Different Ideas of Borders and Border Construction in Northern Ghana: Historical and Anthropological Perspectives

Introduction

The concept of space in sub-Saharan Africa has been shaped in distinct and lasting ways by the international and internal political boundaries that the European colonial powers defined. Borders and borderlands today represent important theoretical frameworks to analyse the processes of identity construction, changes in local political dynamics, and the nature of conflicts in contemporary Africa. Through close studies of African societies and their interactions with each other, particularly via anthropological research, it is clear that despite western conceptions of borders as fixed they have been and remain fluid, highly permeable, and shifting space markers. Indeed, in the absence of a local equivalent to their European understandings of and desire for boundaries and borders in African societies, colonial officials continuously faced conflicting conceptions of space.

In this article, I reconstruct the British process of demarcating borders in the eastern part of the Northern Territories of the Gold Coast, the contemporary Northern Region of Ghana, during the first half of the twentieth century. I am particularly concerned with districts in which Konkomba, Nanumba and Dagomba predominated. Building from archival, ethnographic, and oral sources, I explore contemporary conceptions of borders and their historical context and evolution. I focus on spatial and territorial issues; that is to say, how the contemporary international border and internal boundaries has been delimited and constructed through the imposition of European models and the interaction of different local representations and perceptions of space. My goal has been to understand how and through what process local political equilibriums became so fragile. I also seek to give prominence to the agency of African chiefs and leaders, which is too often neglected when we talk about African boundaries.

To explore borders from an anthropological perspective, it is essential to consider local cosmological constructions of space, colonial and postcolonial representations boundaries, and such markers in contemporary social practice (Barth 1969; Hagberg & Tengan 2000: 24). Following the research methodology offered by the work of Harberg and Tengan (2000), I define cosmology as the way space is perceived and conceptualized by local people and, by extension, how they construct and conceive of boundaries (ibid.: 17).

Ghana’s Northern Region offers myriad local ideas of space and border that intersect and hybridize the local administrative and cultural structures which are legacies of British rule (Bening 1981). During the colonial period, local political systems, essentially determined by a series of alliances disseminated on the territory, aligned with British structures, characterized by a concept of state delimited by fixed borders, to form new communities that one might define as “invented” or “imagined” (Lentz 2000a). This critical phase of African-European engagement involved a multiplicity of cultural and political actors—both African and European – with different intentions, who produced different interpretations of the local environment. Yet, the result was a new conceptualization of borders and borderland within local communities.

The British assumed control of the northeastern section of the Northern Territories in 1919 from German Togoland. The League of Nations defined the area as a protectorate, following Germany’s loss of its colonial territories after the First World War (see Austin 1963; Pyeatt 1988; Schuerkens 2001). The societies in this region had deeply rooted reciprocal connections, which, in many respects, remain today. Yet they are trapped in the colonial typological definitions, which rigidly divide them into “organized kingdoms” (Dagomba and Nanumba) and “stateless” groups (Konkomba). Such a dichotomy, and the profound discordance and conflict borne of it, have been widely discussed in the context of the Northern Region of Ghana (Drucker-Brown 1989, 1995; Lentz 1995, 2000a; Bogner 2000; Brukum 2001; Kunbuor 2002; Skalnik 2002, 2003; Lund 2003; Pul 2003; Lentz and Kuba 2006; Casentini 2008; Wienia 2009; Talton 2010). In the 1930s, the British placed the Konkomba and all the other “stateless” groups under the direct control of neighboring centralized kingdoms that the colonial power recognised as “legitimate tribes” (Rattray, 1932). This system, based on the functional use of the chieftaincy institution, gave rise to an unequal access to resources, principally land.

Local conceptions of borders hinged upon the nature of political organization, the territory’s demographic distribution model, and land use, as well as upon relations between neighboring political entities and societies. Northern societies and political entities considered in this article shared cultural, social, and political spaces, and often resided in mixed villages. There were different systems of both spatial conceptualization and movement even within the groups living in the area; it is important to underscore the fact that those systems were often as different from each other as they were from the British model imposed upon them by the colonial administration.

What we observe today are the products of the intersection of these models. As evolving historical products, those models changed through encounters, clashes and connections, creating new elements to redefine each respective feature. When considering socio-political relations during pre-colonial times, it is useful to consider the concept of “functional region,” a term that allows us to rearticulate and foreground the relationship between people and the territory (Howard & Shain 2005). People’s mobility within a space is the foundation for both their relationships with that space and conceptions of it within a broader regional context in which networks define shapes, limits and borders. In pre-colonial Africa, as Jeffrey Herbst contends (1989), populations were widely distributed. Urban society was the exception, and many cities were mobile and agricultural cultivation was extensive. As a result, political control was not clearly connected to a determined territory, but was more concentrated on people, rather than space. For this reason, population distributions did not provide much information on how to draw stable boundaries (Herbst 1989: 679). Therefore, when reflecting upon borderlands in West Africa, it is necessary to consider the “network map” of this region as a whole, mapping the set of economic and political powers mutually connected through commercial, taxation, patronage and migration ties (Mabogunje & Richards 1971). As Allen Howard explains, the regional dimension, often marginalized by scholars in favor of connections between local realities and global processes, highlights some basic aspects of the current equilibriums (Howard 2005: 51-55). The regional dimension, in fact, enables one to concentrate on the historical profundity of socio-political relations, disclosing the dramatic impact of the various territorial changes that happened in the course of history.

Local sites in Northern Ghana, indeed, have always been connected to wider regions in history, where wars, incursions and trade have generated dependency relationships, hybridization of political structures, and dependency processes justifying the contemporary political requests. The Northern Territories was a “region of encounters” between different groups in history and, therefore, provides an opportunity for a close analysis of the ways in which different populations mutually contributed to shaping the contours of the socio-political environment and local perspectives of it. This diverse, rapidly changing environment—profoundly shaped, as it was, by British rule—illustrates the strong connections between conflicts over land, political representation, and territorial arrangements, which have undergone little change since the end of colonial rule (Herbst 1989; Asiwaju 1985).

The Konkomba and the space

The Konkomba are predominantly farmers and historically moved frequently within a broad expanse of the Oti River plain in search of fertile lands to cultivate or to escape from neighboring groups’ raids, particularly prior to colonial rule. During the colonial period, Konkomba communities frequently relocated to avoid the expansion of the nearby kingdoms (Dagomba, Nanumba, Gonja, Mamprusi, and Anufo). Konkomba predominate in the border area between Ghana and Togo, on both sides of the Oti River. This area has always been a periphery in history, on the margins of different pre-colonial centralised polities, of two different colonial powers, and of the postcolonial states. Konkomba are a good example of inhabitants of what Igor Kopytoff calls the “internal African frontier” (Kopytoff 1987). In fact, they are settled in the “periphery,” a space conceived as an interstice between different centralized kingdoms. From pre-colonial times to now, they have been in a condition of permanent definition and redefinition of spaces and spheres of influence, experiencing a constant negotiation with their neighbours, with colonial administrators and, today, with the postcolonial state. These negotiations have given rise to political transformation in a social space that has proven to be cyclical, dynamic and reversible, and in which the construction of identities and political structures is constantly discussed and debated.

Mobility, in such a context, is a fundamental characteristic that enables the group’s survival and capacity to sustain its social cohesion. While considering migration and relocation as central elements of the historical and social process of the region’s populations, Falola and Usman report on the case of the Lobi, who migrated to Burkina Faso from northwestern Ghana. Through such movement they re-created their society of origin elsewhere, while modifying it according to the new social, political and economic context (Falola & Usman 2009: 8-9).

Therefore, mobility represents, on the one side, a catalyst of identity constructions, and, on the other side, proves its key role in the development of networks between people, groups and the territory. Furthermore, Konkomba communities, who have above all migrated during the second half of the eighteenth century as a consequence of Dagomba occupations (who were themselves escaping pressure from neighboring Gonja), either left their villages or witnessed in many of their settlements the human component hybridization. One result was that Yendi (capital of the Dagbon) was developed on a Konkomba village called Yaa¹. The historical reconstruction of local relations during precolonial times included in the Kitāb Ghanjā provides documentary evidence to the fact that in the Islamic year 1125 (1713-14) the Gonja army entered Dagbon and pushed Dagomba eastwards, which in turn precipitated Dagomba invasions of Konkomba lands on the banks of the Oti River (Wilks, Levtzion & Haight 1968: 97).

The current relationship network the Konkomba communities maintain with shrines—ntingban in Likpakpaaln—that are today located in the Yendi territory confirm and constantly recall such historical transitions. In some specific period of the year, they visit the shrines with special requests, regarding for example the end of sudden and serious drought.² Konkomba migration eastward determined an “on the return” ritual relocation towards the west, the direction of their communities of origin.

According to Carola Lentz, indeed, the shrines have a central role for the High Volta basin settled groups, since they contribute to the construction of a shared ritual world, which is one of the essential conditions for mobility and mutual assimilation (Lentz 2000b). It is necessary, therefore, to make a few considerations on the different ways the local concept of territorial boundary can be embodied and lived. The “borders” of the territory inhabited by clans or family networks are not fixed and steady, instead they can move forward or recede. The dimensions of a clan can, indeed, be modified in time due to population growth or decrease (see also Tait 1953: 215).

It becomes clear that borders between clans and lineages are constantly bargained over and negotiated according to demographic and ecologic conditions. Therefore, rather than borders one might speak of “interaction zones,” since the territorial limits are constantly moving. All the same, the space has different physical and symbolic characteristics, depending on whether it is inhabited, farmed, or holds ritual significance. Discussing the nature of the borders of the North-Western Ghanaian villages, Lentz reminds us that in many agricultural societies the farmed fields’ borders are fixed and linear (Lentz 2005). Similarly, R. B. Bening added that “the northern Ghanaian concept of boundary is not a continuous line but a series of accepted points often widely separated by clearly defined or demarcated features” (Bening 1973:18). Bening’s characterization is equally valid for the Konkomba context. Their fields’ borders are often marked by trees, trails and rocks, as opposed to their villages whose borders are rarely linear to the advantage of a fluid system where limits are themselves constantly changing.

Where cultivable land is abundant, the core of ritual power is geographically well-defined by the presence of a shrine. It is commonly positioned in an uninhabited and regularly cultivated space, from which concentric circles of authority extend toward non-cultivated areas, or bush. The bush is an area of contact rather than separation. Boundaries between shrines are not imagined as linear boundaries, but, rather, as a series of match-points marked by hills, rivers, rocks, small lakes or trees (Zimoń 2003; Lentz 2005).

Konkomba who live in contemporary Ghana and Togo occupy settlements within administrative units called districts which, after the colonial period, underwent a territorial definition process that set fixed borders (Bening 1999). At a local level, though, border representation remained multifaceted and complex. While the concept of border dividing a district from another is accepted and embodied within the contemporary state administration, in the communities’ daily and ritual life such a fixed line loses its meaning. The borders that defined family and lineage space were not fixed. Even today, in relatively urbanized communities such as Saboba, the use of the territory depends upon the dimension and the necessity of a specific lineage, without recurring to physical landmarks. Very often, indeed, the lineages historically occupying a fairly big center with an important market, as Saboba for example, grant their ancestors’ land to families or lineages coming from small villages in search for better jobs and children’s education.

Ethnography and archival suggestions

J.K., hailing from Kukunzori and married with G.K., from Kuntuli, moved to Saboba right after their wedding to guarantee an education to his children. He told me: “I have my land to cultivate back in my village, Kukunzori. Here I only have my home. This is Bichabob’s land.”³

The Bichabob indeed, the largest Konkomba clan historically settled in Saboba, grant land to other clans to build houses through the figure of the earth priest (utindaan). Such debates on the use of space confirm, in the first place, the high degree of mobility within the territory and secondly, the unfamiliarity, within this cultural context, with the use of the concept of border as a fixed line in the social space. Through historically analyzing the establishment of the colonial districts, and the space allocated to the Konkomba in such a territorial redefinition phase, two important aspects of the relationships between Konkomba communities and space are evident. On the one side such analysis confirms the mobility and the malleability of the communities’ spatial characteristics, and on the other side it shows how marginalized their cultural representation of space were when the new colonial territorial layout was established. The Konkomba political and territorial system is, indeed, very far from the European cultural reference points, but it also turns out to be of little use to territorial management and poorly functional within the administrative partition.

In 1921 the Provincial Commissioner Branch communicated to the chief commissioner of the Northern Territories (CCNT) its plan for an internal line of demarcation between the Yendi and Gambaga districts. After setting a detailed description of the rivers and villages that such a line would have to follow, Branch wrote: “All this territory is under the Gushiogu [Gushiegu] and Djereponi [Chereponi] chiefs and the line follows landmarks… This line places the entire Konkomba territory under Yendi”.⁴

This is a rare example of a colonial official mentioning Konkomba with regard to territorial partitions, although the reason was their inclusion in the sphere of influence of other groups. The territorial definition system is expressed on two different levels: on the one side, the colonial administrative structure which sets linear borders and, on the other side, the moving perception of the territory which instead is characterized by exceptional fluidity and a scarcity of fixed demarcation symbols. These two levels can intersect, but they remain two different reference and interpretation systems, which produce in turn two distinct daily life interpretation models: the ritual, economic and social spheres express themselves fluidly, while the administrative and political spheres define themselves and operate through linear boundaries.

The colonial system of government produced a duality of schemes rooted precisely in the territorial partition through which new meanings are created. The Konkomba moved within such a dual world of signs and this is fundamental to understanding the modalities through which they articulate their relationship with the space, their sense of self, and their current political use of the territory. The British colonial project did not eradicate prevailing perceptions of space, which still maintain a central role. Nevertheless, it has undeniably introduced new reference points and new power orders.

The collaboration between colonial powers and organized kingdoms was fundamental in that phase. They, indeed, build a fruitful mutual relationship of convenience in the management of the territory, in particular in the so-called peripheries. Therefore, if on the one side the colonial administration needed the aid of the organized groups’ chiefs to manage the mobility, taxes and punishments of the so called stateless groups, on the other side the chiefs accepted to collaborate because they foresaw the possibility to exploit a new ally, that is the European administration, to include in their territorial spheres the “stateless” groups, which de facto occupied and ritually controlled large portions of land. As Michel Foucault reminds us, what needs to be constantly highlighted is that fact that neither simply a state, or a territory or a political structure is ruled, but groups, individuals, and collectivities. First and foremost to be ruled are people (Foucault 2005).

The Kingdoms’ Borders: Nanumba and Dagomba Territorial Definition

English and French colonial administrators established the management modalities of the new territories annexed after 1919 according to centralized groups’ structures, organizations and territorial dimensions. They wrongly considered other “stateless” groups, as the Konkomba, as both an integral part of the wide influence areas of the surrounding kingdoms and as peripheries on which those kingdoms’ sovereignty exerted an effective and widespread control.

As Wyatt MacGaffey states, it is correct to refer to those northern centralized kingdoms as “mobile unities,” ready to dilate, contract and move following the historical and political situation (MacGaffey 2010: 434). Christine Oppong has also provided a useful explanation of Dagomba territorial representation, defining Dagbon as a “cluster of mobile political unities.” In fact, she explains that, following the general succession rule, in order to achieve a higher status⁵ chiefs must be ready to assume different positions of power, moving frequently from one village to another together with their own entourage (Oppong 1967). Colonial administrators tried to divide the different kingdoms by mapping their own borders. In doing so, their European perceptions clashed with local concepts of space, mobility and authority, and British officials often found themselves—more or less consciously–involved in local power disputes. The analysis of those disputes, and the role played by the Europeans, give us the opportunity to consider the local figures of power as active players in the complex game of power and territorial change that was going on in their own territories.

Below, I present and discuss two different cases focused on the internal boundary definition on the Eastern part of the contemporary Northern Region of Ghana. I use them to illustrate the level of agency of the different actors involved in the process. Borders represent privileged angles from which one can analyze political transitions and identity constructions. At the same time, however, border definition processes must be deeply examined so as not to run into the common error of assuming that Africans have nothing to do with the formation of their own borders (both internal and international). Borders, in fact, are commonly considered the sole heritage of the European colonial agency. Nevertheless, as stated by Paul Nugent in his seminal work on the southern section of the Ghana-Togo border, we must not underestimate the fact that African borders had not always been arbitrarily traced, but they had often been the result of agreements between European interests and local necessities (Nugent 2002).

Case 1: Tagenemo Dispute, 1922-23

This first case study concerns the British colonial administration’s attempt to demarcate a border in 1922 between Dagbon and Nanun territories. After the British acquired the western section of the former German Togoland, which included Nanun and the eastern part of Dagbon, it worked to incorporate the territory into the Northern Territories’ administrative structure centered on indirect rule. Indeed, from 1886 to 1914, what is now the eastern part of Ghana and Togo was part of “German Togoland.” It was subsequently dismembered into two League of Nations mandated territories given respectively to Great Britain and France, the colonial powers that occupied Germany’s colonies during the First World War (see Austin 1963; Pyeatt 1988; Schuerkens 2001). While Britain imposed indirect rule in the new territory, it needed to clearly divide the space among different local powers, with the aim to better rule and control land and people. In the case of the village of Tagenemo, on the periphery between the two kingdoms, the colonial administration evidently could not clearly determine if the village was Dagomba or Nanumba. The simple act of searching for a shared solution gave rise to a true dispute, the analysis of which offers insight into local perceptions of border and periphery. It was very common, in fact, for villages in the periphery of kingdoms, as it is the case for Tagenemo, to have been mixed, inhabited by both Nanumba and Dagomba.

District Commissioner W. Gilbert tried to settle the issue by organizing meetings with community members, elders, and their respecttive paramount chiefs in Yendi and Bimbilla.⁶ The dispute appears to have moved toward a difficult resolution, as negotiations carried on throughout 1923. After his first meeting with both the Bimbilla’s and Yendi’s elders in attendance, the district commissioner gave his first opinion of what could have been the true situation. “There is no doubt that Tagenemo and the other small villages were under the Chief of Yendi for many years before the German Occupation, when they were given back to the Nanumbas.”⁷ Various interviews collected in the following months give us a more complex perspective of the political process experienced by this village, through the political activity and mediation of chiefs and elders. As we will see, the village was formally under Bimbilla, but still claimed by Nanumba and Dagomba authorities. As the Bimbilla Na stated: “The Chief of Bimbilla has always commanded Tagenemo. The boundaries have always been the same as they are now.”⁸

It is likely that the Bimbilla Na sought to leave the British out of an internal dispute, using the exact word (boundaries) that the colonial power could have understood and appreciate, even if, as we will see, in the local context it was extraneous and irrelevant. But at the same time the Ya Na was possibly seeing in the intrusion of the colonial administration a chance to modify the actual power balance; indeed he declared that the village had always been under the Nakwali Na, a Dagomba chief, until the changes imposed by the Germans. Amaru, chief of Tagenemo, declared that:

“I was a full grown man when the White men first came here and was under the Chief of Nakwali. The boundary between the Dagombas and the Nanumbas was about two hours south of here. My mother was a Nanumba and my father was a Dagomba and I was born at Yendi. When the White men first came they gave my village to the Chief of Bimbilla and made the boundary where it is now. I was appointed Head Man by the Chief of Bimbilla soon after the White men came. I was the first Head Man appointed to these villages. We were directly under the Chief of Nakwali.”⁹

The subsequent statement by an important elder of the village, the Nmar Na, provides greater context.

“I am a Nanumba, my parents were born here (Tagenemo) and I have always lived here. I was a man when the White men first came, we were under the Chief of Bimbilla. When I was a young man the Chief of Nakwali took us by force (no fighting) and we were under him until he was appointed Chief of Karaga when we were given back to Bimbilla.”¹⁰

This suggests that this village had historically been under different jurisdictions, both Dagomba and Nanumba. This was common for peripheral villages. They were mobile and under continuously changing authority. For the British administration, such a complex political dynamic was very difficult to understand. Indeed local authorities couldn’t refer to a status quo as a sufficient explanation to define whom the village belonged to, so much so they traced the variability of belonging through the history. In this sense, peripheries continuously moved; there were no fixed lines for British officials to easily mark.

Therefore, there were contradictions in British constructions of the border as a fixed line. Both the Ya Na and Bimbilla Na declared that the border “has always been the same.” The chief of Tagenemo stated that the boundary was “about two hours south of here” before the arrival of the Germans. Different accounts disclose diverse strategies on the part of local actors to negotiate British officials’ interests. Indeed, it is evident that all the actors involved worked to protect their interests, to gain something from the presence of a new authority in the area, namely the British colonial power. The Nmar Na supported the Bimbilla Na’s position, out of personal loyalty and a shared ancestry, while the chief of Tagenemo tried to remain neutral, but actually supported the Ya Na. At that point District Commissioner W. Gilbert went to the court of the Ya Na to learn the history of Tagenemo:

Before the Gonja war about the year 1648-77 Zangina Chief of Yendi appointed Kabinwari a Dagomba Head Man in Tagenemo… Some years after this Asumani Chief of Bimbilla with the assistance of the Gonja Chief of Kembi made war on Albarka Chief of Dakpam another Nanumba Chief. Albarka defeated them. Asumani then requested assistance from Na Mahama Chief of Yendi but was refused as he had had assistance from the Gonjas. Albarka went to live at Tagenemo and sent a messenger to Na Mahama to inform him what had happened. Na Mahama died and Kulunku was appointed Chief of Yendi; he sent for Albarka and at the same time sent a messenger to the Chief of Mampong and told him that he wished to make Albarka Chief of Bimbilla; he agreed to this. Albarka was appointed Chief and Kulunku sent many people to Bimbilla for the celebrations. Albarka gave the following villages to Kulunku: Tagenemo, Tanja… Buli, Gambuya, Mupiegu and Kayanga for the assistance he had received… in the war against the Gonjas.¹¹

The Ya Na (and later the Bimbilla Na) mentioned “the Gonja war” and the election of Na Zangina¹² as crucial starting points to explain “who” actually held authority over Tagenemo village. Following the Ya Na’s explanation, Tagenemo was part of a territorial disposal made by the Bimbilla Na to the Ya Na for political and military support. The Gonja wars, indeed, led to a profound destabilization of the local equilibrium and caused entire populations to relocate, the conquest of villages, and the emergence of new alliances (see Wilks, Levtzion & Haight, 1986).The Bimbilla Na’s version, however, was slightly different, or more precisely, it focused on different issues:

When Na Gungoble Chief of Yendi died… [T]he Dagombas could not agree who was to be the Chief so they all went to the Chief of Mamprussi who appointed Zangina the late Chief’s son. The Gonjas were annoyed at Zangina being appointed Chief and declared war on the Dagombas. Zangina was also unpopular with his own people, he refused to fight the Gonjas and went to live at Gbandi near Sansugu. Zangina then sent to the Chief of Bimbilla and told him he was unpopular with his own people and asked for his assistance, the Chief of Bimbilla sent to Kabinwari Head Man of Tagenemo, Head Man of Bukpali and Head Man of Langedi… There was no war between Asumani Chief of Bimbilla and Albarka Chief of Dakpam.¹³

The Bimbilla Na’s statement suggests the alleged Nanumba internal dispute was incorrect. However, it underlay a Dagomba internal division tied to the unpopularity and weakness of Ya Na Zangina himself and reflects an internal redefinition of power. The British administration proved to be completely unable to understand the situation, even while being fully involved in the dispute. The commissioner for the Southern Provinces declared that: “I found it impossible to reconcile the various statements made. I sat for 3 ½ hours and came to the conclusion that originally Tagenemo belonged to Bimbilla but that for some time the Gonjas of Salaga appointed the headmen… After a war with Gonja Tagenemo was taken over by Yendi.”¹⁴

As a consequence of their inability to understand the local power discourses, the British administrator took the shorter road. The commissioner said:

Before leaving Yendi I saw the Head chief and asked him to use his personal influence to settle the dispute. I also pointed out that he had a big following in comparison with Bimbilla and that the Government did not wish to create ill feeling between the two tribes. From the hints given to Mr. Gilbert and myself I am sanguine that should Bimbilla acknowledge Yendi’s right to the villages by conquest. Yendi will hand them back.¹⁵

The British decided to lean on the Ya Na, who was in that period the better ally of the colonial power, to settle the issue for them, while (consciously) ignoring the fact that the Ya Na was directly involved in the dispute. Ya Na Abdulai, from his side, had gained an advantage once again from his forward looking alliance with British colonial power that had started in 1917.¹⁶ A document signed the following year describes the agreement between Bimbilla and Yendi concerning the Tagenemo village. It showed how much the final decision favored the Ya Na and completely undermined the version of the events collected by Gilbert. “The Head Chief of Yendi is to command Tagenemo and the small villages concerned, and he is to elect Head Men… The Head Man of Tagenemo may go to Bimbilla to salute the Chief.”¹⁷

Case 2: Kulmogba River Boundary Definition, 1935

The second case, reported in 1935, demonstrates the local consequences of the British imposition of fixed borders. By 1935, the British had already implemented their vision for borders in much of the Northern Territories. At that time, British officials took it for granted that there was a fixed border between Nanumba and Dagomba territories, even though the contradictions were evident. The fact that it was necessary to move people from one side of the border to the other, in this case the river, illustrates the imposed nature of the border and the ways in which different chiefs used it to negotiate their own power positions with the British administration. At the same time, it is possible to analyze how the imposition of fixed borders profoundly influenced the immediate and future relationships among different groups. Even those that historically shared the same territory and lived in a condition of true interconnection, were forced to statically belong to a portion of land, confined to their respective “ethnic” categories.

An agreement signed by Ya Na Abdulai and Bimbilla Na Abdulai on the 6^th of July, 1935, offers an example:

The river lying between Dagomba and Nanumba known as Kulmogba is our boundary dividing the two states. All villages and inhabitants with their belongings on the bank of the river on Korle sub-division belongs to the Ya Na and all such inhabitants should render their respective services to Kworle-Na and to take all lawful orders, or instructions from Kworle-Na when required to do so by the Ya-Na in accordance with our custom. Nating-Lana and Gob-Na sub chiefs of Natinga and Gob villages under the jurisdiction of Bimbilla and having their present domiciles at Bagmani a village under Kworle subdivision should remove with all their belongings from the said village and go to settle in their own villages to which they are appointed as chiefs in the Nanumba area.¹⁸

One consequence of the British imposing a clearly defined, static boundary was that people and chiefs had to relocate. The fluid character of the border, as conceived locally, was here radically undermined, in favor of colonial administrative necessities. What does it mean in terms of relationship with the space and in terms of belonging? Settlement in a particular territory meant, from that moment, belonging to a certain group, which linked territory and identity (see Lentz & Kuba 2006). On the right bank, they must be all Nanumba, and on the left all Dagomba. Such an “ethnic map,” drawn by the colonial administrators, did not match the mobility and mixed composition of the villages, especially the peripheral ones. The creation of “ethnic borders” by the European powers throughout Africa, and the conception of those borders as culturally and politically homogenous, fostered an experiment fraught with consequences and conflict, with ramifications for the present, in many contexts (Lentz & Kuba 2006).

Furthermore, local equilibrium had to be considered. District Commissioner Cockey wrote, “The decision is more in favor of Dagomba than I expected, but I have visited the Bimbilla Na subsequently and feel sure he is perfectly satisfied with the decision, and that he and his people were not forced into an agreement which they did not like.”¹⁹ The following excerpt, however, stated the contrary, while the administrator seemed to be not interested in considering this evidence.

I have just received a complaint from Bimbilla Na that the Kworle Na has taken advantage of the settlement of the dispute in his favor, to extend his tentacles still further afield and claim jurisdiction over Chichagi, an undoubtedly Nanumba village… I think it will be advisable to visit and perhaps demarcate the boundary as laid down in the agreement … In his complaint, Bimbilla Na remarks that the decision of Ya Na was not accepted, but there can be no doubt that it was accepted.²⁰

Evidently, the British colonial administration introduced itself into a local dispute over territorial authority of the villages along the river. Yet in spite of the imposition of the colonial model, the persistence of the local uses of the border—again, conceived of as fluid, mobile and flexible—is evident. The Kworle Na used the imposition made by the colonial power for his own benefit.

Conclusions

From the documents discussed here, it appears clear that the district commissioner and his collaborators had little understanding of relations between local authorities. Although the colonial authority imposed its administrative and territorial authority, it was mediated and contested. The centralized kingdoms (Dagomba and Nanumba) as opposed to decentralized societies (Konkomba), were not passively included or excluded from the creation of a territorial map. Local authorities, in the period under consideration, exploited each available advantage derived from alliance with the British in the region. The British administration’s creation of “ethnic borders” enclosed a presumed cultural and political homogeneity. This turned out to be an experiment with paradoxical consequences, strongly reverberating throughout time. If, on the one side, the boundaries’ layouts produced conflicts between kingdoms, on the other side they forcibly included the so called stateless populations, such as the Konkomba, who suffered the systematic exclusion from the colonial “ethnic map”. Colonial violence expressed through territorial partition was a political instrument that accentuated imbalances of power. The British exploited local rivalries and defined relations of dependency as static instead of fluid. The act of narrating the local concept of boundary in this context acquires meaning only by exploring all different local experiences, which in turn express themselves in the continuous encounter between different local realities and actors. As stated by Cohen, indeed, boundaries are largely constituted by people in interaction (Cohen in Hagberg & Tengan 2000: 15). Territorial borders become here ethnic borders: that is why one should always refer to the network approach (Howard & Shain 2005), with the aim at maintaining a fluid perspective and promoting the idea of ongoing process, that is inherent in the local concept of boundary itself.