The Hidden Factor: Experiences of Statehood and Ethnicity in the Long-Term Perspective
Comparing the evolution of forms of group solidarities over the long-term perspective is a complex project. Nevertheless, the case studies discussed in this study allow me to formulate a number of more general conclusions. Before bringing these conclusions into a larger panorama, and engaging in the broader discussion of their implications, I will first recapitulate some results with a focus on a particular point of culmination, which has become evident in our empirical cases: the start of the last phase of reform under late colonial rule, just before decolonisation. As we have already seen in the different case studies, the years between roughly, 1942 and 1957 represented a particular challenge to established community patterns.
In different communities of coastal West Africa, they constituted a phase of aggressive political claims, which in turn produced challenges to established local or sub-regional systems of ‘traditional government’ under colonial rule. In the historiography of sub-Saharan Africa, this phenomenon has only partially been commented on. Normally, in studies on Africa’s decolonisation, local violence has been interpreted as a reaction to a modernising world, in which new forms of political expression had become available, and in which the representatives of the Old Regime – the chiefs, frequently regarded as a product of colonial rule – were increasingly understood as a relic that was to be removed as quickly as possible. Other scholars have insistently referred to waves of discontent in the wake of the decolonisation process, which targeted, through protest against the chiefs, the very structures of European domination.
In both variants of interpretations, which are often intertwined, the focus lay very strongly on issues of ‘modernisation’ and ‘liberation’. It has apparently been impossible for some time to link the protests of the late colonial period to phenomena of ethnic solidarity, and to speak of ethnic mobilisation. Of course, there were those new political parties that, in the literature, were occasionally categorised as ‘tribalist’ – but such parties were, seemingly inevitably, doomed to fail, while the westernised movements demanding a nation state appeared to win the day.1
My analysis of processes in the selected West African coastal areas in the 1940s and 1950s, has given us a distinct picture. Local populations resorted, at certain periods very strongly, to ethnic formulae; these strategies appeared in all case studies, and repeatedly. They were employed in an attempt to carve out opposition against unpopular chiefs, who were now defined by the protesters as ‘strangers’, and as ‘usurpers’ of posts for which they had no ‘traditional’ claim. At the same time, ethnic argumentations made the rounds in the struggle for attractive posts at the heart of reformed administrations, and in newly created territorial governments. These posts were new assets, and, as such, they frequently enjoyed considerable prestige.
We have seen that the Wolof-speakers of the Petite Côte and beyond had previously been mostly passive in their employment of ethnic formulae. In the pre-colonial period, but also under colonial rule, the larger category of ‘Wolof’ was something that existed only at the margins of the community. If it was mentioned at all, it remained an obsession of European visitors and residents, and, in particular, of the colonial officials organising the new local administration in the second half of the nineteenth century and in the twentieth century. The ‘Wolof’ would appear as a category in administrative reports and statistics. However, the practical importance of this category was extremely limited: very few cases of a particular Wolof mobilisation and solidarity existed during this period. Also, no signs can be found that the colonial engagement in categorisation triggered a more frequent self-definition of local Wolof-speakers under the ethnic label of ‘Wolof’. Where groups of this category had been drawn into conflicts, local leaders gladly used European anthropological work, but there was no push from the side of the colonial subjects to employ the category more regularly.
On the contrary, the different groups of Sereer-speaking neighbours of the Wolophone populations challenged Wolof-speakers, who they held as a distinct ethnic group. The respective mobilisation processes were strongest in the two phases of extreme sociocultural change in coastal Senegal, that is, in the second half of the nineteenth century during colonial conquest, and in the 1940s and early 1950s, during colonial reform and territorial democratisation. On both occasions, Sereer-speakers reacted to a new situation. They attacked the ‘Wolof dominance’ (i.e. the dominance of Wolof-speaking rulers and dignitaries) in the pre-colonial political units, and their supremacy in structures of
The Temne of northern Sierra Leone also mobilised in the 1950s against a group of local ‘traditional’ rulers, but the context of this mobilisation was entirely different from the Senegalese experience. Temne-speakers had actively employed the label of ‘Temne’ during all of the conflicts of the nineteenth century. In difficult political situations, that is under the pressure of neighbouring ethnolinguistic groups, they had been engaged in creating group cohesion through the employment of ethnic arguments. While the context in which we can make the use of these pre-colonial slogans visible, is limited to the ways in which Temne-speaking leaders reported upon mobilisation strategies in discussions with the future colonisers, it is, nevertheless, quite evident that these ethnic solidarities were employed during the nineteenth century. The use of the ethnic label was subject to fluctuations according to conditions on the ground, and was less frequent in some particular periods. Even so, it remained a constant factor for self-identification and mobilisation beyond village and town communities.
Only after 1900 did ethnic solidarity among Temne-speakers become less prominent and less aggressive – and only in the 1950s, considerably later than was the case for similar processes in the southern parts of Sierra Leone, did Temne-speakers again engage in widespread mobilisation under ethnic labels. After 1960 (but not before), these mobilisation strategies became increasingly dominant in Sierra Leone’s political arena. Unlike in Senegambia, where the multi-party system introduced in the mid-1940s was from the outset linked, as in the case of the Sereer-speakers, to ethnic sentiment, Sierra Leone’s United Progressive Party and its successor, the All People’s Congress, were only slowly
The experience of the Ewe-speakers is different again. In the late 1940s, Ewe-speaking leaders managed to create a spectacular political movement, which had an impact even at the international level, through the appearance of these leaders on the stage of the United Nations. However, the temporary success of organisations like the All-Ewe Conference stood in contrast to the visible ups-and-downs of Ewe ethnic mobilisation during the second half of the nineteenth and the whole of the twentieth century. Ewe solidarity had experienced its first peak during the 1860s and 1870s, mainly caused by the inability of different Ewe-speaking community leaders to organise an efficient defensive coalition against massive Asante and Akwamu raids. The incursions of the armies from neighbouring regions, under these circumstances, not only had a traumatising effect on the local societies, but they also led to the employment of ethnic solidarity as a ‘last resort’ in order to stop the pillaging groups. As is evident from the many ‘traditional histories’ written between 1880 and 1930, this experience had a lasting impact on the collective memory of the members of Ewe-speaking communities.
However, the German and British occupations of the area east of the Volta River led to the rapid demise of ethnic principles in the region: in spite of the promotion of ‘tribal solidarity’ through missionaries and part of the colonial administration, the local populations focused (or probably refocused) on the political unit that the British would in the future call ‘division’. In the ways in which locals ‘sold’ their larger group identification to the colonial powers, in the form of the above-mentioned local histories, they would still refer to the Ewe solidarity that had helped them during the Asante incursions, but we find no tendency to overstate these issues. In other arenas, such as land claims, succession claims, and other legal issues, the locals were quite content to relate to their ‘sub-regional’ identification, that is, at ‘division’ level. The British administration was frustrated with this tendency of local populations to ignore ‘tribal’ bonds, and some ambitious leaders of the pre-colonial states that now stood under European rule, like Anlo’s Awoame Fia, remained active in justifying an enlargement of their authority through ethnic arguments. Normally, however, Ewe identification did not play a prominent role in political issues during the interwar period.
It was only from the early 1940s, when a group of British-educated and French-educated middlemen discovered the Ewe issue as a vehicle for political mobilisation on a broader, international stage, that the problem of the larger ethnic identification of ‘the Ewe’ again became a point of intense debate. For some years, local leaders and their opponents attempted to jump on this bandwagon
I will now attempt to give an explanation of the tumultuous processes of the years after 1945 in the long-term perspective in view of my empirical results. In three coastal West African areas, we find similarities with regard to challenges facing the ‘traditional structures’ in the post-Second World War period. However, the reactions and types of community mobilisation described were sometimes extremely dissimilar. How can we explain the diverging degrees of recourse, by the different communities, to ethnic mobilisation? In the initial parts of this study, I have pointed out that one technique with which to analyse differences in the forms of local identities is to follow them over the long-term process. In this book, I have limited the perspective to the period between initial stronger European engagement in the respective West African regions (around 1850) and decolonisation (around 1960). If one looks at the numerous examples, in which local spokesmen of groups ‘sold’ their identifications to colonial officials, in direct conversation, during land and succession conflicts, in local histories put together for very different reasons, and regarding alliances between communities that had been forged over the years and even under colonial rule, we encounter a remarkable picture: the variable that defined the recourse to ethnicity, seems to have been ‘the state’ as a political unit – or, to describe this phenomenon in more precise terms, the local and regional ‘experience of statehood’. Both in its pre-colonial and in its colonial meaning, the presence of larger state structures gave a political answer to the frequent problems of the nineteenth and twentieth century, while the absence of such entities deprived those groups of alternative means for creating cohesive communities. Where entities of statehood existed, communities ultimately had no need to rely on ethnicity in moments of insecurity and external pressures.2
As we have seen, among the Wolophone populations of the Petite Côte and beyond, ethnic identifications did not play a central role for the communities. While their neighbours, in particular those belonging to Sereer-speaking groups, frequently formulated their ‘Sereer identity’, Wolof-speaking individuals, in contact situations with the colonial authorities – but also in their direct interactions with Sereer-speaking inhabitants – insisted on their identification as members of one of the pre-colonial states. They identified as being a subject of the Damel of Kajoor, of the Teeñ of Bawol, of the Brak of Waalo, the Buur Siin or the Buur Saluum. This seems to have been far more important for the local definition of individual and group identifications than any overarching Wolof solidarity.
It has to be emphasised that the experiences Wolophone populations had within these pre-colonial states were mixed. The basic mechanism of all larger political units in Senegambia was the extraction of tribute and of contributions: to arrive at these goals, the rulers of the political entities relied on the ceddo system, in which specialised warriors obtained the desired contributions, sometimes through coercion and intimidation, sometimes through outright plunder. Wolophone village communities were subject to regular visits of envoys with ceddo troops, and suffered under this irregular system of tribute-raising. In contrast, in other areas, the pre-colonial state provided a certain platform to rely on, as the administrative units and troops guaranteed a relative degree of protection, and the scenario of conflicts between pre-colonial rulers predisposed certain arrangements in the case of conquest, although the sale of captives as slaves remained a reality until the end of the nineteenth century.
For the Wolof, the ‘statehood experience’ eclipsed the concept of ethnic solidarity. Furthermore, the widespread success of Islam in all the Senegalese regions peopled by Wolof-speakers combined in part with this trend. On the one hand, it offered a potential second path through which to formulate a non-ethnic group identification, which was useful in situations of conflict and rivalled ethnic mobilisation. On the other hand, however, many representatives of local dynasties became engaged supporters of the Muridiyya in the interwar period. Membership of the brotherhood gave them an additional legitimacy.
After the First World War, the direct link to the – now largely disappeared – states became weaker in the memory of Wolof-speaking populations. The pre-colonial entities were, in any case, less frequently mentioned in the traditions through which locals ‘sold’ the characteristics of their group to the agents of the colonial power. Nonetheless, this was not a massive change. In particular, individuals among the paramount chiefs who could claim to be in a family line to the former ruling dynasties continued to enjoy considerable prestige. Therefore, it would be correct to say that the memory of relations with the pre-colonial
An interesting and highly significant alternative path is the experience of Wolophone populations in the territory of what later became the British colony of the Gambia. Here, local Wolof-speaking populations, like their neighbours of other linguistic communities, could not rely on similarly stable pre-colonial state structures. They thus did not have a positive ‘statehood experience’. In the area of the Gambia River, the existing unstable units deteriorated further after the Jihads led by rulers of political entities. In this overall panorama, Gambian Wolof-speakers were far more eager to claim solidarity on ethnic terms than were Wolophone populations further northwards. The situation was similar in Casamance, where Wolophone immigrants lived under instability and threats of violence, and were more eager to insist on their ethnic identification. It is possible to consider this factor of an ethnic counter-mobilisation of ‘Wolofs’ as a principal motive for the Casamance revolt of 1982.
On the Petite Côte – to come back to the region discussed as the case study – the emphasis of the Wolophone elite on state identification instead of ethnic identification created a possible way out of the political struggles of the 1940s and 1950s. The setting allowed opposed chiefs, such as Ely Manel N’Diaye, to ‘disarm’ opposition movements that were working with ethnic claims. The Wolophone paramount chiefs could always claim to represent a ‘tradition’ that had nothing to do with ethnic solidarities; they were the heirs of political structures of authority, and they could claim to enjoy this authority also, at least in part, with the Sereer-speaking communities. The success of this strategy explains, more than anything else, the surprising downturn in ethnically motivated mobilisation in this coastal region around 1951/52; it allowed a number of heavily attacked local leaders to survive politically.
As far as the effects of pre-colonial ‘statehood’ are concerned, the region of northern Sierra Leone was entirely different from our examples from Senegambia. Among Temne-speakers, we had to consider a notable lack of any stronger political institutions. Port Loko, the main centre of the region, was merely a sort of small ‘city-state’. The latter expression could be used for a town controlling a part of its surroundings, with little impact on the life of the inhabitants of the neighbouring villages, and only over certain periods of the nineteenth century. Other town centres, such as Kambia, Magbele or Marampa, were similarly organised, and their rulers also had little impact in controlling any broader regions. The Alkali of Port Loko tried at least to claim a regional paramountcy in the second half of the nineteenth century, but he failed with these plans. He and his fellow Temne-speaking rulers did not find a formula of alliance, nor indeed any political overlordship of larger population groups.
Unlike the rulers of pre-colonial states in Senegambia, the leaders of ‘city-state’ communities in northern Sierra Leone were usually unable even to claim regular tributes from a greater number of settlements in the region. Similarly, they were never in a position to give protection to the populations of broader regions. This situation became obvious in the nineteenth century during the different waves of incursions of Susu-speaking troops, or during the Yoni War – occasions during which the Temne-speaking rulers showed to the fullest their inability to limit the effects of warfare. Under these circumstances, the only means of the rulers of the larger Temne-speaking settlements to affect the lives of populations in their vicinity was through plunder and small skirmishes. Campaigns of this type were indeed carried out regularly.3 They added to the general feeling of insecurity in the region, and created, for the local populations, a ‘statehood experience’ that was entirely negative – at least if one holds that the local political units had any of the characteristics of ‘states’.
In the first half of the nineteenth century, but probably continuing a situation from the eighteenth century, Temne-speaking communities were situated at the margins of two larger pre-colonial entities: the federation of Morea, an alliance dominated by Mandinka-speaking rulers, and the Almamis of Timbo, in the Fuuta Jallon. However, the latter had only been interested in control of the caravan trade routes to the coast, although they remained the object of widespread fear. The Almami of Forékaria, leader of the federation of Morea, was unable to discipline even the chiefs in his closest surroundings, and from
The absence of state structures in the region of northern Sierra Leone made wars smaller as regards scale. However, it is also evident that these wars were far more destructive than the average Senegambian conflict. Rules of warfare were less reliable, if they existed at all, and Temne-speaking villagers could not turn to any entity or institution that would have intervened on their behalf. In the 1830s, the ongoing presence of such destructive warfare brought widespread despair even at the level of local rulers, and led to the invitation to British envoys from Freetown to become intermediaries in the regional conflicts. The type of warfare, conditioned by political fragmentation of the area, also explains with much plausibility the constant employment of an ethnic argumentation. Whenever settlements were menaced by a threat that could be classified as non-Temne, the ethnic card was a last resort to concentrate forces against the attackers.
The presence of Limba-speaking, Loko-speaking, and, above all, Susu-speaking war bands in the region created the motive to resort to these strategies of ethnic solidarity. During the decades of British diplomatic activity in future Sierra Leone over the whole of the nineteenth century, the diplomats from Freetown were again and again confronted with the criterion of Temne-ness used in a number of contexts. Mainly, this criterion appeared as an argument to forge alliances between local Temne-speaking rulers and communities which, given the lack of reliable state structures, needed these argumentations as a sort of last resort whenever they were threatened with complete destruction.
In this logic, it is not surprising that with the imposition of the British protectorate, the eagerness of locals (and local elites) to emphasise their ‘Temne identity’ seems to have declined or even evaporated. The label was no longer really needed. As, after the Hut Tax War, destructive warfare in the formerly typical style was no longer possible, local populations could now turn to less existential problems. This is not to say that British rule over the Sierra Leone Protectorate was benign: as I have pointed out, among other abuses, British officials maintained a system of chieftaincy that allowed for the use of forced labour. In particular, farms and buildings of chiefs were in many cases maintained by the locals without receiving any payment. However, in chieftaincy issues including conflicts for power, there was at the beginning, in the new patterns set by the colonial state, no need to emphasise ethnic configurations. The whole period between the start of the century and 1945 was rather ‘peaceful’
The view became quite different, however, as soon as populations of Sierra Leone’s north started to believe that a government dominated by a distinct group in Freetown was attempting to withdraw from them the state guarantees of protection (and, for the late colonial period, the roughly even distribution of resources) from which they had profited under the colonial state. This general feeling led to a process of counter-mobilisation, starting at the local level, but slowly becoming a larger phenomenon. The mobilisation campaign went on for years, but, in the end, the sentiment of marginalisation in a renegotiated state laid the ground for a renaissance of ethnic Temne sentiment. The argumentation remobilised an old formula. The idea that only the reliance on ethnic solidarity would save the Temne-speakers from a ‘dictatorship’ of the ‘Mende southerners’ had major importance for voting behaviour in post-colonial Sierra Leone, where ethnicity remained the dominant cleavage. Ethnic voting retained its role at least until the 1990s.
In the panorama of results, I have thus pointed out Senegambia as an example of a region whose trajectory was dominated by fairly strong pre-colonial political units (states), which offered an alternative means of mobilisation to ethnic solidarity, while northern Sierra Leone can be considered as a zone where the absence of any larger political entities made the fixation on ethnic mobilisation sometimes almost inevitable. My third case, the Volta River region, can be considered as a case between those two extremes. In the present-day border region between Ghana and Togo, some of the political units were indeed stronger than those of pre-colonial Sierra Leone: Peki in the hinterland of the coast, and Anlo at the Keta Lagoon, developed the most visible activities towards political centralisation during the nineteenth century. The presence of extremely coherent and militarily successful pre-colonial states in their neighbourhood played an essential catalytic role in these processes: Peki had a political and military organisation modelled on Asante and Akwamu, and Anlo had, at the moment of British conquest, at least partly developed in the same direction.
After colonial penetration into the area, the British as colonial rulers further strengthened these solidarities, by transforming both Anlo and Peki into so-called ‘native states’. This decision preserved a part of the authority of the respective two leaders. It is thus no surprise that during political unrest in Keta and Anloga, particularly during the 1940s, the different sides involved all claimed that they had the prosperous future of Anlo State in mind. Although
This played all the greater a role since in other parts of the region, mostly under German rule from 1884 (or at the latest 1890), the impact of pre-colonial statehood, or ‘statehood experience’, was entirely negative. In the 1860s, many Ewe-speakers of the Trans-Volta area had learned a bitter lesson about the limits of state power, in the event of destructive raids from outside. Faced with an Asante and Akwamu invasion, Peki was literally swept away. The Awoame Fia, in contrast, sided with the invaders, which left the rulers of Ewe-speaking communities of the region with an even more significant impression. In this particular situation, it appeared necessary to have recourse to ethnic rallying cries. It was only by invoking a joint Ewe identification – probably also by insisting upon the origin myths of common migration – that the Pekihene and other leaders managed to forge a durable alliance. Although dramatically inferior in its military means in comparison to the Akwamu and Asante raiders, this alliance allowed them to at least organise a more or less effective resistance in the hilly reaches east of the Volta River.
The memory of this cooperation remained a factor within the political situation of the region. Such a conclusion is obvious from the ways in which the events of the 1860s and 1870s, often combined with reflections on the Taviefe War of 1888, continuously appear in the historical ‘tradition’ recounted by members of the most different communities. However, in its practical dimensions, the effects of this ethnic identification were largely dormant until 1945. We do not have any convincing proof for D.E.K. Amenumey’s claim, which insinuated that the Awoame Fia of Anlo and other Ewe-speaking rulers repeatedly referred to this common ethnic heritage before the Second World War. This problem can today no longer be approached through oral interviews, and the written sources (including the transcripts of, and notes on, the ‘tradition’ of local communities written down in the interwar period) suggest that Ewe solidarity was not a reality in the interwar period. By contrast, in the 1940s, Ewe-ness was rediscovered as a successful means of mobilisation. In the regional logic, the reliance on ethnic Ewe mobilisation seemed first – and obviously not successfully – to be a means to obtain the control of an independent territory, although the plans of many of the leaders of the pan-Ewe movement were not very precise. During the 1940s and 1950s, Ewe-ness would then, however, also
The Colonial State: A Limited Impact
The findings of this study reverse, for the examples from coastal West Africa, a number of dominant views on processes of ethnic mobilisation. This is important as conventional wisdom on ethnicity in sub-Saharan Africa still holds that the role of the colonial administration had an immense impact in the process of transforming ethnic solidarity into aggressive and even genocidal activities. While no historian, and probably no anthropologist, would today still defend the idea that colonial officials and European missionaries alone ‘invented’ the existence of the different ‘ethnic groups’, and then transposed these categories upon passive and manipulated populations, it is still a common view that the colonial methods of administration played their part in triggering the clearer bond to ethnic elements in the idenfications of African individuals and African groups. My comparative discussion does not challenge the importance of European rule for the reconstitution of local solidarities and communities. However, it provides a different interpretation of processes and chronologies.
This study does not defy the view that colonial officials of all European systems of rule, influenced by the ideas of racial and national difference that were en vogue between 1875 and 1945 in the respective metropoles, were largely convinced that it was most appropriate to categorise colonial populations in ‘tribal’ ensembles. Nevertheless, there is little evidence in coastal West Africa that they ever implemented these views practically, nor that they forced them upon the minds of passive African recipients. In many cases, the structures supported by the agents of the colonial administration already had their reference points in pre-colonial political units. The administrators might have weakened the political prerogatives of these entities, they might occasionally have made their rulers appear ridiculous, or as mere instruments of European wishes, particularly when it came to tax enforcement or the organisation of forced labour for the colonial state. However, colonial engagement did not automatically cut the connection to previous ‘statehood experience’. The experience populations had with these political entities continued to matter.
Northern Sierra Leonean chieftaincies reproduced, if anything, the rule of established dynasties from the nineteenth century; in the Gambia, the effects were weaker, but in their basic mechanisms similar. Local populations had in both regions used ethnic arguments for mobilisation before, but these
A glance at comparable cases in coastal West Africa seems to confirm these findings – although it would need to be pursued further through analytical research. In the Gold Coast, the British were not very active in encouraging the creation of ethnic identifications in the Colony (the southern part of current-day Ghana) and the Ashanti Protectorate. While they did not block the emergence of a Fante sentiment among coastal populations, the appearance of this sentiment was nonetheless a pre-colonial process of the second half of the nineteenth century, and it mainly transformed a political federation and the traumatic experience of Asante pressure into a formulation of group identifications. Asante as an important ‘traditional authority’ was maintained under British rule, but, then, this British engagement resulted in the preservation of a political structure, where ‘statehood experience’ was in spite of all existing opposition broadly shared, and where it had strong, participatory elements.
In Côte d’Ivoire, the French also did very little to formalise any ethnic rule in the case of the larger language groups, such as the ‘Baule’ or the ‘Agni’. The chefferie supérieure of the Baule was a rather recent construct, and it had not been created as a result of the initiative of colonial administrators. It took a long time to transform ‘the Baule’ into a dominant and politically active group, and this only happened in the face of the political tensions of the post-colonial period.4 In the case of Nigeria’s Yoruba-speaking groups, the pattern was similar. Although the emergence of an autonomous Yoruba literature was an earlier phenomenon which allowed an identity-building process to be pushed forward through a ‘Yoruba historiography’, this was a process not reflected in conscious provincial administration. For this administration, British officials still relied on political and state structures (as they did in the north of the colony with the Sokoto Caliphate and Bornu).
In the case of the future South-western Region of Nigeria, British colonial planning meant a practice in which the authorities of the leading city-states like Oyo or Ibadan still retained a leading role. The role of these authorities survived for a time, in spite of the massive socio-economic transformations triggered by internal migration movements.5 These led only in the end to new challenges and conflicts in which the power of these authorities was subject to sometimes violent contestation, of which the first electoral campaigns of the 1950s were the beginning. Moreover, the ethnic variable could be activated in the repertoire of many locals as a means both in internal struggles and in the defensive position against groups and populations coming in from other parts of Nigeria.6 This, again, was not a colonial activity. It was due to a process of change from the late colonial period, which relied on principles of cultural homogenisation that had eagerly been absorbed by local elites.
The patterns of behaviour as related to ethnic identifications and the state, as I have proposed them here, also give an explanation for the violent
However, it is obvious that these hopes were quickly disappointed. On the European side, the erection of a system of ‘native tax’ was regarded as one of the vital and necessary parts of colonial rule. The motivations for the creation of exploitative tax systems, and even for the tax obsession that some European colonial administrators would show, were manifold: attempts to make the territories lucrative for the metropole, and racist preconceptions that saw the imposition of constraint to force Africans to work as inevitable, both led to these tax decisions. The impact of the relatively high tax levels was evidently felt in many parts of West Africa, and led to resistance. However, regions in which local populations had not been confronted with previous systems of taxation, and where, normally, tribute was demanded on an irregular basis by oppressive outsiders, became geographical centres of revolt and tax evasion.
Most of the themes in African history that have gained great prominence over the last twenty years have been contributed to by cultural studies: syncretist religion, the impact of gender, and questions of generational relations, for instance, are all subjects that have greatly broadened our view of African societies. However, if we wish to understand decisions of groups to choose identifications, and to show solidarity in given moments, we have to bring the experience of ‘statehood’ back into the picture. In a way, this conclusion also strengthens the case for a new political history of African populations – but one that analyses the agency of locals choosing their form of mobilisation and solidarity in the face of manifold challenges, including the conditions of colonial conquest.
West African Ethnicity and Global History
The comparison of the three coastal West African examples, in view of their historical dimension, suggests a model of explanation for ethnic mobilisation in a long-term perspective. The influence of pre-colonial states, of a ‘statehood experience’ (even if in still rudimentary forms), and the ways in which state structures were taken as a point of reference, is quite important. Where such structures were at work, they normally eclipsed the ethnic argument. In other words, where stronger state structures existed, the recourse to ethnic mobilisation was not normally regarded as a necessary strategy. By contrast, in the absence of reliable state structures, locals eventually turned to an insistence on ethnic solidarity to reach their goals.7
What could these results mean for the debate on global history? Would it be possible, as Toyin Falola has claimed, for the experiments with nationhood in sub-Saharan Africa after independence to put ethnic mobilisation in a global perspective?8 As I have pointed out in the first two chapters of this book, many of the categories of group cohesion and mobilisation – ‘ethnicity’, ‘nation’,
I suggested in the discussion on the use of categories in a global perspective that ‘ethnicity’ was artificially reserved for sub-Saharan Africa, ‘native America’, and ‘tribal Asia’, as if these world regions had been characterised by an entirely different type of group organisation. Obviously, the combination of colonial domination and repression, and the belief of the European colonisers that they had to classify local subjects according to racialised and negative stereotypes, had an impact on group organisation. However, it is dangerous to see in the importance of ethnic labels the import of such labels from European invention. Global history has notably made it possible to approach the agencies in entangled processes such as European colonialism (and beyond Eurocentric interpretations), but at the same time to take into account the fact that European expansion changed the perspectives of this agency.
In the decades before colonial conquest, and in the conquest phase itself, many parts of West Africa were characterised by massive insecurity for communities. This insecurity was at least partly provoked by European activities on Africa’s coasts: the European demand for slaves to be transported into the Americas fuelled violent conflicts between communities; the change in export patterns from ‘human merchandise’ to ‘legal’ export crops intensified the use of slave labour in some agrarian zones, and therefore slave-raiding; in the conquest phase, European residents became the allies of local rulers who wished to increase their regional position and who used this support to attack neighbouring communities. The presence of European residents also had an impact in eighteenth- and nineteenth-century South and South-east Asia, where European participation in local conflicts in some cases changed the balance and ultimately intensified violence. However, the pressure on communities where state structures were relatively weak, was probably strongest in the context of West and West Central Africa, given its involvement in the late period of the slave trade and subsequent systems of enslavement.
After the establishment of colonial structures, these pressures diminished. Clearly, the colonial states – in sub-Saharan Africa, but also with regard to several communities in South and South-east Asia – had their own impact in creating ‘traditions’ according to European imaginations of either useful or ‘authentic’ rule. However, the period between 1918 and 1945 may have been characterised by abuses and the creation of systems of patronage in rural areas, but not as much by ethnic mobilisation.9 The problems of exploitation under
The late colonial period after 1945 has, for many contexts, been described as a phase of attempted modernisation and of experimentation with enlarged democratic structures. Interestingly – at least in the context of sub-Saharan Africa – the evolution of the colonial systems after the Second World War seemed to create a massive feeling of instability in the regional zones, far from the processes of centralised politics in the capital cities of the late colonies. As I have argued, in the case of many African regions, this situation can be regarded as the starting point of processes – processes that, after the end of the colonial period, led to a preference of local populations for ethnic instead of other forms of mobilisation. It is for the global historian, of course, a challenging question to ask whether other parts of the colonial world knew a different path to the creation of post-colonial societies. A principal difference might lie in the existence of stronger pre-colonial administrative traditions in large parts of South and East Asia, so that the transition left fewer opportunities for mobilisation via ethnic solidarities.
It is interesting to discuss the effects of state and administrative structures versus ethnic mobilisation in view of the many conflicts discussed as ‘ethnic’ in the post-colonial world. From the late 1970s, the emergence of ethnic mobilisation in parts of the Middle East and of Central Asia seems to point in this direction – although we are probably still too close to the events and they would indeed need to be disentangled from religious elements of mobilisation and freed from the simplifications that one often finds in political science discussion.
The important European cases of ethnic mobilisation that we find in the twentieth century also fit into the model of interpretation that interprets the results of my comparative view on West Africa. The cases of ethnic conflict in the Balkans, for instance, seem to indicate a rise of such mobilisation at the principal moments of the breakdown of administrative structures, both before the First World War and after the collapse of the Yugoslavian state in the early 1990s. Ethnicisation of debates in the case of Spain has frequently had to do with principal conflicts over
Therefore, it would be totally erroneous to assume that ethnicity was a factor of mobilisation ‘automatically’ belonging to African communities. The mobilisation of regionalist solidarities based on an idea of cultural-linguistic unity in a circumscribed region, was an option in many parts of the globe. It was an adequate reaction to many cases of insecurity: an insecurity that could be created by the absence of stronger administrative structures to offer protection under conditions of violence; but it could also appear as a response to changes in more formalised administrative patterns, in which particular communities regarded themselves at an unfair disadvantage.
The principal factor of insecurity, as a central background condition for ethnic mobilisation, also builds the link to studies on migration, namely in a global historical perspective, in which ‘ethnicity’ is employed as an element of group identification, but with a different meaning. The bridge between both ways of formulating ‘ethnicity’ lies in the creation of solidarity through bonds that substitute in part for a lack of protection. However, European and Chinese migrants in the principal global migration systems – and more so from the eighteenth century – already had a reference to larger existing state structures that they mixed with references to a common language. By contrast, populations living in contexts of insecurity and within very small political-administrative units, did not have any such points of reference. For the global historian, this distinction needs to be very clear. An interpretation that takes both types of group solidarities as the same phenomenon of ‘ethnicity’ is not helpful.
Regional solidarities in Europe, the settler Americas, China or India, for example, in particular if discussed from the end of the early modern period, are not automatically different from those of groups analysed as ‘ethnic groups’ in the African continent, or in parts of Asia and the Americas. The need to activate these regional solidarities, and the modalities of their employment by local populations and their leaders, are, however, different, as structures of rule (or state administrations) offer another framework on which these populations can rely. The existence of the latter often makes it unnecessary to mobilise the more regional forms of group identification in the interest of the group’s security. In other words, it appears that if the structures of states and administrations provide a somewhat reliable set of rules, ‘ethnicity’ as a factor of group mobilisation is usually not needed.
Chabal, Patrick, ‘Emergencies and nationalist wars in Portuguese Africa’, Journal of Imperial and Commonwealth History 21(3), 1993, 235–49, 242.
On the relationship between weak states and warlordism, see Reno, William, Warlord Politics and African States (Boulder – London: Lynne Rienner, 1999), 18–35.
This can be considered as a counter-model to the evolution of the pre-colonial state of Segu, as described by Jean Bazin in ‘Etat guerrier et guerres d’Etat’, in Jean Bazin and Emmanuel Terray (eds.), Guerres de lignages et guerres d’Etats en Afrique (Paris: Editions des archives contemporaines, 1982), 321–74, 336–40.
Lesourd, Michel, ‘Une remise en cause de l’ethnicité: Le comportement sociospatial des Baule émigrés dans le sud-ouest de la Côte d’Ivoire’, in Jean-Pierre Chrétien and Gérard Prunier (eds.), Les ethnies ont une histoire (second edition, Paris: Karthala, 2003), 76–90.
A different approach is taken in Apter, Andrew, ‘Yoruba Ethnogenesis from Within’, Comparative Studies in Society and History 55(2), 2013, 356–87. While Apter’s method discounts the question of political opportunities much too quickly, his main findings are complementary with my results for other groups of coastal West Africa, especially when it comes to flexibility and the question of gains.
Ikpe, Ukana B., ‘The patrimonial state and inter-ethnic conflicts in Nigeria’, Ethnic and Racial Studies 32(4), 2009, 679–97; and, especially, the new Adebanwi, Wale, Yorùbá Elites and Ethnic Politics in Nigeria: Obáfemi Awólowo and Corporate Agency (New York: Cambridge University Press, 2014).
This relationship between ethnic identity and security is also alluded to in a number of studies and handbooks, which, however, do not normally draw any larger and analytic conclusions from the matter, see Collins, Robert O., and James M. Burns, A History of Sub-Saharan Africa (Cambridge et al: Cambridge University Press, 2007), 366–7.
Falola, Toyin, ‘Writing and Teaching National History in Africa in an Era of Global History’, Africa Spectrum 40(3), 2005, 499–519.
Mamdani, Mahmood, Citizen and subject: contemporary Africa and the legacy of late colonialism (Princeton: Princeton University Press, 1996).
Falola, Toyin, Colonialism and Violence in Nigeria (Bloomington – Indianapolis: Indiana University Press, 2009).