Landscapes of Power: Ownership and Identity on the Middle Kavango River, Namibia*

Abstract

This article examines traditions of origin among riparian communities living in a semi-arid environment as a means of exploring how the environment and socio-political identities are mutually constituted. Within Namibian Kavango communities over the past half century, such traditions have fallen into two general categories: those of the royal clans (accepted by most residents) which state that they were the first to settle this section of the river, and those of other groups who claim that their own clans were the real first-comers but were later subordinated by the immigrant kings. Each group of traditions claims prior origins from similar riparian environments as a means of demonstrating how their group was able to tame the wilderness. Indeed, both the royalist and the indigenous traditions rest on shared assumptions about how ownership is established: through first-comer status and through creating ‘country’, or a place fit for human communities, out of ‘bush’ or wilderness. Autochthon–immigrant relationships were probably long fraught with tension here, as they were in other parts of southern Africa. Embedded within some of the indigenous traditions are hints of ‘charters’ establishing relations between the two groups. However, the colonial and postcolonial experiences have resulted in the erasure of autochthonous communities from royalist historical narratives. Chiefs needed to prove ownership of their communities to colonial officials seeking to impose a form of indirect rule, and colonial and postcolonial interventions into riparian communities affected all residents and broadened the definition of indigeneity to include all Kavango people. Indigenous traditions survive to some extent within the community, but they are increasingly irrelevant in a political landscape in which everyone has become an indigene at the mercy of strangers from afar and in which ownership is no longer properly exercised by anyone.

Notes

^*Research for this project was supported by fellowships from the National Endowment for the Humanities and the Woodrow Wilson International Center for Scholars as well as grants from Georgetown University. I especially wish to thank Kletus Muhena Likuwa and Michael Uusiku Akuupa for their invaluable assistance in conducting interviews in 2005 and 2006–07.

¹ A.G. Stigand, ‘Ngamiland’, The Geographical Journal, 62, 6 (1923), pp. 410–11, discusses the highly localised nature of river names, and speculates that the ultimate origin of ‘Kavango’ came from its upper course in Angola as no variants of the name existed in the lower course.

² J. Poroto and T. Kupembona, interviewed, Gumma village, Namibia, 18 July 2005; S. Hausiku, interviewed Sauyemwa village, Namibia, 21 July 2005.

³ The best summary of the river's hydrology and surrounding terrestrial environment is found in J. Mendelsohn and S. el Obeid, Okavango River: The Flow of a Lifeline (Cape Town, Struik Publishers, 2004).

⁴ While the Namibian section of the Kavango is often conceived as part of the lower Kavango, in this article I refer to it as the ‘Middle Kavango’ in recognition of the distinctive human-environmental relationships that emerged there, centred as it was between upstream population clusters in much wetter environments and the downstream Delta communities.

⁵ The Kwito, its major tributary, joins the Kavango in Namibia, but it too receives no water from this area and is simply bringing water from rainier points to the north.

⁶ The presence of these fertile riverbeds was recognised by European visitors quite early, indicating that in the mid and late nineteenth century, settlement patterns very much adhered to these patterns of soil fertility. See E. Wilmsen (ed.), The Kalahari Ethnographies (1896–98) of Siegfried Passarge, Nineteenth-century Khoisan- and Bantu-Speaking Peoples (Köln, Rüdiger Köppe Verlag, 1997), p. 288. When political conditions in Angola have permitted, these are favoured sites for cultivation. On the heavily settled southern bank of the Kavango in Namibia, however, these soils have been altered – for the worse – through the introduction of ploughing in recent decades. This has mixed infertile layers of subsoil with the fertile but thin layer of topsoil, resulting in a decline in overall fertility. Mendelsohn and el Obeid, Okavango River, pp. 41–45.

⁷ See, for example, L. Jacobson, ‘The Archaeology of the Kavango’, Journal of the SWA Scientific Society, 40–41 (1987), pp. 155–56.

⁸ J. Vansina, How Societies Are Born: Governance in West Central Africa Before 1600 (Charlottesville, University of Virginia Press, 2004).

⁹ Themes of loss and nostalgia may be common in oral traditions dealing with environmental history. See for example T. Giles-Vernick, ‘Doli: Translating an African Environmental History of Loss in the Sangha River Basin of Equatorial Africa’, International Journal of African Historical Studies, 32, 2/3 (1999), pp. 373–94 and J. McGregor, ‘“Living with the River”: Landscape and Memory in the Zambezi Valley Northwest Zimbabwe', in W. Beinart and J. McGregor (eds), Social History and African Environments (Oxford, James Currey, 2003), pp. 87–106. Enfield and Nash refer to this tendency as ‘environmental nostalgia’ and note that it operated among missionaries and local people alike (albeit for different reasons). G. Enfield and D. Nash, ‘Drought, Desiccation and Discourse: Missionary Correspondence and 19th-century Climate Change in Central Southern Africa’, The Geographical Journal, 168, 1 (March 2002), pp. 33–47, especially pp. 40–1.

¹⁰ A. Fleisch and W. Möhlig, The Kavango Peoples in the Past: Local Historiographies from Northern Namibia (Köln, Rüdiger Köppe Verlag, 2002).

¹¹ B. Sandelowsky, ‘Kapako and Vungu Vungu: Iron Age Sites on the Kavango River’, South African Archaeological Society Goodwin Series, 3 (1979), pp. 52–61, found early Iron Age pottery in conventionally Bantu styles dating back to perhaps the ninth century, and dated artefacts clearly related to the current inhabitants' material culture to the seventeenth century. Also see J. Kinahan, ‘Settlement Patterns and Regional Exchange: Evidence from Recent Iron Age Sites on the Kavango River, North-Eastern Namibia’, Cimbebasia, 3, 4 (1986), pp. 110–16; Jacobson, ‘Archaeology of the Kavango’.

¹² G.T. Nurse and T. Jenkins, ‘The Kavango Peoples’, Journal of the SWA Scientific Society, 30 (1976), pp. 55–58.

¹³ K. Klieman, ‘The Pygmies Were Our Compass’: Bantu and Batwa in the History of West Central Africa, Early Times to 1900 C.E. (Portsmouth, NH, Heinemann, 2003), pp. 68, 74; see also I. Kopytoff, The African Frontier: The Reproduction of Traditional African Societies (Bloomington, Indiana University Press, 1987).

¹⁴ This argument has been made in various instances of autochthonous-immigrant relationships, perhaps most sweepingly by Klieman, ‘Pygmies’. See also D. Gordon, Nachituti's Gift: Economy, Society and Environment in Central Africa (Madison, University of Wisconsin Press, 2006); D. Lan, Guns and Rain: Guerrillas and Spirit Mediums in Zimbabwe (Berkeley, University of California Press, 1985); C. Crais, The Politics of Evil: Magic, State Power, and the Political Imagination in South Africa (Cambridge, Cambridge University Press, 2002); M. Salokoski, How Kings are Made, How Kingship Changes (Helsinki, University of Helsinki Research Series in Anthropology, 2006).

¹⁵ Klieman, ‘Pygmies’, p. 74.

¹⁶ E. Kreike, Re-Creating Eden: Land Use, Environment, and Society in Southern Angola and Northern Namibia (Portsmouth, NH, Heinemann, 2004); Salokoski, How Kings Are Made, p. 69.

¹⁷ Klieman, ‘Pygmies’; Gordon, Nachituti's Gift.

¹⁸ All of these narratives were recorded on the Namibian side of the border, for the simple reason that, by the 1950s, relatively few people appear to have been left on the Angolan side – and those few crossed during the era of the liberation wars beginning in the 1960s. By the 1980s, the Angolan side of the river was virtually depopulated. Residents from the Kavango banks were not seen as ‘refugees’ per se, but rather as people who had never broken ties with their particular polity. This movement from Angola to Namibia is reflected in the large number of people I interviewed who were born on the Angolan side of the river.

¹⁹ See, for example, the traditions recorded in Fleisch and Möhlig, Kavango Peoples in the Past; also G. Gibson et al., The Kavango Peoples (Wiesbaden, Steiner, 1981). Traditions in the two western, upstream polities of Mbunza and Kwangari (closely related groups) recount two possible points of origin: either the Linyanti region to the east or a point to the northwest, where the Kunene, Cuvelai and Kavango rivers flow close to each other. See Fleisch and Möhlig, Kavango Peoples in the Past, pp. 179–80; Gibson et al., Kavango Peoples, p. 38.

²⁰ Fleisch and Möhlig, Kavango Peoples in the Past, pp. 37–54.

²¹ Some of the traditions, not surprisingly, suggest ultimate origins from places further away, but the Zambezi link is most clearly and consistently emphasised.

²² Vansina, How Societies Are Born, pp. 183–6, notes the contradiction between the linguistic and oral evidence. He says that Kwangari, Mbunza, Shambyu and Gciriku communities all speak a language that is most closely related to Umbundu, although none of their traditions posits historical links with that area. But he does not really explore the reasons for the contradiction, in part because he is not particularly interested in how Middle Kavango communities view their own identity. Rather, he (correctly) sees the area as a long-time magnet for refugees from all directions, as evidenced by a mix of political and linguistic influences from many places. But his conclusion, that the area is marked by ‘a mix of impoverished reminiscences of rich features that are typical for one or another of the surrounding regions but nothing original to the region itself’ with ‘no original forms of social institutions’ (p. 185), is likely to offend most Kavango residents and certainly flies in the face of their own historical experiences.

²³ M. Kandambo, interviewed Katere village, Namibia, January 2007.

²⁴ This is a common opposition in Bantu-speaking African communities, from as far away as the central African forest (Klieman, ‘Pygmies’) to neighbouring Ovambo communities in Namibia and Angola, as described in Kreike, Re-Creating Eden.

²⁵ M. Kandambo, interview, January 2007.

²⁶ All names in these traditions, including all place names, appear as Bantu in origin with the sole exception of ‘Tjaube’. Other versions of this tradition draw a distinction between Bushmen and the Tjaube, arguing that the Tjaube preceded Bushmen into the area.

²⁷ The entire Chronicle is reprinted in Fleish and Möhlig, Kavango Peoples, pp. 29–56; a very similar version was recounted by D. Kashera, interview, Kambowo village, 20 January 2007. Iron-working is attested at these sites by archaeological evidence; the tradition also refers to a wetter climate phase in the region since it argues that routes were traversed by boats that would be impossible today.

²⁸ D. Kashera, interview, 20 January 2007.

²⁹ Fleisch and Möhlig, Kavango Peoples in the Past, pp. 48–51.

³⁰ S. Hausiku, interview, 21 July 2005.

³¹ Sandelowsky, ‘Kapako and Vungu Vungu’, Kinahan, ‘Settlement Patterns’.

³² L.L. van Tonder, ‘The Hambukushu of Okavangoland: An Anthropological Study of a South-Western Bantu People in Africa’ (PhD Dissertation, University of Port Elizabeth, 1966), pp. 44–50, has the most detailed version of this migration tradition.

³³ See, for example, A. St. H. Gibbons, Africa from South to North through Marotseland, Vol. 1 (London, John Lane, 1904), pp. 217–18.

³⁴ G. Gibson, T. Larson, C. McGurk, The Kavango Peoples (Wiesbaden, Franz Steiner, 1981), pp. 99, 163, asserts the absence of people in Shambyu and Gciriku settlement zones, while the histories of the Kwangari, Mbunza and Mbukushu essentially assume the absence of prior inhabitants. This study argues that //anikhwe and Yeyi-speaking inhabitants along this stretch of Kavango are more recent arrivals, whereas the indigenous traditions argue for them as first-comers (p. 23). Some Shambyu traditions describe prior Bantu-speaking inhabitants along this stretch of the river, claiming that they bought a river valley from Mbunza who were already settled upstream. In Fleisch and Möhlig, Kavango Peoples in the Past, the Chronicles of the Shambyu, Gciriku, Kwangari and Mbunza all leave the impression of uninhabited territory upon the founders' arrivals.

³⁵ A. Sipandeka and S. Kandere, interview, Kaisosi village, 20 July 2005; Poroto and Kupembona interview, 18 July 2005.

³⁶ National Archives of Namibia, ‘unpublished report by the Kavango Native Commissioner’, 4 January 1963. I wish to thank Mattia Fummati for giving me a copy of this document.

³⁷ D. Kashera interview, 20 January 2007; also S. Kandere, interview, Rundu, 28 January 2007.

³⁸ Fleisch and Möhlig, Kavango Peoples in the Past, pp. 51–2.

³⁹ Ibid., p. 54. This is in direct conflict with the Mbukushu tradition recorded in Van Tonder, ‘Hambukushu’, p. 49, which argues that part of the //anikhwe community were made slaves of the Mbukushu while others moved away; earlier parts of Van Tonder's version do however indicate an earlier relationship through marriage of chiefs and autochthons, p. 45.

⁴⁰ There are many indications that none of these polities was ethnically ‘pure’. The Tjaube Chronicle is one; another is the findings of a South African researcher in 1960 who concluded that nearly one-third of those in Shambyu belonged to the vaKwandjadi clan, making it the largest matrilineage long after its conquest. J.L. Bosch, ‘Die Shambiu van die Okavango, ’n Volkekundige Studie’ (PhD, University of Stellenbosch, 1964), p. 130. Similarly, there are stories of Gciriku men being wiped out in a punitive expedition led by Europeans and being replaced by migrants from other Kavango communities.

⁴¹ Fleisch and Möhlig, Kavango Peoples in the Past, pp. 53–4.

⁴² Fleisch and Möhlig, Kavango Peoples in the Past, pp. 132–75.

⁴³ Kandambo, interview, January 2007: the interviewee and her daughter had an extensive debate about the historical role of the Kwandadji clan. The interviewee, the mother of the hompa in Angola, noted that the Kwandadji clan was her father's clan, indicating that perhaps marriage patterns between royal families and purported autochthons do exist in this area.

⁴⁴ Fleisch and Möhlig, Kavango Peoples in the Past, pp. 137–8.

⁴⁵ I.J. Haididira, interview, Kauti village, 31 December 2006.

⁴⁶ The Mbukushu royal title was fumu, again reflecting distinctive origins of their language and political institutions.

⁴⁷ Vansina, How Societies are Born, p. 185. He does not deal with Mbukushu kingships.

⁴⁸ The only figure who seems to have asserted any real authority over life and death was the Mbukushu fumu, and his seems to have been restricted to sacrificing members of his family as part of the rain-making ceremonies. When the fumu took to selling his subjects into slavery in the late nineteenth century, they deserted him in large numbers, moving downstream to the northern regions of the Okavango Delta.

⁵⁰ Fleisch and Möhlig, Kavango Peoples in the Past, p. 53.

⁴⁹ Salokoski, How Kings Are Made, pp. 88–92, has carefully examined Oshiwambo oral traditions to the west and has also concluded that Bushmen were erased from royal histories. She assumes this happened at the very beginnings of the consolidation of royal power, despite the fact that traditions of crucial San roles in the creation and maintenance of royal authority survived into the 1960s.

⁵¹ Quoted in A. Eckl, ‘“Reports from ‘Beyond the Line’”: The Accumulation of Knowledge of Kavango and its Peoples by the German Colonial Administration 1891–1911’, Journal of Namibian Studies, 1 (2007), pp. 21–2.

⁵² In interviews, some people said kings did sell local people into slavery, but portrayed this as a partnership between the chiefs and families looking to get rid of their troublesome members. D. Kashera, interview, 20 July 2007.

⁵³ Many of these monarchs had their courts on the Angolan side of the river when Europeans first arrived but relocated to the Namibian side in the early twentieth century as the Portuguese consolidated their rule.

⁵⁴ K. M. Likuwa, ‘Rundu, Kavango: A Case Study of Forced Relocations in Namibia, 1954 to 1972’ (MA thesis, University of the Western Cape, 2005), p. 109, argues that under colonialism chiefs went from being owners of the land to being salaried officials under a government that owned the land – and that the chiefs did not fully realise this. But this alliance, while certainly forced upon the chiefs, did provide an external authority to back up their own power locally over their subjects.

⁵⁵ This movement, too, is conceptualised within a framework of loss today, as informants report that the Angolan side of the river is far more fertile and was usually the preferred site for permanent settlement prior to Portuguese colonial abuses.

⁵⁶ W. Werner, draft report, WERRD October 2003 report.

⁵⁷ Eckl, ‘Reports’, p. 16.

⁵⁸ Poroto and Kumpembona, interview, 18 July 2007.

⁵⁹ Likuwa, ‘Rundu’, is a study of forced relocations and removals – beginning with a 1950s-era relocation from a flood-prone zone and continuing through the liberation war of the 1960s and 1970s.

⁶⁰ Quoted in Likuwa, ‘Rundu’, p. 95.

⁶¹ Sipandeka and Kandere, interview, 20 July 2007.

⁶² Sipandeka and Kandere, interview, 20 July 2007

⁶³ Sipandeka and Kandere, interview, 20 July 2007; Mbunza Hompa A.K. Mattias, interview, Sigone village, 20 July 2005; Kandambo, interview, January 2007; Poroto and Kumpembona, interview, 18 July 2005.

⁶⁴ K. Mberekera, interview, Kambowo village, January 2007.

⁶⁵ S. Kandere, interview, 20 July 2007.