status, were trying to carve out a new position of power in the interstices of colonial administration. Their actions were clear attempts at adapting political ideologies and coercive practices derived from their society to a new and unstable socio-political context. Facilitated by their linguistic and cultural proximity to noncentralized populations, Koutou's horsemen tried to reassert their nakomce rights by installing themselves as chiefs in this region on the fringes of colonial control. What could be read as an abuse of their position thus also takes on the connotations of a reproduction of a centralized social system based on coercion, hierarchy, and genealogical distinction. In other words, the horsemen tried to regain the political legitimacy taken away from them in the colonial wars, at the expense of the noncentralized societies. A specific case that illustrates this is that of Asana Moshi.

Between 1901 and 1907, the town of Bolgatanga in the Gurensi area became the base for a part of Koutou's horsemen who were enrolled in the British army. One of Morris’ flagbearers, called Asana Moshi, was reported to be the head of a group of horsemen that looted the region in this period.⁴³ Asana and his retinues raided the neighboring settlements for slaves, acting in the name of the British administration.⁴⁴ It is possible that they had the complicity of some locals, too.⁴⁵ Asana compelled the people in Bolgatanga to build a ‘castellated residence’ for him, and he started to act as chief, hearing and judging cases, as a Mossi chief would have done in his settlement.⁴⁶ Asana's behavior fostered hostility and suspicion against him, so much so that some of the Gurensi victims reported him directly to the chief commissioner and brought his actions to the court.

The records of the trials against Asana Moshi demonstrate that he had effectively managed to obtain a position of chiefship for himself and that he justified it through his affiliation with the colonial administration. He was directly involved in two trials: in the first, he was charged with the murder of a man, seemingly because he had failed to comply with a marriage agreement.⁴⁷ In the second case, Asana was accused of killing two men, acting on a grievance.⁴⁸ In both murder cases, Asana did not act alone but was accompanied by a dozen other Mossi horsemen who actively participated in the events. Furthermore, the witnesses asserted that they did not report the facts to the British officers because ‘Sana had told them he has been sent by Colonel Morris to command all that District’,⁴⁹ or ‘at that time we regard Sana as [a] whiteman’.⁵⁰ These depositions allow us to confirm that Asana not only performed the script of a Mossi nakomce (thus allowing himself to judge and dispose of the lives of his subjects) but that he also justified his actions through his affiliation with the colonial administration. He was finally arrested in June 1907, and in August of that year, he was tried for murder. He received a guilty sentence and was publicly hanged in Gambaga.⁵¹

The execution of Asana Moshi was viewed with optimism by the administration.⁵² Some officers saw his arrest as a new occasion to reaffirm their power in the unsettled area, since they hoped it would convince the population of the righteousness of the colonial mission and motivate the opening of a colonial station to maintain order. In fact, in 1910, the Company of the First Battalion of the GCR moved to Zuarungu.⁵³ However, the settling of the soldiers in the still unsettled area laid the foundation for the development of the impenetrable hedge. A hint of what would happen in the following years could already be found in the last lines of the proceedings against Asana Moshi: the chief commissioner was quite disappointed by the silence of one of the soldiers who witnessed the events.⁵⁴ Sergeant Abuchi Grunshi — not a member of the Moshi horsemen unit — was indeed remarkably reticent to speak about the defendant, a telling silence which revealed his complicity with the horsemen. The Mossi horsemen's exploitation of their position set an important precedent for subsequent intermediaries. In later years, coercion and abuse remained widespread practices in all ranks of the administration.

The impenetrable hedge

The militaristic configuration of the administration in the Northern Territories was maintained until January 1907, when it shifted to a civilian one, which was considered more likely to bring the population closer to British methods of government.⁵⁵ Nonetheless, it was characterized by a profound vacuum when it came to the chiefs’ powers. The absence of a well-defined and articulated colonial policy did not mean that there was no policy at all, however. The political spaces left empty by officers were filled by constabulary and interpreters, many of whom had been involved in the war efforts of previous years.⁵⁶

The relationship between the chiefs and intermediaries intensified to comply with colonial administrative prescriptions, and in order for both parties to maintain their positions. Indeed, in this period, they developed patterns of mutual collaboration that altered the verticality of the colonial hierarchy. Since 1899 at least, the individuals recognized as chiefs were regarded as the direct agents of British power in the Northern Territories.⁵⁷ In the eyes of the officers, the chiefs’ functions were essentially limited to complying with British requests, such as preservation of peace, collection of taxes, and gathering of labor. In particular, forced labor modified the way chiefs approached and managed power.⁵⁸ In the nineteenth century, chiefs’ powers were characterized by an absence of control over both force and coercion.⁵⁹ Violence was a part of everyday life in noncentralized societies, but power was not exercised through intimidation.⁶⁰

The next step undertaken by the British was to try to convince the local population to obey the people that officers wished to be chiefs. If there were no preexisting chiefs, the administration installed new ones. If the ones who were already present were not in agreement with the officers, they were replaced. In this way, British officers sought to impose a hierarchical and centralized system of authority that clashed with the local noncentralized and nonhierarchical forms of power. Chiefs did not have the public legitimacy to collect taxes or force people to work for them, and they lacked a coercive apparatus to enforce their decisions. But this political transformation could be eased by the close association of the intermediaries. British intervention was guided by these actors, and, in many cases, colonial officers were not fully aware of the context in which they were operating, as in the case of Asana Moshi.

In the first two decades of the twentieth century, chiefs (many of whom were illiterate) did not usually communicate directly with the officers, primarily because of linguistic barriers: most of the officers did not speak the vernacular.⁶¹ Interactions were presided over by the constabulary who served as mediators between the parties. Beginning in the late nineteenth century, the constabulary of colonial Ghana was recruited mostly in the Northern Territories and trained in Tamale.⁶² In this way, a large part of the soldiers could converse in local languages and English, thus being able to act as interpreters. Commissioners relied mainly on autochthonous soldiers as interpreters, but, when these were not available, they made use of soldiers fluent in Hausa, thus translating local languages into Hausa and Hausa into English, which very often led to errors and misunderstandings.⁶³ The only literate intermediaries tended to be clerks; they however came mainly from the southern regions and were therefore not familiar with northern languages. Moreover, they had limited relations with the chiefs. Soldiers, on the other hand, were involved in carrying out a commissioner's orders from the moment of pronouncement to concrete implementation.

The collaboration with the constabulary proved very useful for those chiefs who struggled to impose their authority. The chiefs’ newly acquired power over the population was neither taken for granted nor freely accepted. In the first decade of the twentieth century, for example, locals contested the role acquired by the chief of Bolgatanga, one of the most densely populated settlements in the area.⁶⁴ He found in the British officers and the constabulary powerful allies: this led him to come into conflict with other settlements that contested his legitimacy, which were attacked and destroyed in 1899.⁶⁵ Ten years later, it was reported that the chief was still encountering strong opposition, as many headmen of the nearby settlements did not accept his requests for labor.⁶⁶ In response to these ‘insubordinations’, the administration paved the way for Bolgatanga's paramountcy by jailing these men for them to ‘learn sense’.⁶⁷ In this situation, colonial intervention nullified any direct opposition, and the chief of Bolgatanga was able to gain authority not only in his community but also among the neighboring settlements. A new hierarchy within the Gurensi communities was created, based on compliance with British orders and cooperation with their intermediaries.

In the 1910s, most chiefs continued to use military support from the constabulary to assert their power over the people and comply with British requests. When faced with a request, such as road making or carrier services, the chiefs usually asked for a military escort to carry out the orders.⁶⁸ The officers viewed this as a necessary step, as long as the chiefs’ lack of power remained ‘ineradicable’.⁶⁹ On the other hand, the chiefs explicitly endorsed the presence of the constabulary to show the people that they held limited responsibility for executing colonial orders.⁷⁰ However, chiefs also resorted to this sanctioning of violence to accumulate personal wealth while acting in the name of the ‘white man’, as Asana Moshi and other ex-soldiers had done in previous years. Some of the chiefs promptly adopted this stratagem in association with soldiers to request labor or seize cattle for their gain.⁷¹ In this period, chiefs also began to act as judges in the newly established courts in the district headquarters.⁷² They now had military backing to enforce their decisions, a situation that led in some cases to abuses of authority.⁷³

These new patterns of political control remained precarious, however. Contestation and resentment against chiefs and intermediaries remained a pervasive feature of the region. In March 1909, after the death of the late chief of Nangodi, the chief commissioner appointed a new chief in the settlement. The choice was evidently an ‘administrative error’, as the provincial commissioner would later define it.⁷⁴ The error resided in installing a chief without the people's consent and support. The newly selected chief was beheaded and his compound was burned to the ground by the people of the neighboring settlements who did not accept his legitimacy. These events eventually led to a ‘punitive’ expedition.⁷⁵ A few days after the murder, the settlements of those recognized as the culprits were leveled to the ground and at least 35 people lost their lives in the battle against the constabulary.⁷⁶

To better control the area and avoid similar incidents, in the same year colonial officers opened the first administrative station in Zuarungu.⁷⁷ The building of the post was a landmark event in the history of the region. For the first time, the British officers had a permanent structure to reside in. Their presence became more stable, even though most of the time the commissioners would be busy touring the district. The chiefs, in turn, had their actions supported by the constant presence of the constabulary who were transferred there. Although in the long run most of the colonial personnel ended up peacefully settling in the area and intermarrying with locals, initially this relationship was conflictual.⁷⁸ The officers admitted that the ‘sudden influx of troops and followers of the government has fallen hard on the immediate neighbourhood’.⁷⁹ Soldiers were a nuisance the officers knew well, coming to their attention either for abducting women, looting settlements, or disturbing the people frequenting the markets.⁸⁰ In the case of Zuarungu, however, even clerks were suspected of extortion and a ‘good deal of pilfering’ from the local people.⁸¹ Colonial reports were quite optimistic about the opening of the post, but there is no doubt that, in the beginning, it had a detrimental effect on the population, reducing trust and heightening tension against the intermediaries.

Once the station had been opened, and clerks and constabulary permanently stationed there, they carved out a position of prominence for themselves in the colonial echelons by exploiting the gap of ignorance, prejudice, and linguistic barriers that separated the European officers from the local population. In 1910, shortly after his transfer to Zuarungu, Captain Nash defined this newly-emerged class as an ‘impenetrable hedge’:

in a district like this, where the natives are so ignorant of our methods and procedure, it is most difficult for the district commissioner to prevent himself being surrounded (by native officials who are supposed to assist him) by an impenetrable hedge through which no complaints are allowed to come – soldiers, constables, their wives and numerous retainers will not allow complaints to be brought to the court if they are implicated in the matter.⁸²

According to the officer, the interest of the people composing the hedge was to keep the commissioners divided and isolated from the rest of the population. The class of intermediaries that the British selected to operate in the region was in fact managing power outside the colonizers’ gaze.

The people composing the impenetrable hedge were quite aware of the importance of their positions in the colonial administrative structure. They exerted a certain degree of control over some of these positions. For instance, they were in charge of mediating the selection of interpreters. A commissioner remarked in 1910 that the lack of reliable interpreters knowledgeable in the local languages was due to the actions of the constabulary, as these latter were determined to maintain a certain degree of manipulation in the process of selecting interpreters. Indeed, the young man selected by the officers quit his job, apparently because the soldiers had subjected him to blackmail and brought false charges against him.⁸³ Another clear example of this was the move to appoint a retainer of Asana Moshi as the interpreter in Gambaga, the British headquarters in the province. This man was seemingly ‘hand in glove with the Sergeant and other N.C.O.'s of the constabulary’.⁸⁴

The chance to manipulate the recruitment among intermediaries gave the constabulary and chosen interpreters a degree of leverage in the process of selecting chiefs. On October 1913, a report denounced that a large number of individuals who aspired to become chiefs ‘certainly realized that the officials … were there only temporarily, and they and the interpreters took full advantage of their lack of experience’.⁸⁵ That same year, the court interpreter in Zuarungu, Salifu Frafra, was removed from his office because he ‘carried corrupt practices to an almost unprecedented pitch’ by influencing the selection of chiefs, accepting bribes, and giving false interpretations.⁸⁶ The provincial commissioner clearly expressed his suspicion that the chiefs and this man were colluding in the process of appointing chiefs.⁸⁷

This state of affairs persisted in the following years. In 1915, the White Fathers missionaries noted that the process of court interpretation in Navrongo was quite inaccurate and represented a source of falsification and profiteering.⁸⁸ The illicit and extortive methods employed by the interpreter Bassana Moshi, put in charge of the Zuarungu district beginning in 1914, have already been documented.⁸⁹ In this case as well, the man was implicated in a bribery case in connection with the election of a head chief.⁹⁰ The process of chief selection and appointment and the enactment of colonial policies were thus not controlled entirely by either the officers or the chiefs. Instead, this process involved intense negotiation among the colonial administration, its intermediaries, and the aspiring chiefs, thereby making the commissioners’ decisions malleable and alterable. And this in turn created an informal horizontal structure of power where chiefs and intermediaries could collaborate to reach their objectives.

Reprisals against African intermediaries

Despite their privileged status, the intermediaries in the Northern Territories found themselves often in a precarious position. Supporting the colonial administration would have led to insubordination and hostility from the rest of the population while backing the population would have attracted punishment from the colonial officers. The degree of ‘despotism’ exercised by the chiefs was thus nuanced and negotiated.⁹¹ Officers’ complaints concerning the ‘incapacity’ of the chiefs to comply with their orders persisted, indicating the line of compromise that some of these political figures started to pursue.⁹² Relations among the people of the district, officers, and intermediaries worsened during the First World War. As soon as the constabulary moved from Zuarungu to Gambaga in August 1914, complaints of insubordination followed in the next months.⁹³ From the beginning of British occupation, people had believed that the colonial presence was only temporary and in this period the hope intensified, giving rise to an increase in insubordination and threats to the chiefs and the intermediaries.⁹⁴

In 1916 the situation had deteriorated to such an extent that violence finally erupted in the so-called ‘Bongo riots’. What the officers had labeled ‘riots’, were in fact no riots at all. The term riots suggests that these were sudden outbursts of violence disturbing the peaceful conditions established by the colonial state. In reality, however, ‘they were symptomatic of continuing popular discontent with the structure of African administrative intermediaries that the British had imposed’.⁹⁵ Many actors came under serious attack in this period, not only intermediaries but also the officers themselves, and this uprising was only the most evident expression of a resentment that could be traced to the first years of colonial occupation.

A fight between two different sections in Bongo erupted, apparently over a land dispute. As the chief of Bongo was not able to quell the conflict, constables were sent in by the commissioner. The soldiers were likewise unable to calm the situation and fighting ensued. In the following quarrel, one of the constables was killed. The brutality of the colonial response was reminiscent of the violence of the so-called pacification period. The military raids that followed were backed by a ‘strong line of policy’ that led to the ‘usual compound-destroying and driving in of cattle’.⁹⁶ The following statement, made by Lance Corporal Bayoro Grunshi to justify the constables’ cattle looting in connection with the civil unrest that occurred, is very revealing of the constabulary's attitude. He declared: ‘If your father sends you to the farm to collect ground nuts, he expects you to take some for yourself’.⁹⁷ The violent actions of the Mossi horsemen at the beginning of the century, as well as those of many other colonial intermediaries who sought personal profits for their performances, seem to echo in these words.⁹⁸ The constabulary stationed in Zuarungu persevered once again in maintaining the monopoly of violence established during the war, with the connivance of the officers who kept a laissez-faire approach. The officers’ and intermediaries’ actions in response to the unrest were so violent and heavy-handed that Governor Clifford accused them of ‘inhumanity, hypocrisy, insufficient travelling, and a failure to investigate incidents before taking indiscriminate punitive action’.⁹⁹ In the ensuing years, acts of insubordination and dissent against the colonial political apparatus endured, albeit without attaining the same magnitude witnessed during the Bongo uprising.

In 1917 the same factions that had been involved in Bongo fought once again, this time over an attempt to recruit soldiers.¹⁰⁰ The initial method of holding meetings to persuade people to join the army did not prove very successful, and recruiters then proceeded to make direct requests to headmen in every settlement.¹⁰¹ However, the chief commissioner found this approach too coercive and soon replaced it with one in which the chiefs were ordered to produce a certain number of men.¹⁰² The results of this last method are particularly revealing of the widespread resistance and noncompliance with colonial requests. In the first instance, insubordination and resentment against chiefs increased. Even though chiefs presented a certain number of men, these individuals either quickly deserted or were ‘unsuitable’, being too young, too old, or not physically fit. This was probably a strategy some chiefs employed to comply with colonial requests without depriving their settlements of able-bodied men and causing too much resentment.¹⁰³ However, the recruitment shortly became ‘a high-stakes game of hide-and-seek’, with the officers looking for recruits and the young men hiding from them in the fields.¹⁰⁴

The role of the chief of Bolgatanga as a recruiter was praised and awarded by the officers, but it led his chiefdom to the verge of violent reprisal.¹⁰⁵ In November 1917, he reported a case of resistance and death threats encountered during his recruiting attempts.¹⁰⁶ This time, however, the memories of the previous year's uprising were still fresh, and the constables were ordered not to provoke open fighting. Some of the men connected with the threats were arrested, but one, Akanyele, was able to escape and hide in a settlement not too far away. Nothing occurred until the chief of Bolgatanga notified the district commissioner of Akanyele's return almost six months later. In May 1918, an expedition composed of the district commissioner A. W. Cardinall, three constables, the interpreter Bassana Moshi, the chief of Bolgatanga, the chief of Zuarungu, and some of the chiefs’ followers marched from Bolgatanga to Akanyele's compound. A stalemate was followed by a shooting. Akanyele was killed with five shots. On the other side there was only one injury: Bassana Moshi took an arrow to the shin. Apart from the support that the constables and chiefs comprising the expedition gave to the district commissioner, the governor himself praised Cardinall for his ‘courage’ and defended his actions. This man's death did not lead to any violent reprisals against the chiefs or the intermediaries in the following years. However, nearly twenty years after the Bongo uprising, intermediaries still felt in danger. In 1933, Bassana Moshi asked permission to keep a carbine for his protection.¹⁰⁷ The interpreter was scared because people were still threatening him.

Breaching the hedge

Violence and abuse of power by colonial intermediaries in northern Ghana reached a peak with the Bongo uprising. After the First World War, for a variety of reasons, their position became more consolidated within the administration and offered less opportunity for indiscriminate coercion and misconduct. Colonial administrators were certainly aware of the manipulation to which they were vulnerable, particularly when learning the ropes at a new posting. Internal transfers of officers were therefore gradually reduced, in order to obviate the problem. Moreover, in the interwar period, knowledge of the local languages began to be mandatory for commissioners.¹⁰⁸ In this period, important steps were also taken in terms of education and the production of colonial knowledge on colonized societies. Although the first Western-style school in the region was opened in 1906, it was not until 1919 that investment in education increased significantly.¹⁰⁹ Moreover, especially in the early years, the main pupils were the children of the constabulary and chiefs.¹¹⁰ Yet the 1920s marked the entry of a new class of colonial intermediaries, trained as literate Christians and more conversant with the ‘word of paper’, such as Bali Cypriani and Victor Aboya.¹¹¹ They took on a pioneering role in colonial knowledge production, serving as interpreters and translators for both commissioners and the first anthropologists who worked in northern Ghana.

Bali Cypriani was the first autochthonous Christian literate who worked as a clerk and interpreter in northern Ghana, yet has received the little attention in the literature. He was born in a Kassena family in Ouagadougou, where he met the White Fathers, converted to Christianity, and received primary education.¹¹² He followed the fathers to Navrongo in 1906 when they opened the first Christian mission and the first Western-style school in northern Ghana. Through his education and zeal for proselytism, he became the White Fathers’ ‘homme de confiance’ and got involved in their educational efforts.¹¹³ Cypriani was the first teacher in the Navrongo school, but his lessons were directed not only to the newly converted Christians: he also lectured the missionaries on local languages. Due to his skills as an interpreter and translator, he was hired by the colonial administration as a clerk starting in 1915.¹¹⁴

His work led to a significant improvement in the quality of interpretations. Up until then, the administration had not been able to rely on literate local civilians who knew the local languages, as already pointed out. Cypriani's work allowed for reliable interpretations from vernacular to English for the first time, outside of those provided by the constabulary. He continued to work for both the colonial administration and missionaries, not only operating as an interpreter, clerk, teacher, and proselytizer but also actively contributing to the ‘colonial library’ in the region. For example, he compiled and collected information for the 1921 census in the Northern Mamprusi District.¹¹⁵ He was also the main advisor to Cardinall for his 1920 monograph on the Northern Territories and to commissioner P. T. Whitall's report on Kasena religious and land tenure practices.¹¹⁶

Victor Aboya, in contrast to Cypriani, is a much more present figure in the literature. And, like Cypriani, he contributed significantly to the change that was taking place among intermediaries. Born in the late nineteenth century and enslaved in his infancy, he lived part of his life in Kumasi, where he was freed by the Basel missionaries, received primary education, and converted to Christianity.¹¹⁷ In 1917, he managed to return to his birthplace in the Northern Territories, where he began working with the White Fathers. In particular, he became involved in translating the catechisms and gospels into local languages.¹¹⁸ He is mainly remembered, however, for his collaboration with R. S. Rattray, the first colonial ethnographer who worked in the Northern Territories. Rattray was an amateur ethnographer sent to northern Ghana to facilitate the implementation of indirect rule. Shortly after his arrival, he started a collaboration with the White Fathers who taught him the local languages and introduced him to Aboya, who became his principal research advisor. Much of Rattray's work in the region is the result of their collaboration and it is based on a methodology that had been developed directly by Aboya.¹¹⁹

The chances for falsifications and abuses decreased drastically also for chiefs. Beginning in the 1920s and even more so in the 1930s, their position was consolidated and the boundaries of their power more precisely defined. By the end of 1929, under the mandate of Governor A. R. Slater, the new policy to implement indirect rule was launched. This reform intended to amalgamate the chiefdoms of the Northern Territories, thus placing the chiefs at the head of the new Native States from a territorial, juridical, and financial point of view.¹²⁰ In particular, beginning in the 1930s, many chiefs began to develop a new political strategy that endured in the postcolonial period, defined by Alice Wiemers as ‘progressive chieftaincy’.¹²¹ The political language, which had previously centered around coercion, was gradually replaced by a new rhetoric increasingly based on modernization and development, to achieve recognition and state funding. The methods of the impenetrable hedge eventually became obsolete and anachronistic.

As many scholars have already pointed out, colonial records provide faded and unclear images of the intermediaries, who tended to leave traces in archives whenever they did not act cautiously.¹²² They formed a hedge almost as invisible to the historian as it was impenetrable to the colonial officers. Despite this, this article has shown that through a careful reading of the sources, it is still possible to identify connections and collaborations among intermediaries, to follow their paths in the records, and, ultimately, to confirm their central role in the colonial administration. Intermediaries often created niches of power, taking advantage of the authority given to them by the colonial administration, and exploiting the incomplete control the officers had over them.

Intermediaries were crucial in defining colonialism's local permutations in Africa. Colonial intervention was so pervasively mediated by Africans that it would be difficult to conclude that the officers’ intentions were always thoroughly enacted. Indeed, intermediaries were rarely passive actors of colonialism. The analysis of these figures allows us to understand the weaknesses and cracks that colonial administrations developed over the years and provides crucial insight as to their degree of hegemony, as well as the ramifications and the performances involved in transforming Africans into colonial subjects. The impenetrable hedge defined and influenced how early colonialism was realized in the Northern Territories. Even when not evident in the records, African mediation and cooperation were always driving forces in the enactment of colonial policies. Therefore, rather than being eminently European or African, colonialism in the Northern Territories was the result of an encounter between different ways of conceptualizing power, an encounter that was ‘quintessentially colonial’.¹²³