yet, the ostensible promise of ‘protection’ represented no definite guarantee against annexation. In the wake of the South African war, the British government ‘began to press for the transfer of the Protectorate to the control of the nascent Union of South Africa’ (Ramsay 1998: 82). Once again the dikgosi headed campaigns against such plans, and in 1909 they made a second journey to London, where ‘the Batswana leaders’ views had an immediate, if immeasurable, effect on public discussions'. The Times remarked on their ‘skill in elocution’, concluding that ‘[t]he speeches of these barbarian chiefs…are far better reading than the speeches of most European statesmen’.13
Although the success of the Tswana dikgosi certainly owes considerably to their personal capacity, it must be stressed that their endeavours were decisively conditioned by the conjunction of particular historical circumstance. First, the strength of their respective polities, centred round large, well-organized royal towns and their extent of control over many outlying communities perfectly matched the British strong desire of running the protectorate at minimal cost by implementing extensively practices of indirect rule from the outset by leaving ‘the traditional organization very much alone’ (Ashton 1947). Second, the dikgosi's development of a close, supportive attachment to the London Missionary Society facilitated the development of the British-Tswana relationship. Third, the new orientations emerging amongst ruling groups of the Bakwena, Bangwato and Bangwaketse underpinning the dikgosi's efforts to ensure British protection, was conditioned by the threat of being captured into the violent domains of the expanding settler regimes immediately east and north of their own territories.
Finally, despite their capacity of incorporating and assimilating foreign groups, these processes were, however, not complete: during the latter part of the eighteenth century a number of communities, some of which large and strong, were located within the territory claimed by one of the three Tswana merafe. As we shall now see, with a privileged position within the colonial state, the dominant position of their ruling communities in relation to other communities was progressively reinforced.
Expansion of the Dominant Tswana Merafe under the British Wing
After the Bechuanaland Protectorate had been established in 1885, the eight Tswana merafe officially recognized by the British were each designated a distinct, demarcated territory, known as ‘native reserve’ (see Schapera 1943a: 7ff.; Motzafi-Haller 2002: 86ff.).14 A kgosi was established as the ‘native’ authority in charge of all the peoples living in each of them. Furthermore, during the three dikgosi's visit to England in 1895, the British made it clear that they would be given full backing, if required, to prevent junior sections of the ruling dynasty from branching off and forming separate establishments. The rulers of these merafe were hence installed by the British as the supreme authority of all those communities living within their respective reserves. This also meant, as suggested previously, that groups that had seceded from the royal centre and taken residence elsewhere, were now subjected to the overlordship of the Tswana kgosi recognized by the British (Schapera 1952: 17). The dikgosi were assured full support in case subject communities should challenge their overlordship. In this section I shall explain how the dominant Tswana expanded the network of power with their respective reserves and how the increasingly repressive structures were countered by resistance.
As suggested in the Introduction, there were huge differences among the eight ‘native reserves’, both in territorial range (see Map 1) and population size (see Schapera 1952). Moreover, they differed sharply in the extent to which foreign groups were assimilated into a recognized morafe.15 The three smallest ones – the Barolong, the Malete and the Batlokwa – occupied minute areas and were quite homogenous, with no unassimilated groups of any significance except for a small category of servants (batlankha). The Bakgatla were also relatively homogenous, small in number and confined to a relatively small territory, yet, as we shall see, with one distinct Tswana community with non-Kgatla identification. Of the four larger territories – the Bangwaketse, the Bakwena, the Bangwato and the Batawana – the first two included some sizeable communities which identified themselves as Tswana, but originated from sections different from the three ‘hosting’ merafe. In addition, a considerable number of people were scattered around in small villages or hamlets: as discussed earlier in this chapter, these consisted partly of small groups originating from Sotho-speaking communities in the Transvaal, and partly of people who had ‘always’ been living in the area and who were classified by the ruling Tswana groups as Makgalagadi or Masarwa.16 When the dominant Tswana communities increased their cattle wealth from the eighteenth century onwards, the latter categories constituted a source of free labour and consequently were exploited as herders. This was particularly the case among the four large merafe. The Bangwato and the Batawana were granted the most extensive ‘native reserves’; they also were assigned overlordship of great numbers of people who had retained their non-Tswana cultural identities.
By the exercise of their dominant position, all communities within the confines of each native reserve were captured into the hierarchical structure of authority radiating from the royal centre of the officially recognized merafe. The Tswana ruling group exercised their power by means of this network: formally on behalf of the Crown, they administered the allocation of natural resources, collected tax and exercised jurisdiction. The heads of provincial communities were recruited either from the local group or from members of the royal centre who were placed in provincial communities to embody the authority of the kgosi officially recognized by the colonial state. During the colonial era such delegation by the dikgosi increased, with a system of ‘chief's representatives’ originating from the royal centre and also an extensive intelligence service.
This system worked more efficiently in compact communities, especially those closer to the royal centres, than in peripheral communities in large ‘reserves’ such as those of the Bangwato and Batawana. On the other hand these peripheral communities, composed mainly of hunter-gatherers and horticulturalists, were politically inactive and thus represented no threat to the Tswana rulers. Moreover, although the majority of the population in the reserves lived in compact villages, being agriculturalists and pastoralists they operated over a wide area: a family's arable land could be located 500 kilometres from their village and their cattle post even further away. As all land resources were ‘communal’, and under the trusteeship of the kgosi access to land could be obtained only with the consent of the kgosi or one of his deputies, by appointing ‘overseers’ to administer all the areas used for agricultural and pastoral purposes (see Schapera 1943a: 143, 224ff), the dikgosi kept control of this particular part of the hierarchical network.
It needs to be stressed, however, that communities foreign to the Tswana merafe were brought under their domination to a highly different extent. This had partly to do with space. For example, while one Kalanga group was so closely connected to the royal centre of the Bangwato that it constituted a separate ward in the royal town, other Kalanga groups were living far away at the fringes of the territory which the Bangwato attempted to control, allowing them considerable de facto autonomy (Werbner 2004: 37; cf. van Binsbergen 1994: 156; Schapera 1952: 65ff). Partly this was a matter of power: while most ‘foreign’ Tswana-speaking communities were tightly integrated with the royal centre, there were some large Tswana communities which, as we shall see subsequently, were not located far from the royal centre yet always challenged the supremacy of the ruler of the ‘hosting’ morafe.
In any case, the domination of the Tswana merafe recognized by the British prevailed. The force of the processes and structures of subjection is testified by the fact that during the colonial era – spanning some eighty years – there were very few rebellions or serious conflicts between the dominant Tswana groups and their subject communities. It also reflects the fact that the leaders of these subject communities often perceived some benefit in submitting to the dominant morafe. For example, in 1922 a large contingent of Herero (originating from the present Namibia) sought refuge among the Bangwato and were accommodated by the establishment of a separate ward in 1922 (see Durham 1993: 72, 130–31).
