The Ewe: A Case of Aggressive Solidarity
Unlike for Wolof-speakers or Temne-speakers, the ethnic identity of ‘the Ewe’ was an issue of public and even international debate from the 1940s. Their spokesmen won some fame as the first case of an African political movement attempting to use the stage of the United Nations.1 The Ewe unity movement seemed to be an institution with ancient roots, defined through identification with a common ethnicity.2 However, for the Ewe-speakers like for the Wolof and Temne, the interrelations of the different layers of identifications are complex and it is crucial to understand how they ‘sold’ themselves to European residents and colonisers over the decades.
The relation between Ewe identification and Avatime identification is reminiscent of the complex relationship between the categories of ‘Wolof’ and ‘Sereer’ in coastal Senegambia. Today, the Avatime speak both the Central Togo minority language of the same name and Ewe. Their historical vision regards the Avatime as second-comers in an area having been inhabited by a quasi-mythical older population, the so-called ‘Bayas’, and as having arrived before any Ewe-speakers. However, the more significant event in this respect is the Asante invasion of the late 1860s.3 During these struggles, the Avatime presented their relationship to other Ewe-speakers as a military
The Avatime thus had a continuous relation with the larger ensemble of ‘the Ewe’ in the late nineteenth century and the twentieth century.
At first glance, Ewe group mobilisation seems exceptional, and the experience of the group has attracted much interest in scholarly research on sub-Saharan Africa. Their activity appears to be an impressive protest against artificial division by a colonial border (Maps 6 and 7).5 As Togo became, from 1919, a League of Nations trusteeship territory, then a United Nations mandate administered by the French and British colonial powers, the question of Ewe
In post-colonial Ghana, the Ewe seem to have followed a distinctive voting pattern. They were opponents of various Ghanaian presidents before 1979. In the 1980s, they turned into staunch supporters of Flight-Lieutenant Jerry Rawlings and his successors under the National Democratic Congress (ndc).7
In the more authoritarian post-colonial country of Togo under Gbassinje Eyadéma, the military ruler in power since his coup d’état of 1967, the role of Ewe-ness in politics has been more severely limited. While local populations in the south of Togo continue to regard themselves as ‘Ewe’, ethnic identification has receded as a factor of political group behaviour. Before Eyadéma’s introduction of single-party rule in the late 1960s, both dominant parties – the Comité d’Union Togolaise (cut) and the Parti Togolais du Progrès (ptp) – were
An additional problem is the question of whether particular communities belong ultimately to the group of Ewe-speakers or not. The community of Accra, under the rule of the Ga Manche, and the communities of Ada on the western side of the Volta River close to its mouth, and those of Krobo with the towns of Odumase and Kpong, form a western socio-geographic boundary of ‘Eweland’. Those communities speak Gã or Adangme, which distinguishes them from the Ewe, although they actually claimed political influence over Ewe-speaking communities during the first half of the nineteenth century.10 One particular Adangme-speaking group, the Agotime, live as a pocket among Ewe-speakers. Further to the north-west, Ewe-speakers bordered Twi-speaking groups. Akwamu, a small pre-colonial political entity, was the most immediate neighbour, but the relationship with the Asante was the crucial variable in the broader region. The Asante demanded tribute from a number of Ewe-speaking communities, and their invasions in the late 1860s were a traumatic experience for the latter.
To the north, the region of Buem was dominated by Lefana-speakers, a smaller language group, and by the Akposo. The Avatime live in the region between Kpandu to the north and Ho, a second important town centre in the region, to the south. In the east, the regions of Agu, Kuma, and Be, and the area of Notsie, are clearly part of the Ewe-speaking ensemble, but it becomes more complicated with Ge on the coast. Here, the language employed is Guin or Mina, which has similarities to Ewe, but whose alleged origins from the Gã language of the Accra region allow this community to claim that it is distinct from the Ewe-speaking cultural ensemble. While the rulers of Aného, the major town centre of the Ge community, have sometimes presented themselves as overlords of all of southern Togo, it remains questionable whether they ever were in such a situation. The representatives of other communities in Togo’s south-east, such as Glidji, are today unwilling to categorise themselves as ‘Ewe’, and refer to their manifold bonds in the eastward direction.11 As a whole, it is, therefore, quite complicated even to give the current geographical limits of Ewe identifications.
Both Nugent and Lawrance treat the question of local group identifications as being at the heart of their particular perspectives. Nonetheless, while both approaches integrate several examples of conflicting layers of group identification, their implications do not point to the same problems that I intend to analyse from a comparative point of view. Nugent is interested, in particular, in how local populations made use of the colonial border, first between the British and the Germans, from 1884 to 1914, then between the British and the French from 1914 to 1957. Moreover, his study favours Likpe and the region of Buem over the Ewe-speaking areas to the south and south-east. Finally, Nugent does not concern himself with Ewe or non-Ewe group relations before the First World War.12
Lawrance focuses on a very particular perspective, which he calls the ‘peri-urban’ dimension of ‘Ewe identity’: he concentrates on Ewe-speakers as inhabitants of the surrounding towns of a large city – Lomé – where an exceptional network of roads and railroads would have created an extraordinary setting for political mobilisation.13 This approach is stimulating – and Nugent’s even more so – but it does not question the basic principle of Ewe ethnic solidarity, nor does it discuss any alternative concepts to Ewe-ness. It is therefore challenging to focus on the engagement of the different groups and individuals
Sandra Greene regards Ewe identification as a principle mainly formulated in the 1930s, as a weapon in the struggle for resources in the Gold Coast, which was on its way towards modernisation. For the Wifeme group in Anlo – a pre-colonial state ruled by Ewe-speakers – Greene has shown that the view of the Wifeme as ‘strangers’ was increasingly questioned over the years, as members of the group tended more and more to appeal to a joint ‘Ewe identity’ that connected them to the other clans and kinship groups within the state.14
Ewe-speaking groups had contact with the Atlantic world through their early participation in the slave trade. Many of the local traditions collected by British anthropologists, above all in the 1920s, point to a strong engagement of different local groups in the trade.15 Aného (‘Little Popo’) was the largest of the small ports of the immediate coastline of the later colony of Togo.16 Far more important, however, was the port of Ouidah with its three European fortresses, in present-day Benin, which was linked to the Ewe-speaking areas further westwards. This also brought the Ewe-speakers into indirect contact with the Kingdom of Dahomey, although Dahomean political activity was oriented eastwards.17 On the western side of the coastline, Keta was the most important slaving port for communities living close to the Volta River.
After colonial conquest, the different European powers drew borderlines that had an impact above all on the viability of trade networks in the region. They also proceeded with a rationalisation of power structures, and modified the latter through the organisation of ‘indirect rule’. German officials had a tendency to weaken rather than to strengthen existing political structures, while the French gave part of the local power back to the chefs de canton, but held them on a short leash.18 In contrast, British administrators believed, long before the protest movement of Ewe spokesmen in the 1940s, in the prior
Today, the Ewe-speakers – if we exclude Adangme-speakers and Guin-Mina-speakers – are a community of roughly one million individuals.20 According to Jakob Spieth, missionary of the Norddeutsche Missionsgesellschaft (or Bremen Mission) and author of a monumental account on the ‘Ewe tribes’ based on interviews conducted in the early twentieth century, the Ewe-speakers were proud of their language, and used it strongly for their self-definition.21 Nevertheless, we have to take into account the bilingualism of many individuals in the region, which makes the definition of one single ‘Ewe’ identification through language alone rather complicated. Also, concrete regulations for marriage and family structure seemed to vary ‘from tribe to tribe’.22
This problem was again expressed, in 1968, by B.W. Hodder who held that ‘the area in which the Ewe call themselves Ewe for purposes of political action is not the same as the area inhabited by the Ewe “tribe”’.23
In the last four decades, we have not really come closer to a profound analysis of these contradictions and to a definition of Ewe claims for a collective identity. These complications go back to interpretations of testimonies from as early as 1912.24 It is unsurprising that European officials were frequently quite confused about local genealogies and conflicting claims to chieftaincies. Regents were nominated by the communities and often constituted an additional force, as did stool fathers, often being the authority to nominate the candidates, and Mankradowo, i.e. leading political councillors.25 ‘Amalgamation’ of divisions in British Togoland led to even more confusion – in extreme cases, such as Nkonya, a divisional chieftaincy could remain vacant for eight years.26
At the heart of these struggles, we find references to a certain notion of Ewe-ness. This is related to the idea, uncritically reiterated in some of the literature, that ‘the Ewe’ relied on ‘decentralised’ institutions, expressed through the role of the dufia (chief), with restricted powers, and under the obligation to cooperate with a council of elders representing the important lineages, the fomewo.27 A certain political language was connected to such institutions, as different contenders argued about ‘the customs of Eweland’.28 In such cases, European officials were frequently only too ready to accept such references to customs, and to attack what was presented to them as ‘most unusual in an Ewe tribe and [that] should be regarded with suspicion’.29 However, states whose rulers claimed to be at the core of ‘Ewe identity’ had political institutions that were very similar to those of neighbouring ‘Akan’ or ‘Adangme’ states. Thus, Anlo was, from the eighteenth century onwards, organised with a ruling chief with his ‘wing chiefs’, similar to those of Twi-speaking communities like Akwamu or even Asante.30
Patriotism here is extremely local, and very strong: The natural psychological tendency of the Togoland people is not to unite to create greater groups, but to split up into ever smaller ones. I once reported that the people of the Akpini State, for instance, do not call themselves by the name of their State, Akpinis, but rather by the names of their divisions – Kpandus, Sovies, Alavanyos etcetera [sic] in fact the process is more extreme than that; even inside a division people will call themselves by the names of their own towns and think primarily of the interests of their towns before they think of the interests of their division, − thus in the Gbi division they call themselves Wegbes, Attabus, Kpoetas, Blas etcetera, before they call themselves Gbis. And it does not stop there; every week a Captain in some Sub-Chief’s town collects his people together and they go off to form a new town elsewhere; having set up on a new site he will call himself a Sub-Chief with Captains or Asafohenes of his own.32
Local divisional chiefs who were pressured to become part of larger politico-administrative entities were not at all happy with such developments. British administrators were certain that most chiefs would have preferred simply to continue with the ‘German situation’, where hundreds of so-called ‘divisions’ had had their autonomous jurisdiction. In 1942, 96 so-called sub-chiefs even petitioned to reinstate the system existing before 1931, in which each group had had its own native court.33
While in the 1920s the different populations of the Ewe-speaking zones of the Gold Coast and the Togolands would intermittently mention their Ewe identification, this is difficult to corroborate as being a long-standing tradition. In the eighteenth century, Danish and British residents at Accra and Keta had had early contacts with the rulers of Akwamu, then a pre-colonial state in the interior, but they knew nothing of any ‘Ewe’ category.34 British and German officials, or missionaries from the Basle Mission or the Bremen Mission, were active in carving out this identification. There is nothing to indicate that Ewe-speakers defined themselves as a homogeneous group; only later, did the idea of Ewe unity become fashionable, and linguistic studies seemed to sustain it.35
Moreover, pre-colonial states like Anlo and Peki had a complicated relation to Ewe culture, whether or not we take it as a historical construct. The ‘traditional ruler’ of Peki, the Pekihene (or Deiga in Ewe), who later claimed the overlordship of much of the Volta Area, stood, during the first quarter of the nineteenth century, in a close relationship to the very efficient military state of Akwamu; only around 1833 did the rulers of Peki finally break this bond. By then the Pekihene was attempting to profit from close cooperation with Ewe-speaking rulers further eastwards. The case of Anlo is even more complex and needs to be studied with caution.36 The Awoame Fiawo (the ‘traditional rulers’) of Anlo were long-standing allies of the Akwamuhene, in spite of ‘the sharp cultural, social and linguistic differences between the two states’.37 Although in Anlo oral traditions from the 1980s, elders had an
We find a dispute amongst Ewe-speakers during the colonial period as to which were the ‘traditionally’ important Ewe states. Most traditions emphasised the role of the Pekihene, Kwadzo De iv, and of his successor, built up in the wars against the Anlo and the Asante, and which gave a pre-eminence to the Peki stool (the ‘traditional’ throne). However, even these claims were challenged. In 1946 Adai Kwasi Adem ix of Awudome – then belonging to the administrative Native Authority of Peki State – demanded Awudome’s separation from Peki on the grounds that both ‘states’ had only once formed a military alliance against the Asante, and that Awudome had by far been a larger territory than Peki. The British district commissioners were unable to verify this argument: hereditary rights seemed unclear.39
Why then has Nana Akuamoah iv, whose people could be said to belong to the same stock and speak the same language of Lefana as the Buems of Jasikan, Borada, Guaman and Nsuta, found it necessary to declare himself independent of Nana Akpandja [the paramount chief of Buem]?41
The two different groups of people, the Kedjeomas (Kudjes) and the Boradas and others, were total strangers to each other and although they all spoke (and still speak) the same Lefana dialect yet that, in itself alone, is no ground for any assumption that all of them were one and the same people similarly as all the Ewe-speaking peoples of the Volta Region or the Nkonyas, Anums and Bosos or the Akims and Akwamus, for instance, do not, besides their common citizenship of the Republic of Ghana, claim a common origin because they all speak Ewe or Guan or Twi, respectively.42
Occasionally, however, local authorities and elders formulated clear antagonisms based on language. In the founding legend of the Ewe-speaking Anfoega community, the Asante war of the 1860s was explicitly explained as a conflict between language groups: the Twi-speakers from the west of the Volta, led by the Asante, battling the Ewe-speakers.43 The idea of solidarity between
Founding legends helped to sustain this possibility. The famous Notsie myth was a principal point of reference for Ewe-ness; and it appeared indeed in many of the local histories collected by Europeans in the first decades of the twentieth century. A number of allegedly Ewe-speaking groups were said to have escaped from the cruel methods of government of a legendary ruler of Notsie, Agokoli iii. The myth appears within the account on the history of Ho by Spieth, and in the early anthropological work of the British scholar-administrator, John Sutherland Rattray, in the 1920s; it was also told to the British officials in the Keta Lagoon Area. The myth is ever-present as a symbol of a broader group experience.44 Even so, a number of non-Ewe communities also claimed to have been in Notsie; details of the myth changed over the decades; and we do not have sufficient archaeological data to confirm elements of the legend.45
The accounts of scholar-administrators such as Captain C.C. Lilley and John Sutherland Rattray were well known to literate persons in Togoland communities, who re-employed these British attempts at classification. Thus, in a succession dispute, the elders of the small community of Tokokoe criticised Lilley for describing their group as not having come from Akwapim sixty years before, but having migrated over hundreds of years.46 These elders held that they knew the community’s history better than any outsider! Like many other examples, this incident points to the flexibility of information on group origins. At best, an Ewe ethnic identification existed in principle in the late nineteenth century, but perhaps only as a latent possibility in competition with
European administrative information was subject to a number of changes. After the slow disappearance of the Danish presence by 1850, the British showed only lukewarm interest in territorial control over regions close to the Volta River. Until the 1880s, the British were mainly focused on neutralising the power of the Asante State in the west and north-west, and their alliances in the Volta Region were subsidiary to campaigns financed from Accra by the local community. The British interpretation of events east of the Volta River was therefore clearly biased through an ‘Accran’ perspective.
The sudden German interference in political issues on the ‘Slave Coast’ changed the whole picture. The protectorate agreement signed in Aného in 1884 consigned large parts of the coastline between Keta and Ouidah to German administration; it provoked a more aggressive competition and led both the British and the Germans to hastily collect information on the local communities. In 1890, the colonial border was finally negotiated, allowing both colonial powers to establish their own style of rule. British administration adhered to the principle of indirect rule and wished to take the cultural and political structures of the ‘divisions’ of Peki and Anlo, which had become part of the colony, as a model. The 1912 report of the Secretary of Native Affairs, Francis Crowther, reflected this bias in favour of claims coming from the rulers of Anlo and Peki, which continued well into the 1920s. In the 1930s and 1940s, the British administration remained generally interested in ‘traditional’ issues, which, they felt, had to be addressed in the regular reports to the League of Nations in Geneva.
The German version of colonialism, on the other hand, was far less attentive to the formulation of local identifications.47 While the Bremen missionaries were quite active in ‘research’ on Ewe culture, the interest of administrators in these issues was limited. The French, who occupied a part of the German colony in 1914, and received Lomé and large parts of the Ewe-speaking zone around the town centres of Kpalimé and Notsie at the end of the decade, shared these attitudes, and concentrated on the economic exploitation of the colony and on taxation. Only with the outbreak of a spectacular revolt at the heart of the Ewe-speaking zone under French rule, did these attitudes begin to change. On 24 and 25 January 1933, Lomé, now the capital of the French mandate in Togo, was the scene of an immense tax revolt, which led to widespread violence and a brutal reaction from the colonial government. As a consequence
Apocalypse Now: The Test for a Larger Ewe Community, 1867–1914
The label ‘Ewe’ is not a colonial creation. Paul Nugent’s claim that ‘Ewe’ as a category probably appeared for the first time in 1884 does not take into account the missionary activities of the Basle Mission and, in particular, of the Bremen Mission in the Volta area.49 As early as 1858, Reverend J. Bernhard Schlegel wrote an article on the history of the ‘Eweer’, based on local sources.50 However, it is plausible that this broader label was only reactivated during the traumatic incursion by armies coming into the region from the exterior, that is, in the late 1860s. In that sense, Nugent would be right.
The Asante campaign of the 1860s polarised the whole region.51 Asante commanders and individual Asante war-gang leaders cooperated with some local communities to attack villages and political units on the eastern side of the Volta River. Asante was allied with the Twi-speaking Akwamus – a small
During these conflicts, the alliances in the region bridged linguistic differences between the groups and complicated the picture: for instance, the Ewe-speaking communities of Volo and Dufor sided with Akwamu and were active as pirates on the Volta River; the Ewe-speaking state of Anlo remained allied to Akwamu and Asante; the Gã-speaking community of Accra, and the ruler of Ada, were opposed to the raiders, and allies of the majority of Ewe-speaking groups east of the Volta River. Among the latter groups, Peki was the most prominent of the groups in the anti-Asante coalition. The British only had more detailed information from the areas close to Accra and Keta, and from Basle and Bremen missionaries who at the end of the 1860s lived in Keta, in Odumase in the pre-colonial state of Krobo, and in Ho. This distorts our picture of political activities, because the missionaries were not very interested in political matters and gave only a minimal amount of information. Mounting pressure on them by Asante raids increased their activity in information-gathering.
Anlo, Akwamu, and Asante invasions have a particular role in ‘traditions’ amongst Ewe-speakers and other neighbours. For the Avatime, Lynne Brydon described the experience of the invasion of the late 1860s as a ‘cataclysm’ shaking the very fundaments of the group – which was anchored in collective memory – and this seems to have been also the case for other communities.54 The different groups regarded as ‘Ewe’ and interviewed by Rattray in the 1920s integrated these events into their accounts: the Akwamu wars and the Asante incursions appear as traumatic incidents, which could be reinterpreted as a common ‘Ewe experience’.55
For European officials, the 1860s brought the very first experience in the interior of Voltaland. From 1869, British administrators and military officers were active in battling ‘pirates’ on the Volta River. Thereby, they met Ewe-speaking communities in the hinterland of the Volta Estuary, such as those of Volo and Dufor. In the late 1870s and the first half of the 1880s, the British enlarged their control of the coastline to include Anlo. In 1888, British troops initiated a campaign to support the pre-colonial state of Peki against groups that had, two decades before, sided with the Asante invaders.
This latter operation made it possible to mark out territory against German activities. It also represented an attempt to avenge the murder of a British envoy in the Volta River Area, who had been assassinated by the followers of the ruler of Taviefe, a small Ewe-speaking community in the vicinity of Ho. For many of the local groups, this British campaign was quickly identified as a chance to take revenge against the few communities who had sided with the Asante and the Akwamu, to which Taviefe notably belonged.56 Local communities participated strongly in ‘informing’ about group relations, as they bombarded the British with a huge number of narrations and petitions. The same participation by Ewe-speaking informants happened on the German side and featured in Spieth’s account.57
Some of the early British views are influenced by the Accra Ga Manche (the highest ‘traditional ruler’), Tackie, who claimed the overlordship of the south of the territories in the Volta River Region, including Anlo.58 Informants from Accra described the ‘Volloes and Doffors’ as ‘part of the Aquamboe tribe’, and thus ‘tribally’ juxtaposed with the other groups in the Volta Region.59 This does not go together with the regional pattern of linguistic groups, but it did not matter: the populations in question were often simply referred to as ‘eastern tribes’. The Accra leaders only had rudimentary ideas about the political organisation of the respective groups. The same applies to the coastal region beyond Anlo, where the Accra dignitaries had contacts with allies in the region of
In 1869, the Ada Manche took the role of the advocate of the Agotime populations in dealings with the British authorities, and accused the Anlo of banditry and treason. In Agotime, the situation was obviously critical at the beginning of the dry season. The Ada ruler insisted that he was particularly concerned by the events in Agotime; he named the Agotime his ‘brothers’ and strongly solicited a British intervention in the matter.64 This behaviour strengthened local claims of difference from their local neighbours and set the emphasis on language. Agotime informants confirmed this allegiance in 1888, insisting that the Agotime, while now being under ‘Krepi rule’, ‘belong to the Adangme tribe’, and ‘to Mr. W.N. Ocansey, of Addah [the Ada Manche]’.65
For the Asante incursions, Spieth believes that ‘while this war brought few advantages to the Ewe, it had nonetheless as its consequence that they began to get an understanding of their common tasks, tasks that would lead them to a slow but gradual sense of unity’.68 In the British documents discussed above, however, the question of Ewe solidarity – so prominent also in retrospect in the interviews held with community leaders and chiefs in the 1920s – did not play the slightest role. On the contrary, one of the conflicts in which British troops and auxiliaries from Accra were most involved, the conflict between the small constituency of Bator and the warriors of Volo and Dufor, and a second that involved the Mafi and the ruler of Anlo – the latter four being allied to the Akwamu – was exactly a military conflict between different Ewe-speaking communities.69 The Fiawo of Volo still commemorated the lost battle against British troops more than 80 years later, as the employment of a British warship left Volo populations with traumatic memories, but without remembering the
The role of the Agbosome, a coastal Ewe-speaking community east of Keta, which had built up, in the bights of Lomé and Ve, a sort of free harbour for contraband trade into the territory of the British colony, was another apparently complex case.72 The British became increasingly interested in this community after the German intrusion and during the partition of the Volta Region between Berlin and London. The Agbosome were active as raiders on Anlo territory, the territory of the mighty Ewe-speaking Awoame Fia, and were long-standing enemies of an Anlo sub-group, the Anyako. Once again, cultural identification was not the principal motive for solidarity here.73
The British tested the coherence of communities on the coast during their conflict with Anlo.74 In January 1885, Assistant Inspector Stewart was sent to Keta and Anloga to obtain information about the prospective enemy. Based on oral reports of his Keta informants, Stewart described ‘the Awoonahs and the Anyakos’ as the two major distinct groups, ‘with many tribes who pay them feudal allegiance’.75 With such terminology, Stewart managed to point out the
Still more interesting for our analysis is the fact that informants from the interior described community relations in the region in the sense of a clear difference between Anlo and ‘the Crippies’ (Krepis). However, for years it remained uncertain what exactly ‘Krepi’ meant.79 This label for an imagined community in the interior of the coastal region practically ceased to exist at the moment of the final Anglo-German partition of the Volta Region.80 The British claimed that they had bought the right of protection over a community named ‘Krepi’ from the Danish residents in Keta in 1850, but, once again, this was a label without concrete meaning!81 In 1858, Thomas Birch Freeman in his report of a journey into the interior described Kwadzo De iv as ‘Paramount Chief of the Krepis’. Even so, afterwards he commented that the Pekihene only ruled over his ‘subject towns’, while neighbouring ‘divisions’ such as Sokode
Thirty years later, most observers regarded ‘Krepi’ as the name of the territory dominated by Peki. Basle missionaries who visited Kpandu and Peki in the 1880s were convinced that the Pekihene was the ruler of a federation over ‘the Efe’ (including Kpandu), referring to the alliance against the Asante. The Pekihene, Kwadzo De vi, did his best to give credence to this view, claiming ‘Krepi’ to be his territory.83 However, most of the other communities involved on the side of Peki and Ho in the Taviefe war, were not at all amused by such discourses of the Pekihene. They resisted accepting Peki’s lasting overrule, although Kwadzo De initially had British support, as the British hoped to use his ‘rights’ as an argument against German territorial claims. The German Commissioner of Togo, Jesko von Puttkamer, in June 1888, obtained information on the western parts of the German protectorate, mainly on Agotime and Ylo, and described ‘Krepi’ as an ancient ‘federation’, which had included Peki, Akwamu, Avatime, Agotime and others of the Volta Region communities and had ceased to exist sometime before 1850.84
The Avatime made a competing claim. In 1888, Adzatekpor of Vane, the ruler (and later paramount chief) of Avatime, had still been virtually unknown to British officials.85 However, in 1894 the Avatime received a British confirmation that they were the rulers of the former state of Krepi (although they were
In 1888, the Taviefe War, starting with the murder of British Assistant Inspector John Scott Dalrymple on 11 May 1888, again allowed for a refinement of local identifications and regional solidarities between the British conquerors and Ewe-speaking groups.89 Dalrymple’s mission had been to discipline Taviefe, Matse, and Adaklu, former allies of the Asante and Akwamu in 1870 and opponents of Peki between 1875 and 1877 – information on their ‘behaviour’ had come from Kwadzo De vi and the ruler (Howusu) of Ho, and they were described as ‘traitors’ to the Ewe-speaking community.90 The British inspector had defined Kwadzo De vi as legitimate king ‘of all of Krepi’ threatened by unruly subjects. The British neither understood the interest of the Peki ruler in presenting the facts in a particular way, nor did they see that the alliance of the Pekihene included a number of Twi-speakers, in particular from Boso and Anum.91
After the failure of a planned peace conference with the Taviefe ruler, Bele Kobina, and Dalrymple’s death, an alliance organised by the British district
Buem, further to the north and at the margins of the Ewe-speaking area, had interested the British because of its strategic position for inland trade. The British had also (quite erroneously) concluded that the inhabitants of Buem spoke a language similar to the Ewe-speakers and were thus ‘naturally’ members of a British protectorate.95 Further south, the German and the British governments continued to discuss the ‘possession rights’ over Kpandu, Ho, Avatime, and Adaklu, until the border was fixed in 1890, with most of these ‘divisions’ remaining in German territory, and the Adaklu villages divided between the Keta District of the Gold Coast and the Misahöhe District of German Togo.96
As British officials struggled to come to an ‘authentic’ organisation of local political units, many participants from the African side used their chance to sell themselves as rulers of their respective community. Issues of being an ethnolinguistic group did not play a role in that process. Smaller communities, such as Taviefe, Adaklu, or Ho, and pre-colonial states with Ewe-speaking rulers, such as Anlo and Peki, dominated questions of group identification. However, the hardening of the colonial border had an impact on categories, as had missionaries of the Bremen Mission in defining Ewe culture and politics. These missionaries and some linguists and early anthropologists, showed a
It is very easy to claim, with hindsight, that the different Ewe-speaking communities had already had a strong feeling of unity under German rule.99 However, there is no proof for such a hypothesis. German administrators were very much focused on labour issues, and attempts at engaging with community structures were mainly to be found where the ever-present problem of labour and labour evasion was particularly acute, as in the region of Sukpe. In this region, the administrator defined local populations as being members of the ‘Aveno tribe’, who had settled in the region only in 1850. These ‘Avenos’ would have bought land from the Ave, Agotime, and Adaklu communities, earlier settlers in the region.100 However, while in these concrete cases the Germans commented on the existence of different ‘divisions’, we find little inclination (and information) to point out larger solidarities or to install
In contrast, the British (as in the Gambia or in Sierra Leone) wished to create a chieftaincy based on ‘traditional rules’ – which in Togoland opened the gate to constant disputes. In Ewe-speaking communities, a typical cause of disputes was the previously mentioned role of the ‘stool father’ (the zikpuito), whose function was unclear to the colonial rulers.102 With their agenda, the British had a clear objective of classifying populations according to their imagined larger affiliation. After 1915, the British imported these patterns into the western part of former German Togo, based on some established models. At the same time, not only British rule but also the Ewe language expanded. Ewe entered in the 1890s into regions such as South Akposo; in 1900, the Bremen Mission station official at Amedzofe, Ernst Bürgi, also reported a spectacular rise in the use of the language.103 This process started with an influx of refugees from the area between Lomé and Kpalimé, fleeing northwards from the reprisals of German soldiers and their auxiliaries. Other local populations also discovered the language to be useful.104 Moreover, German Catholic and Protestant missionaries contributed to this process by using Ewe as the principal language in education and codifying it, which was grudgingly accepted by the German colonial government. From 1904 onwards, an Ewe-speaking elite formulated its own interests, including in language politics.105 These elites were very vocal after the British conquest in 1914.
However, European rulers remained unsure as to how to employ ‘Ewe’ as a group label. In geographic terms, the Germans called the area of Asome,
In general, the experiences after colonial conquest do not confirm the hypothesis of a strong and continuous pan-Ewe sentiment that challenged colonial boundaries. Many locals attempted to obtain improved positions by reference to former allegiances and dependencies. ‘Ewe-ness’ was a marginal part of this picture at best – in spite of the linguistic efforts of the missionaries.
The New Border and the Quest for ‘Authentic’ Arrangements: British and French Readjustments of ‘Ewe’ Institutions after the First World War, 1914–1945
The French and British invasions of the German colony of Togo led to an extension, and, eventually, to an exacerbation of the administrative separation of different groups of Ewe-speakers. Local elites attempted, however, to ‘choose’ what they regarded as the more benign European power of tutelage; and to improve trade contacts, as Lomé merchants hoped for free access to the port of Keta. Important spokesmen of the coastal elite, such as Octaviano Olympio, actively lobbied for British rule over Lomé. However, it has to be asked how far these actions were motivated by a feeling of Ewe-ness, as was suggested by D.E.K. Amenumey.109
Two examples show the complexity of this interpretation. During the First World War, the Awoame Fia, Togbe Sri ii, continued with the claims of overlordship of the eastern Volta Region, which Anlo rulers had already been formulated in the 1860s and 1870s. This explains the Anlo war effort in favour of the British side. Further east, the Lawson family in Aného appears to have claimed in 1922, under French rule, overlordship of populations in the south of Togo. However, again the label of ‘Ewe’ is nowhere employed in the evidence.110 From a broader perspective, it is also significant that we find no attempt by local populations to bring the question of the perils of the ‘Ewe community’ before the League of Nations. European officials would have been sensitive to identifying what they saw as correct ‘tribal’ settings. ‘The Ewe’, however, did not play a prominent role in European reports on the Togoland mandate written in the 1920s. In the British 1927 report on the Togoland mandate sent to the League of Nations, the authors, while briefly mentioning the Ewe language, recommended installing ‘Akan’ structures of local government wherever possible!111 Even in the anthropological work of the famous scholar-administrator John S. Rattray, who interviewed communities about their Notsie experience, the fact of ‘Ewe-ness’ and wish for Ewe unity do not at all appear.112
After the British conquest of western and southern Togo in 1914/15, claimants from the Gold Coast Colony and the Gold Coast Protectorate demanded their rights of political rule over ‘natives from Togoland’. The ‘Togoland communities’ themselves were more interested in revising local hierarchies installed by the Germans, like of the Adele paramount chief over the Adjuti. The other goal was to prevent a return of the German administration. In Ho, Noepe, Adame, Aflao, and Lomé, the chiefs petitioned for the continuation of British rule, pointing to German atrocities with regard to forced labour, violent tax requisition, and corporal punishment; and the Anfoega tried to free themselves from the Kpandu authority, imposed on them by the
Contrary to Amenumey’s assumptions, local populations did not mention any idea of Ewe cohesion, and although some held that they ‘do not want the Krepe tribe again split up between two separate European Governments’; others, like the Agotime, now insisted that they had nothing to do with ‘Krepi’.114 Indeed, communities in Togoland under British mandate were content to see the power of Anlo and Peki curtailed. The British obsession with the creation of ‘native states’ generated new tensions, as many chiefs refused any participation in that. To give one representative example: when the ‘Divisional Chief’ of Taingbe applied to exert a levy to improve the infrastructure of Taingbe Town, he learned he was no longer the legal authority entitled to charge such a levy.115 In British administrative politics, ‘native jurisdiction’ and ‘native treasury’ became the privilege of ‘states’ that had been a product of ‘amalgamated divisions’, which only created more problems. The French officials, while choosing a less elaborate terminology, basically attempted the same: local communities were put into chefferies de canton, without consultation.
In subsequent community conflicts before the 1940s, previous struggles between pre-colonial states were much more a point of reference than ethnic solidarity. One good example for British Togoland is the land conflict between the inhabitants of Taingbe and Tokokoe (in the zone north-east of Ho). Both sides mobilised historical ‘tradition’, but did not refer at all to their common ‘Ewe past’. During a stool dispute within Tokokoe, between Agamah and Buankrah ii, which remained a constant problem through the whole of the
The question of the new border and its effects were another obvious problem. Most extreme was the situation of Agotime, where 29 villages became part of the French zone, while two villages, including the village of Kpetoe of the Agotime head chief, remained under the British.117 Another difficulty was the continuous importance of the former border between German Togo and the British Gold Coast Colony, now translated into an administrative border between Gold Coast and Togoland under British mandate. Some chiefs wished to settle old scores, as in the case of Agbosome, which continued its aggressive politics from the nineteenth century. The chief of Agbosome laid claim to the towns of Some and Have and thereby challenged Anlo rule.118 In all these conflicts, demands of the former pre-colonial states and the interests of smaller communities were far more important than any idea of larger (ethnic) group solidarity.
On the French side of the mandate, the situation was similar. In a 1930 report on the situation in the Cercle of Klouto, with the administrative centres of Kpalimé and Misahöhe, the French dealt with the ‘Ewe problem’ in a very ambivalent way. The district commissioner argued that even in view of the common Notsie legend, the ‘Ewe race’ did not seem to know any more sustained solidarity. ‘The village is the true ethnic group’, the French commented on the fragmentation of political solidarity in the area. The same comment applied to the border regions of Mission Tové and Akoviepe.119
The main pressures of the interwar period came, once again, from the ‘traditional states’ in the region, Anlo and Peki. In particular, the Pekihene hoped after 1915 for the creation of a larger regional paramountcy and reminded the British of earlier promises. The chiefs of Anfoega, Hlefi, Dsocho, Goviefe and Akrofu, Adaklu, Agotime, Dakpa, Zofe, Logba, and Botoku were all against any such paramountcy; the inhabitants of Akuope and Taingbe were even more explicit about the historical background, and they accused Peki of passiveness during their 1860s conflict with Anlo; and even in Abutia, which was initially positive, integration into Peki State was complicated.120
In the latter case, the Pekihene demanded the integration of Abutia into the state between the early 1920s and 1945, until the final refusal of the head chief of Abutia in July 1945 and Abutia’s admission into the newly created Asogli State in Togoland. The Asogli solution helped to favourably solve land conflicts with neighbours.121 Within Abutia, the sub-chief of Abutia Kloe – one of Abutia’s communities – tried to obtain Peki’s help against his own paramount chief and nearly provoked a Peki intervention under Kwadzo De X. This initiative was brought down with British support.122 The ruler of the second largest community in the Abutia ‘division’, the Dufiaga (sub-chief) of Agove, Adja Dra V, had long been uninterested in the Peki initiatives. However, in 1933, he suddenly decided to back the petitions coming from Peki, and sent representatives into the Peki State Council. The conflicts with the paramount chief of
In 1945, in a last-minute attempt, the Agove sub-chief, S.K. Kumah, and the asafohene (military society leader) of Teti (the central division), Okai Debra, allied to call for integration of Abutia into Peki. The situation was little favourable for such an attempt: the Peki Ruler, Kwadzo De X, was seriously ill, and the British were not eager to change their established line of politics.124 Even so, Kumah and Debra mobilised part of the local populations, and organised a showdown at a divisional meeting of Abutia, where the Regent of Peki, Ayim V, and the Howusu of Ho as President of Asogli State, explained their claims. Even so, the attempt did not win a majority except in Agove, and in spite of the massive protests of the Peki authorities, the British urged Ayim V and his successor as regent, Donko, to renounce their claims. In the end, the Peki initiative remained fruitless.125
We have discussed the Peki-Abutia affair in so much detail because it was exemplary for many similar cases of the interwar period. Against their earlier preferences before 1914, in which the British had favoured a regrouping of
The experience of the Anlo rulers was similar: after their engagement for a larger political unit with common cultural characteristics during the First World War, they were unable even to create stronger relations with the Adaklu chiefs as a neighbouring community. Both groups were Ewe-speakers, adherent to a common version of the Notsie legend, and had been allies of the Asante in the late 1860s. However, this was insufficient.127 Discord between the two communities was expressed in an ‘oral tradition’ centred on an attack of Anlo units against Adaklu villages shortly after 1870, which had forced the Adaklu to flee into their strongholds near Adaklu Mountain. The respective tradition was rebuilt in the 1920s against attempts at integrating Adaklu villages into Anlo State.128 In a second oral tradition, reinterpreting events during the mythical exodus from Notsie, the Adaklu founding father, Foli Kuma, was described as equal of the founder of Anlo, Wenya. This tradition thus refused Anlo any pre-eminence.129
Thus Adaklu elites moulded their own, independent politics in the region, and refused integration into Anlo, but also into the new Asogli State around Ho. With regard to the latter, the Adaklu spokesmen claimed that subjects of the Howusu had regularly stolen land from them and were therefore long-standing enemies.130 When British pressure on Adaklu community leaders became stronger, the Adaklu adapted their traditions: they now held that their
Both these conflicting goals of different communities, and the very existence of the territorial border between the Gold Coast Colony and Togoland, prevented a strengthening of Anlo State as a genuine Ewe-speaking political unit. British officials believed that it was easier first to tackle the question of Ewe-speaking groups that stood under the rule of the Ada Manche – also situated in the Gold Coast Colony – and to profit in this context from the circumstance that the paramount chief of Ada State had been suspended.134 The district commissioners of Keta and Ada assembled the ‘Ewe chiefs’ of Ada State in Tefle and inquired whether they wished to leave Ada overlordship. Most of those chiefs indeed claimed they would prefer to belong to Anlo, as they said ‘for reasons of tribal links’ – using the British ‘tribal’ argument. Some chiefs, such as of Sukpe and Bator, who had been suspended and expected to be destooled before the British initiative took shape, used the development to save their position.135 Thus, the divisions of Agave, Sukpe, Tefle, Vume, Bakpa, Bator, and Mepe – ‘the Ewe Divisions’ – within Ada State, made a ‘request for
Agave is the best illustrated of these cases. The Agave did not argue with questions of ‘Ewe identity’, but focused on relations between political entities: they claimed they had fought constant wars with Anlo, and that they had been the overlords of the Ada Manche and were thus on the same level as the Awoame Fia! Language was only an additional argument, with the Agave pointing out that as Ewe-speakers they were distinct from Ada’s Adangme-speakers. It is remarkable that they styled themselves with Ewe titles on this particular occasion, but relied elsewhere on a political terminology, the ‘wing system’, that was normally defined as ‘Akan’.137 As a whole, the demand of Ewe-speaking rulers from Ada State was pending for a period, and then merged into the large current of Ewe activities of the immediate post-Second World War years, which I will analyse below.
In Anlo State itself, the rhetorics of power also became increasingly confused, as the Awoame Fia lost control over the state council based in Anloga, which many chiefs of the neighbouring ‘divisions’ did not attend anymore. In 1943, the ‘Ewe Union’, a nascent pan-Ewe institution counting in its ranks many ‘intellectuals’ and a good part of the less influential chiefs of the state, showed its strength. The Awoame Fia’s tax policy had enraged many of the local chiefs even further, and had provoked the opposition of the Ewe Union.138 However, the main thrust of resistance against any more centralised politics came from the old Anlo capital of Anloga, whose headmen, in 1944/45, virtually paralysed Anlo State.139
Doubtless, the colonial border between German and British, then between French and British possessions, created its own dynamics that had an effect on identifications. Illicit commerce and seizures of cattle and even land were typical.140 In Kuma on French territory, villagers complained about the confiscation of their lands lying in the now-British zone by Togoland villagers, with the alleged tacit support of the administration; in the opposite direction, villagers of Bogo Achlo in the French zone had occupied land belonging to peasants of Baglo, who were now living in British Togoland.141 Such acts, started first during the general instability of the war years, created new, long-lasting enmities.
Migration from the French into the British mandate also became typical. Those migrants were described by the inhabitants of the Volta area as ‘natives of the French zone’ and became the occasional victims of xenophobic rhetoric from a ‘proto-national’ perspective, even if they were Ewe-speaking. Therefore, aside from local discourses, and before the idea of Ewe unity became important, the image of the ‘proto-national’ stranger, with reference to the colonial territory of origin, already had an impact.142 There were some more ‘ethnically’ oriented exceptions to this rule. An early example was given by the political evolution of Buem in the British zone, where the populations were, in their large majority, not Ewe-speakers and did not share in the Notsie tradition.143 As in the cases of Asogli and Awatime States, British administrators wanted to
Hence, we do not find an ethnic Ewe movement in the interwar period, and Ewe identifications were weak if they existed. The question is whether, as Amenumey suggests, such solidarities were simply implicit, or if incidents like the tax revolt in Lomé on 23 January 1933 can be read as expression of Ewe unity, as is held by Benjamin Lawrance. We do not have the slightest evidence in that regard, to say the least.147 As exemplified by the Bund der
Succession conflicts under the French mandate were similar to situations in British Togoland – and frequent in the rural constituencies, in particular those close to Kpalimé, even with the theoretically stricter French control of African authorities. In Agu-Nyogbo, one of the large districts, the death of the paramount chief, Kofi Pebi, in 1939, led to weeks-long troubles.149 In the case of Danyi, historical ‘traditions’ were mobilised: Paramount Chief Bassa having been replaced by Gabla, there was an intense production of evidence relating to the dispute, promoted by an Ewe Presbyterian Church (epc) preacher from Kete-Krachi, Reverend Wampah. This ‘tradition’ separated two groups in the region, the Daye Kakpa and the Daye Atigba, which both had Ewe origins and had been in Notsie before migration. According to this ‘tradition’ it had always been the first-comer of the two groups that had had the right to the paramountcy.150
In other cases, land was at stake: such as between the two Kpele villages of Goudève and Elé, or Woame and Mayondi. Issues were complicated if the land in question was to be found between two villages lying in different cantons, such as between Bogo Achlo and Daye.151 Only rarely did the creation of ‘traditional histories’ to be used in land disputes have to do with Ewe myths. Finally, even the immigration of Muslim populations from the north into the zongos (‘Muslim quarters’) of Lomé and Kpalimé failed to trigger a discourse on ‘otherness’ directed against northerners on the part of the Ewe-speakers.152 Thus mobilisation under ethnic banners never became a particularity of the Ewe-speaking communities of Lomé, Kpalimé, and Tsévié, at least not before 1945.
We are Natives of that part of Togoland known as the French Togoland and we are called and styled ‘Ganyi’ by all the people who are described by the Gold Coast Government as the Ewe speaking people. We are entirely a race or tribe different and distinct from the rest of the Ewe speaking people. We have different customs and ceremonies. Our ceremonial rites are entirely different. Our names are distinct from the rest of the Ewe People and thus you see, we do not form part of either the Anlos and the Pekis who are known as the Ewe speaking people.154
Exaggerations abounded during this power struggle between different Ewe-speaking or related groups in Accra. However, whenever such groups felt threatened, as by Paramount Chief Keami Osaabo in the Akwapim town of Nsawam, they eventually applied to the British authorities pointing to their common ‘Ewe nationality’.155
Even so, in the British-controlled territories, most conflicts and discussions between Ewe-speakers tended towards rather ‘regular’ land conflicts, and were unrelated to any questions of Ewe solidarity. The relations between the settlements of Ziavi and Klepe (west of Ho) are typical for such conflicts.156 Sometimes, these land conflicts were reported upon in connection to histories of loyalty and betrayal, such as in the conflicts between the Fiervier community and the chief of Sukpe, both Ewe-speaking units within Ada State before 1945. Each side claimed that the other had immigrated into the land and had been sheltered by their own group, with the Fiervier suggesting that the Sukpe had illegally taken over power. This conflict went on in the 1940s when both entities belonged to the Tongu Confederacy, with the Sukpe paramountcy finally being confirmed through the intervention of the colonial power.157
In Sokode Division, smaller communities also mobilised history to escape from the allegedly abusive Sokode paramount chief. Thus the Hoviepe elders complained they had once, in the battles against Asante, Akwamu, and Taviefe, accepted the Fia of Sokode, Ampim Danku ii, as leader in a war alliance. Now, the current Sokode chief, Tenkloe Koku ii, had created a false myth of the Hoviepe settlers coming from Klepe. Gbogame, another ‘subdivision’ of Sokode, followed the Hoviepe initiative and also claimed ‘independence’ from a false political dependency.158
Finally, we need to come back to identifications in Agotime during the interwar period. The Dufia of Agotime living in Kpetoe, Hene Hoe Keteku ii, was subject to opposition in the 1920s, as he was challenged by a legal complaint from the Chief of Agotime Afegame, Mahumasro, before the Supreme Court of the Gold Coast. The latter claimed that the paramount chiefs of ‘the Agotime nation’ had ‘always’ come from Afegame and that the preference for Kpetoe had been a product of German misunderstandings and manipulation by one Mati Sukpor. After Keteku’s destoolment, i.e. his removal, the Afegame claim was again intensified in the early 1940s, with the mankrado in Afegame also advocating Agotime’s integration into Asogli.159 Again, an opposition between rightful
All in all, the British may have hoped, in the interwar period, for a pan-Ewe ‘tribal’ feeling that would have allowed them to better control the territory. As Anlo or Peki were too weak to be used for political matters in Togoland, some British authorities had indeed hoped in the 1930s that ‘an Ewe Confederacy will in due course settle this matter’. This did not seem to happen.161 Ironically, however, the category of Ewe unity suddenly enjoyed unexpected successes after 1945, and the British would in the end be forced to fight this idea of unity.
The International Agenda: Ewe-ness as an Anti-colonial Weapon, 1945–1957
The panorama of group mobilisation in the Trans-Volta region, which in the interwar period gave little room for discussions of Ewe community sentiment, changed completely in the 1940s. In 1945, the question of Ewe unification seemed easily to overshadow all the other political problems. A strong lobby group of so-called Ewe ‘unificationists’ had a clear intuition for the French political and military weakness after the war, and its members started an impressive campaign.162 The transition of the official administration of the
The pan-Ewe movement operated, during its roughly two decades of existence, under different labels, but the All Ewe Conference (aec) was the most famous of all those groups. aec leaders initially enjoyed enormous prestige at the local level, instead of being just an elitist movement: from 1945 to 1947, the offspring of the aec as a political party, the Comité d’Union Togolaise (cut), won the elections in Togo under the French mandate; Sylvanus Olympio became a deputy in the French National Assembly, and the cut dominated the Territorial Assembly in Lomé. By 1948, the party could count on a considerable grass-roots support in southern Togo. This was followed by a period of decline, which was due both to French repression and to disappointment of party members with the apparent failure of the rather ambitious unification plans. The Parti Togolais du Progrès (ptp), the main competitor party, also had its strongest base among Ewe-speakers, but distanced itself from any pan-Ewe programme.165
In British Togoland, the All-Ewe Conference was also influential, but it soon lost out against the Togo Union (tu), later transformed into the Togoland Congress. As aec activists appeared too radically ethnicist, the non Ewe-speakers of the Volta Region refrained from taking part in their initiatives. The tu easily captured these voters and the more moderate Ewe-speaking electorate.
In its early phase the pan-Ewe movement sought a particular ‘Ewe tradition’, which was both reinvented and popularised through the effort of Ewe-speaking journalism. An important role in this context was played by the Ewe News-Letter, a journal edited by the Anlo-born activist Daniel Chapman. Chapman and his collaborators greatly extended the length and scope of the ‘tradition’, stretching ‘Ewe legends’ beyond the Notsie myth back to Oyo and the Nigerian city-states. The authors of the Ewe News-Letter also tended to use an inclusive approach that involved all the communities of southern Togoland, whether they were in the end Ewe-speaking or not.166 Chapman also came to the (rather problematic) conclusion that the ‘pure’ Ewe groups had been ‘Peki’, ‘Anlo’, and ‘Tongu’, in spite of the fact that the creation of the Tongu Confederacy after the Second World War was mainly a colonial idea.167
British and French officials attempted to formulate a joint position with regard to the Ewe claim, but this was complicated, given that the French regarded this claim as subversive and invented by Communists, and did not have much interest in discussing the ‘truth’ behind the ethnic arguments.168 Inside the British administration, the position was more ambiguous. In June 1945, the French commissioner of Togo, Jean Noutary, turned the Ewe argument against them: if a ‘majority’ of the Ewe lived under French rule, this was all the better for them. To prove their point, the French counted all speakers of languages related to Ewe as ethnic Ewe, changing their older categorisations, according to which not more than 14.4 per cent of their mandate was to be seen as ‘Ewe’. It was a particularly weak point of the aec that the movement had no response to these claims.169
For British officials, it was seemingly easier to accept the basic assumption that ‘the Ewe’ as defined by the activists, were indeed a single unit and that there had previously been ‘Ewe states’, such as Peki and Anlo.170 ‘Ewe identity’ now became an issue in the Togoland Reports to the United Nations committee, where the British authors claimed that ‘Ewe’ was the majority culture in all of the southern section of Togoland, with the single exception of Buem State, where ‘Akan’ was said to be dominant. Smaller, non Ewe-speaking groups were said to be totally ‘eclipsed’ by ‘the Ewe or the Akan’. British officials had thus entirely bought the story of ‘Ewe unity’: the Central Togo minorities, or the Lefana in Buem, no longer counted.171 Nevertheless, British officials argued that the two mandates could not be integrated into the Gold Coast – although they sympathised with Ewe ‘tribal’ sentiment.
Officials on both sides of the colonial border were in agreement about seeing a ‘strong national feeling’ among Ewe-speaking populations. When Ewe-speakers from Anlo went to clinics recently built in the nearby Lomé area, British officials held that ‘the Ewe people do not consider themselves French and British, but simply Ewe’.172 Even so, these observations frequently overlooked the other, parallel, discourses. One of those favoured Togoland identifications over Ewe identifications, and drove some of the All-Ewe Conference activists under British rule into the arms of Kwame Nkrumah’s Convention People’s Party (cpp). The latter benefited from these contradictions by winning a part of the Ewe-speaking electorate of the southern section of Togoland, during the Legislative Assembly Elections of the 1950s and during the Togoland Referendum of 1956. Togolese politicians in French institutions also played on both concepts, that is Togolese versus Ewe identifications.173 Some of the former sympathisers of the Ewe reunification idea
British administrators accepted the claim for an ethnic ‘Ewe identity’, but worried from the outset about the ‘chauvinism’ of ‘the Ewes’ who wished to marginalise other southern Togoland populations, as for example in Buem.175 However, during the peak of pan-Ewe militancy, in the late 1940s, we even find a conflict within the movement about the definition of ‘Ewe-ness’. The activists did not want to see a Peki or Anlo dominance within the movement, as those communities were not original ‘Togolanders’.176 The meetings of the chiefs of the four existing ‘native states’ on Togoland territory under British mandate, Akpini, Asogli, Awatime, and Buem – officially discussing the nomination of a delegate of the southern section of Togoland to the Legislative Council in Accra – had indeed an ambiguous position towards pan-Ewe statements. The Togo Union rapidly dominated these occasions, and most chiefs lost their sympathies for the more radical positions of the All-Ewe Conference. Therefore, ‘Ewe unification’ disappeared from the agenda by the late 1940s.177
Between 1944 and 1949, however, the question of ‘Ewe identity’ also became part of a generational conflict. In Anlo, the younger literate populations supporting the Ewe Union campaigned for pan-Ewe activist Daniel Chapman of Achimota as future ‘Ewe representative’ for the Gold Coast Legislative Council in January 1945, against the candidate of the Awoame Fia, Julius Tamakloe.178 In British Togoland, young activists rebelled against their limited rights under ‘Ewe customs’, which subordinated them in political communication to their asafofiawo (age set leaders) and their stool fathers. Under the cover of Ewe
The Togo Union and the Togoland Youth Association pushed the Ewe issue increasingly into the background in favour of the joint Togoland discourse – which turned against the ‘Gold Coast Ewes’ organised first in the aec, and, later on, in the cpp. The inhabitants of southern Togoland complained that ‘foreigners’ were attempting to take over political power in ‘their country’.180 In French Togo, the political competition between two ‘Ewe’ leaders pursuing distinct programmes lowered the interest in the Ewe issue from the late 1940s. The colonial administration believed in ‘Ewe propaganda’ as a conspiracy by the Nkrumah government in Accra to annex the French-mandated territory.181 This was a misinterpretation. While Sylvanus Olympio and the cut achieved unexpected political success in 1958 and took over political power in the mandate, the Ewe issue had by then long disappeared from the cut’s agenda.
In British Togoland, we find until 1953 an odd parallelism between local demands and British belief in the amalgamation of divisions that followed the strategies of the interwar period. This became obvious during a 1949 meeting of representatives of the four ‘native states’ and of delegates from the ‘unamalgamated divisions’ in Sovie. The Ewe issue had disappeared; the All-Ewe Conference was not present; and Ewe-speakers from the Gold Coast Colony were excluded from the discussions.182 In local conflicts, some ‘Ewe
In 1951, in Togoland under British mandate, the pan-Ewe idea had lost its impetus. The question now was whether Togoland could become an independent territory outside of the Gold Coast, and leaders of the Togo Union now accused the Gold Coast cpp of ‘subverting’ the region and lobbied for a common territory for all ‘Togolanders’.184 Nevertheless, Nkrumah’s cpp made inroads in parts of the populations of Voltaland’s southern section and won the support of discontented populations of several smaller areas. This was not sufficient to win the referendum for integration into the Gold Coast in the southern part of the trusteeship territory – they received the necessary votes through an overwhelming cpp majority in the north – but the regionalist ‘Togolander’ sentiment, even in the south, only had a narrow majority. As the votes were counted as a single regional block, British Togoland finally became Ghana’s Trans-Volta Togoland District in 1957.185
In French Togo, the cut’s 1958 electoral victory made Olympio the Prime Minister of a now autonomous state, and brought about the country’s independence in 1960. This could have brought the unification issue back onto the agenda of the larger region, but it did not. Leaders of Olympio’s party were content to establish cut rule in the ‘recalcitrant’ parts of the country and to punish ptp supporters through levies for having voted for the ‘wrong party’.186
‘The Bigmen Get Small; The Small Ones Big’:188The Regional Scope of the Battle for Autonomy and Resource Allocation
Apart from the international discussion, the reorganisation of Togoland in the British-mandated part, and communal reform on the French side, created a hidden local dimension of conflict and change, in which ‘Ewe mobilisation’ had a different sense.189 Some local struggles involved established paramount rulers and sub-chiefs, others mobilised so-called ‘youngmen’.190 These conflicts appeared, above all, in Anlo and the Tongu Confederacy: they led to the creation of militias – such as by the fiawo of Tefle, Sukpe, and Vume – and to violent clashes in the first half of the 1950s.191 Everyone referred to United Nations trusteeship: in September 1945, the divisional chiefs of the mainly Ewe-speaking communities of Liati, Fudome, and Vli refused to pay
Anlo in the 1940s was a particularly violent case. The nobility of Anloga was annoyed that Togbe Sri ii preferred to live in Keta, close to the office of the British district commissioner.193 The issue of ‘Ewe-ness’ was strategically used to impress the colonial power: the Anloga headmen claimed that the Awoame Fia’s ‘absence’ damaged the administration of taxes, which led to a lack of funds for the New Africa College, according to them the best such institution ‘in the entire area occupied by the Ewe speaking tribes’.194 Also, during the internal debates in Anlo before 1949, representatives of the ‘Ewe parties’ were duly invited to boost the prestige of the chiefs.195 Reference to Ewe-ness was, however, most useful when it came to excluding groups on ‘traditional grounds’. In 1946, the Awoame Fia of Anlo withdrew authority from the Atiavi Council, and defined the Lostofi clan as autonomous from the Atiavi. Furious, the Atiavi chiefs first tried to convince the Awoame Fia of his ‘error’. Then, in a long petition to the British authorities, they explained that they were the descendants of the respective Ewe groups coming from Notsie and settling in Anlo, and that the Lostofi were former slaves coming from somewhere else. Thereby, the Atiavi managed to bring ethnic matters to the fore against a local decision regarded as unfair.196 The second issue involved discussions about pre-eminence amongst clans within the states: the members of the kinship group
The Pekihene encountered a period of similar difficulties in the 1940s and the rest of his power eroded. The Fia of Awudome and Benkumhene of Peki State, Togbe Adai Kwasi Adom ix, withdrew from this native state, challenging the ‘traditional paramountcy’ of Kwadzo De X of Peki. The Awudome supported this decision with alleged ‘historical tradition’, claiming that their community had been the first to revolt against Akwamu rule, back in 1829. They also demanded, repeatedly, a United Nations inquiry into the matter. However, the British, who at this point were still convinced that the future lay in Ewe
In the ‘amalgamated states’, problems were similar, and the power of the paramount chiefs waned. In the case of Akpini, the state council in Kpandu failed to intervene in the struggle for the chieftaincy of Wusuta, which had become increasingly violent.203 In Alavanyo, the Akpini State Council also attempted to press the reinstallation of formerly destooled ex-chief Godwin Anku as Atakora V, but failed as much as in Tsrukpe.204 As no established mechanisms had been created for conflict resolution, ‘native states’ such as
In these internal conflicts, the attention was focused on hostilities between the ‘central division’ of the ‘native states’, and groups in other ‘divisions’, which led to memoranda and reflections on pre-colonial history. The asset of ‘Ewe unity’, and the related ‘traditions’, seemed useful for a short time. Even so, very quickly after the peak of international pan-Ewe campaigns, the Ewe issue disappeared as an argument. In 1953, divisional chiefs within the Tongu Confederacy even demanded the deletion of the reference to ‘Ewe states’ in the State Councils (Colony and Southern Togoland) Ordinance No. 8, to the astonishment of the British administrators.206
Many Togoland communities believed nonetheless that they needed legal protection against future interventions of the Pekihene or of the Awoame Fia: the fear of the ‘classical’ states lingered on. This was in part irrational, but chiefs such as in the Tongu Confederacy tried to shut out the Awoame Fia and the Pekihene from Trans-Volta Togoland regional entities forever. Local chiefs feared historical prerogatives of these ‘traditional Ewe rulers’. Even so, most local rulers did not refrain from strategic alliances with Anlo and Peki, whenever they were politically advantageous.207
A good example of the manoeuvres of population groups between states are the Mafi, who were divided between Anlo and Ada, and later the Tongu Confederacy. In 1912, the British had described this separation as ‘traditional’. This idea was based on Mafi behaviour in an 1865 war, and indeed supported by the Anlo Mafi Union – an association that included many local dignitaries.208 In the 1940s, however, many leaders of Mafi villages wished to bring the
In a pending case whose resolution remained unclear until 1953, representatives of Mafi communities from both sides of the Gold Coast-Togoland border employed the Notsie myth of ‘all Ewe tribes’ and the joint exodus.211 A minority part of the communities on Gold Coast territory refused the new arrangement
In some more marginal communities, the use of the Ewe argument in local conflict was even more attractive. We have already seen this for rulers of ‘Ewe’ communities which had previously been part of Ada State, who knew how, when it suited them, to play on the concept of Ewe unification to further their interests. In January 1945, spokesmen of these communities had convinced the district commissioner that ‘56,000 Ewes being controlled by the Ada Manche’ needed to be liberated. It was unclear how this was supposed to happen. The chief of Agrave, one of those ‘Ewe units’, refused to become a future member of a joint division of the eight communities as he feared Anlo interference.213 Other chiefs agreed with this critique.
Another issue was the relationship of the communities of the Volta Region to the state of Akwamu. The relations between Ewe-speaking locals and the Akwamu authorities – descendants of the ‘plunderers’ of the 1870s – were particularly interesting. In the 1940s, the region of Volo, now a small fiagadom (rulership), had been exempted from Akwamu rule and become part of the newly created Tongu Confederacy.214 After some violence in 1937, the Akwamuhene also accepted that Dufor would leave Akwamu rule.215 However, the Akwamu Native Authority reserved for itself a part of its rights
Frankadua, a small town but with an important market, was a test case. Kwasi Abliza iii, the Fia of Volo, claimed rights over this town.216 The British administrators east of the Volta intended to solve the issue, but their colleagues in the Gold Coast Colony were more interested in maintaining the prestige of the Akwamuhene.217 For a period of five years, the situation of Frankadua thus remained unclear: tax payments were suspended; police forces from both Akwamu and the Tongu Confederacy occasionally entered the town and several inhabitants were shot.218 The ‘Volos’ in Frankadua produced numerous pages of ‘traditional claims’, explaining why the town had to be separated from Akwamu. The Akwamu councils argued with ‘traditional rights’.219 Kwasi Abliza iii accused ‘the Akwamus’ of having usurped the land Volo had been given in 1873.220 Part of Abliza’s interest was in improving his own position in Dufor, where he was a contested ruler, through success in the Frankadua issue and the hereditary conflict with Akwamu.221 In the 1950s, the debate became even more heated. The Volo headmen pointed to a census in which 607 inhabitants of the market town defined themselves as ‘Ewe’ (and, oddly, 50 more as ‘Ewe-Volo’), while only 12 had identified themselves as ‘Akwamu’.222 Volo rulers were now insistent that the inhabitants of Frankadua were ‘Ewe’ who wished to live with their ‘kinsmen’
In 1953, the village chief of Frankadua, Kofi Tulasi ii, and a number of the community’s leaders won over the deputy of the Volta Region in the Gold Coast Legislative Assembly, P.D. Adjani, to their cause. They complained bitterly in Keta that ‘from time memorial [sic] we the inhabitants of Frankadua are Ewe speaking people’; that they felt like Ewe and were part of the ‘Tongu sub-group’ of their ethnic grouping; and that they wished to become part of a local council that only involved Ewe communities.224 British officials were sympathetic to the campaign, but the final decision was left to the Nkrumah government.225 Even nowadays, Frankadua retains its difficult legacy, after becoming a part of the Volta District in post-colonial Ghana. As recently as in 1999, ‘ethnic tensions’ provoked a near-battle between ‘Anlos’ (now meaning the Ewe-speakers of the town, which is somewhat curious) and ‘Akwamus’.226
Smaller groups like the Togome and Fodjoku followed the Volo and Dufor initiatives, and in November 1946 they obtained, after protests, the promise of the paramount chief of Akwamu to be integrated within the Tongu Confederacy. The question of symbolic allegiance to the Akwamuhene was, however, a problem.227 In a meeting before the Akwamu Native Authority, ‘traditions’ were mobilised. The Wirempihene challenged the Togome claim, criticising the group because they ‘speak Eve, yet…stay on Akwamu land’. According to this
British officials agreed that those groups ‘are quite clearly Ewes’ and ‘should join their brothers in the Tongu Confederacy’, but they struggled to understand the old Akwamu-Togome-Fodjoku alliance or the realities of local rule. Nkwanta, the Fia of Togome, insisted on the existence of a historical border with Akwamu; the Fodjoku added to the confusion by holding that the Togome were ‘Anlos’ and, hence, another type of ‘Ewe’. Also, the accounts of behaviour during the Akwamu invasion were very contradictory.229 At the end of the 1940s, the senior district commissioner in Ho was desperate: the Togome and Fodjoku communities were ‘Ewe speakers’ and wanted to ‘join their Ewe brethren’, but had various claims against each other. This led colonial officials to fear the worst for other, larger cases.230
‘Ewe’ solidarity in times of the impressively large, ethnically expressed group mobilisation of the aec and of similar organisations, was often only a part of the picture. Elsewhere, the old, local conflicts continued unaffected by the pan-Ewe idea. At the heart of the various ‘native states’, conflicts were often local: within the Tongu Confederacy, Tefle inhabitants claimed they had to defend themselves against Bakpa attempts at occupying their land; the paramountcy of the Ho Division – the central division of the Asogli Native State – was contested over decades between the villages of Dome and Bankoe.231 In the latter case as in many, ethnic arguments were sometimes employed, when convenient: thus the contenders from Ho-Bankoe accused the Howusu of being a member of a ‘stranger’ group that had only learned its proficiency in the Ewe language after its arrival in the Ho region.232
As we have seen, in the Adaklu community the break with Anlo had been an obvious goal until the end of the Second World War. However, after 1945 the Adaklu suddenly wished for a ‘return’ to Anlo rule. They referred to the ancient political Anlo–Adaklu alliance formed against Ho and ‘the other Ewe divisions’ and their old cooperation with the Asante invaders.233
The debate on Ewe-ness could even be employed as argument for local debates in non-Ewe-speaking regions, such as Buem. It was used in the conflict about local education policy and the creation of schools in the first half of the 1950s. Buem elites refused to send their children to Ho Secondary School, as they feared compulsory education in the Ewe language. Under pressure, the British officials had to assure that ‘no student is compelled to study Ewe’. However, as the conflict between Twi-speakers and Lefana-speakers escalated in the 1950s, the Ewe question faded into the background.234 In that respect, the Omanhene of Buem accused the Twi-speaking militants of ‘tribalising’ the conflict. Other, minority linguistic communities such as the Bowiri joined in the protest; and it also became connected to internal conflicts about the hierarchy of divisions and their leaders within the Buem Native State, but this was no longer about the Ewe.235
However, ‘Ewe mobilisation’ was used as a model outside of the Ewe-speaking areas. In the case of the tensions between the Lolobi and Akpafu in Buem, the Lolobi profited from their knowledge of the Ewe process to express political protest. Lolobi dignitaries argued that they should have been represented in the important debates at Lake Success in the United States, where the spokesmen ‘of the Ewe’ formulated their claims towards the trusteeship council of the United Nations. They claimed they did not, of course, have an ‘Ewe identity’, but they believed that even smaller ethnic groups should have the same right of mobilisation.236 Therefore, the Ewe example had an immense impact as a blueprint for local rhetorics.
The border continued to have its own dynamics. In 1943, the Fiaga of Bator had, in a land claim, described his opponents as ‘Kpele strangers from Eve’; we find no solidarity between ‘Ewe’ when land rights and immigration were involved. Frequently, Ewe-speaking local councils voted for the expulsion of Ewe-speaking immigrants from community land, and if the latter came from the other side of the colonial border, they would often be insulted as ‘French strangers’.237 By the same logic, in Agu Tafié, at the French side of the border in the Kpalimé region, the elders refused to accept the nephew of the retired paramount chief, Aboyo, as new ‘traditional ruler’, on the grounds that he had lived for the largest part of his life in the Gold Coast and thus adopted foreign manners!238 The colonial border could in that regard be exploited, in the perception of locals, beyond the ethnic solidarities.
In the region of Kpalimé, cut leaders attempted to restore the popularity of their movement, which had come under mounting pressure in 1950, through
Most of the conflicts in the Kpalimé area remained on the strictly local level. They essentially reflected problems with decisions taken in the pre-colonial period, such as, in particular, fusions of villages in the Kuma region. The inhabitants of Totsi and Yokele complained that, during the confusions of the Asante incursions, the leader of Tsame had usurped the chieftaincy over the sub-region of the canton. This was then connected to longstanding land conflicts between Tsame and Totsi, which went on from the 1930s to 1970. Other, internal, conflicts as in Kuma-Apoti, or in Agu-Apegame, were also linked to the traumatic experience of the Asante invasions, but not the Ewe legends.240 Finally, the fate of the chiefs of Agu Nyogbo Agbetiko was discussed during the whole of the late colonial and the post-colonial phases: in the 1950s, this particular conflict was linked to the battles between the ptp (still in territorial power) and the cut, hostilities that reappeared once again in 1970.241 As in
Here we refer to the notion of the Councillor of the French Union, Mr Savi de Tove, who defends a principle adopted as well by our representatives in the Local Assembly, which says the following: ‘The nomination of the Ewe chiefs by the administration is not in line with the indigenous customs. This is the reason for great troubles. From this motive, it is necessary to abolish the decree [speaking of the text from 1st March 1945, about indigenous rule in Togo] and leave the people their liberty to choose their chiefs to their convenience. This would be far more democratic’.242
In spite of such threats coming from local elites, most local inhabitants, including the Ewe-speakers and Mina-speakers, were not at all interested in creating a real conflict. When the chief was elected, the ‘stranger’ Adama Dali managed to maintain, with support of a group of elders, that he was the ‘traditional heir’ of the ruling family, and he celebrated a crushing victory over his Ewe-speaking
Therefore, the reference to Ewe identifications was subject to rather strong limitations. First of all, it obviously needed an ‘otherness’ present on the spot that could be defined as ethnically different, as against a ‘Fon’ in Fongbe, or between Ewe-speakers and the linguistic groups of Buem. Reference to Ewe-ness did not serve as a language of reconciliation in matters of conflicts between different smaller Ewe-speaking communities. Even in the territory of Togo under French mandate, which lacked other larger political entities to rely on, ethnic mobilisation did not, in the local practice, become a particularly important principle.
Ewe from Outside: The Avatime and the Question of Ewe Solidarity
To illustrate these points through a local case over the decades, we will now go back to the Avatime communities. The Avatime as speakers of a Central Togo language are, of course, at first glance distinct from Ewe-speakers in linguistic terms.244 Nonetheless, in the late nineteenth century the community’s mastership of the Ewe language was so outstanding that German Governor Jesko von Puttkamer remarked that, in contrast to other Ewe-speakers, the people of ‘Awatimé have the pure Ewe dialect’!245 As in the mixed and fluent communities of Joal-Fadiouth and Port Loko, Avatime settlements thus represented a local view on different options and cleavages at different times.
In the 1920s, Avatime informants redefined ‘historical tradition’ into a more independent narrative: only some of them had come from Notsie, the rest from Ahanta in the Gold Coast (setting them apart from the majority of Ewe communities). Also, according to this version, the Avatime had formerly been Twi-speakers, and were linguistically part of the groups west of the Volta River. This version seems to be a clear invention and is particularly curious.246 Under
European residents helped with this idea, as German missionaries tended to treat the Avatime as an ‘Ewe tribe’. African missionary personnel recruited for work in the area around Gebi Mountain were obliged to undergo training in Ewe, even if destined for the more isolated Avatime villages, such as Kolenin. Moreover, the growing group of Christian converts was, in this early phase, rather eager to accept an all-embracing Ewe culture (including the Notsie founding legend).247 Missionary attitudes in the 1900s were ambivalent and mixed with very practical issues. Consequently, as one example, the Bremen Mission recruited local recruits from Avatime, such as Godwin Banimanve, who were installed as auxiliary teachers in places like Amedzofe, because they were able to teach undergraduates in the local language. In 1890, Andreas Aku, a Bremen Mission catechist from Keta, described villages such as Amedzofe and Gbadzeme as bilingual. In Spieth’s account, the Avatime are implicitly considered to be one of the ‘Eweland tribes’. Around 1900, the Avatime thus had the two options of ethnic identification in their repertoire.248
After the First World War, the Avatime populations attempted to maintain the image of being a particularly ‘independent’ community, and as especially anti-German. They accused their neighbours, the Tafi, of having been the first to ‘defect’ to the Germans in the 1890s, and described them and the Logba as ‘weak’ as opposed to the fierce cruelty of their own warrior community; the Agotime and the Adangbe north-east of Lomé have the same origin legend that emphasises cruel acts during wars. It makes good sense that three Trans-Volta communities thereby underline their
The British ‘amalgamation’ project of the late 1920s brought the Avatime under stronger pressure to define their group identification in relation to neighbouring communities. They continued to reject Peki’s leading role, as the Peki had been, from that perspective, a weaker ally during the Akwamu and Asante incursions.250 Their discussion with British administrators did not yet touch on the question of ‘Ewe identity’. With regard to Anlo, the Avatime claimed that there had never been any tributary arrangements, but only trade relations between partner ‘states’ on equal terms.251
In the 1920s, Adzatekpor nevertheless became nominated an arbitrator in stool disputes within ‘Ewe’ communities. For the Wadze Stool Dispute, he also claimed to have ‘knowledge of the native customs’ of the Ewe.252 The Avatime chief managed to impress the British so much that in 1931 he was exclusively spared from confiscations to enforce the payment of debts. However, this relative success made the head chiefs of neighbouring divisions strongly suspicious of Avatime intentions: the divisional chiefs of Honuta, and, obviously, of Tafi were hostile to the creation of an ‘amalgamated state’. The Avatime reputation for violence did not help. In early 1928, Avatime warriors from Dzokpe destroyed the village of Tafi Atome, which antagonised not only the Tafi but also the Logba, and poisoned the atmosphere in the region.253
While Avatime informants presented the Tafi as spoilers of peace agreements, the rulers of the other communities were infuriated by the Avatime attack and refused any future cooperation. In June 1930, Adzatekpor V finally gave in and agreed to the restitution of pillaged Tafi possessions and to legal arbitration from Accra. This also opened discussions on identifications. The villagers of Tafi stated that they were the autochthons on the land in question, and the Avatime held that their Dzokpe branch had ‘spear-won’ the same land. Some reference was made to the Notsie legend, but none to ‘Ewe’ identity.254
In 1932, ‘Awatime State’ nonetheless became created as a larger political unit; this new ‘native state’ promoted by the British was an entity that included many Ewe-speaking communities including former adversaries. In the 1930s, the new State Council began to engage in questions of infrastructure, land use and heritage in Ewe-speaking places like Dudome or Izola. However, Adzatekpor, as president of the ‘native state’, was regularly shunned by the fiawo of the different Ewe-speaking communities in the state who preferred to apply directly to the British authorities.255 Awatime State remained a loose alliance of practically equal partners, in which the rulers of Avatime did not really manage to mark any dominance.
With regard to group identifications, the Avatime and other members of the ‘native state’ were hostile to immigrants regarded as ‘Peki’. The question of making Philip Tekedu the headman of the ‘Pekis’ of Honuta was thus controversially discussed; the communities feared that this headman would act as a
During the 1930s, Adzatekpor’s rule over the core lands of Avatime became shaky. The paramount chief had reached a peak of unpopularity by introducing the head tax, which was attacked by many elders and other opponents as a return to practices from German times.258 The resistance of the ‘youngmen’ with open and outspoken support by many elders became in the end untenable, and the stool father, Traugott Tekpe, even called for the destoolment of Adzatekpor in November 1937.259 In 1938, the Avatime chief was no longer being invited to the court sessions of his own State Tribunal, and he was replaced there by the chief of Amedzofe. However, neither the British nor the members of the other ‘divisions’, agreed to Adzatekpor’s destoolment. In 1939, an uneasy peace was brokered between Adzatekpor V and the ‘youngmen’ engaged in a symbolic cleaning procedure of the Vane-Dzolo-Kpoeta road.
The emergence of the pan-Ewe campaign, in 1942, came at the right time. It gave a new opportunity to Adzatekpor to change the odds, and the tone of the Avatime paramount chief became different. He appeared at the forefront of the Ewe unity adherents, where he emphasised the necessity of Ewe reunification. After seven decades of a discourse insisting on Avatime’s distinctiveness, this community was suddenly no longer different from the ‘Ewe’. Adzatekpor V demanded Ewe unification on the grounds that the borderline ‘deprives us, the Ewe from our privileges for trade as due to the effects of the war of 1914’ and he wished ‘that we the Ewes enjoy our former privilege for harmony of our works in Togoland’.262 The ruler obviously tried to satisfy some of the divisional chiefs inside Awatime State, but also to profit from the widespread enthusiasm. The British official in charge in Ho commented in 1945 that Awatime State as an entity was indeed ‘entirely Ewe’!263 Over
This surprising flexibility of ‘tradition’ – which demonstrates the particular usefulness of our approach of ‘colonial history on the ground’ through written accounts – can notably be compared with Agotime. Like the Avatime, the members of this community normally presented themselves as non-Ewe, but could appear as part of ‘the Ewe’ whenever this was useful. In the 1920s, ‘traditions’ showed a clear sense of flexibility with regard to relations between Agotime and Ewe-speakers.265 Some of these ‘traditions’ insisted there had been Ewe settlers as part of the original Agotime community foundation, and argued that the schism between the Agotime and the Howusu of Ho had appeared through a misunderstanding during the Asante war. Mahumasro as a chief and candidate for the stool of the head chief in 1932 probably had his own reasons in promoting such a version, which became accepted.266
Therefore, members of the so-called Central Togo Minorities or the Adangme-speaking Agotime were quite capable of managing their identifications according to the regional necessities. In the late nineteenth century, the political entity was the central point of reference. However, these entities suffered under the German intrusion, and the Awatime Native State was only a shadow of the pre-colonial states. During the Ewe unification campaign, these chiefs supported the grand project, although their communities had over the years quite often insisted that they were not Ewe. Where the interest of the ruling dynasty was at stake, flexibility in identification was very possible.
Outlook: Political Ewe-ness in the Ghana Volta Region, Neutralised Ewe-ness in Togo
British Togoland lost its status as a separate territorial unit in 1957. In spite of more than twelve years of Ewe unification campaigns, and of some ten years of Togoland reunification propaganda, the cpp was successful against the programme of the Togoland Congress (tc). Some of the local chiefs participated actively in the defeat of the tc, such as in the case of Anfoega where Togbe Tepre Hodo intimidated tc activists and organised riots against tc electoral campaign meetings.267
After the cpp victory in the Trans-Volta Togoland Region, the Nkrumah government started to become more relaxed about Ewe issues in its rhetoric, and even to invoke ‘Ewe solidarity’ whenever this was useful. With regard to communities like Mafi, Awudome, Fodjoku, and Togome, Nkrumah’s cabinet members urged politicians and chiefs from Togoland to consider more cooperation with ‘Ewes in the Gold Coast’. This could now easily be used as an argument to refuse political reform.268 Nevertheless, and in spite of the electoral victories of the cpp in the tvt Region, the representatives of the new Ghanaian administration remained distrustful. In Ho, in the first half of the 1960s, the majority of civil servants in place were deliberately not taken from the group of Ewe-speakers.269 Such decisions alienated former supporters in the area from the cpp politics. The idea that particular ‘forces’ in the region worked for secession and armed resistance existed over decades.
In Togo, there was a similar uneasiness about a possible arms trade and a possible Ghanaian invasion in the region of Kpalimé.270 However, the post-colonial centralised state focused on controlling the local communities and their chiefs through the préfets (district commissioners), and on local distribution of resources. These conditions played against ethnic allegiance. In the canton of Agu, the seat of the paramount chief was removed from Kebu Dzigbe to Toubadji.271 In the neighbouring canton of Kuma, the long-standing
The category of ‘Ewe’ retained some of its importance in land issues and similar administrative practices. This can be shown in the numerous conflicts in the Kpalimé region of Togo during the 1960s and 1970s, as during those between Village Chief Léléklélé ii and the youth association of Agbétiko.273 Similarly, when in 1972 an alliance of dignitaries and elders attempted to remove Théophile Dom Gamety V as paramount chief (chef de canton) of Kuma, they claimed that among other alleged misdeeds he had betrayed common Ewe origins. The fact that Dom Gamety V had not participated in the yearly festivities in Notsie, was employed to demand his retirement.274 This shows that sensitivities with regard to Ewe culture could indeed be employed under particular circumstances.
However, most group conflicts in Gbassinje Eyadéma’s Togo remained concentrated on the local level, including on issues of identification and group mobilisation. In the context of migration to Lomé, locals organised in associations of originaires. As in the case of Agu Nyogbo Agbetiko, these associations would attempt to influence developments in their village community of origin, but would no longer represent a larger group idea. Ewe-ness was unimportant in comparison to local affiliation.
Amongst the Ewe-speakers, ethnic solidarity was one option of mobilisation. It became more attractive as pre-colonial political entities such as Peki or Anlo often did not manage to offer a convincing alternative. Even so, it remained in the background for a long period, as ‘divisional’ group identities continued to be the principal form of group idenfications in the Trans-Volta Region. The pan-Ewe mobilisation of the 1940s was a spectacular exception. However, it had a surprisingly low impact on the long-term situation, as it did not enduringly eclipse local orientations.