Victorian Field Notes from the Lualaba River, Congo

ABSTRACT

This article examines nineteenth-century British documentation on the Central African village Nyangwe and its inhabitants produced by the three British explorers (David Livingstone, Verney Lovett Cameron, and Henry Morton Stanley) who visited this settlement in the 1870s. In doing so, the article has four principal objectives. First, the article extends recent work on nineteenth-century African urbanization to consider how such urban history might be written with recourse to representations of Nyangwe in the archive of Victorian exploration. Second, the article highlights some of the issues inherent in using the archive for this purpose. Third, the article argues for a new methodology that uses recent historical, anthropological, and linguistic scholarship to link Victorian exploration literature to the local African circumstances out of which that literature emerged. Finally, the article draws on this new methodology to examine Victorian explorer representations of the Wagenya, an African tribe inhabiting the Lualaba River in the vicinity of Nyangwe, and to consider the broader influence of this tribe on the development of Victorian geographical representation and knowledge.

KEY WORDS:

• Congo
• Zanzibar
• Livingstone
• Stanley
• exploration
• darkness

Acknowledgements

This article is dedicated to Dane Kennedy in appreciation of his scholarship and in gratitude for his support. The author wishes to thank the following individuals, all of whom assisted with research for this article: Tansy Barton, Roy Bridges, An Cardoen, Karen Carruthers, Gary Clendennen, Ian Cunningham, Lutgard Doutrelepont, Felix Driver, Debbie Harrison, Judith Harrison, Michael Hawkins, Dorothy Helly, Nancy Hunt, Tim Jeal, Chris Lawrence, Mathilde Leduc-Grimaldi, Anne Martin, David McClay, Sharon Messenger, Alison Metcalfe, Sara Duana Meyer, Caroline Overy, Allen Roberts, Aldwin Roes, Kate Simpson, Sarah Strong, Herbert Tucker, Tom Turner, Stuart Whittaker, Devin Zuber, and the members of the VICTORIA, SHARP-L, and Digital Medievalist listservs. In addition, research for this article directly led to the development of the Livingstone Spectral Imaging Project ([Online] Available at: http://livingstone.library.ucla.edu/), the results of which, in turn, enabled the author to complete research for the present article, so the author would also like to thank the many individuals acknowledged for their support of that project ([Online] Available at: http://livingstone.library.ucla.edu/1871diary/acknowledgments.htm). The author is also grateful for questions and observations presented by the audiences who attended talks based on research for this article at the Sixth Conference of the International Society for Travel Writing (University of South Carolina, Columbia, 24 September 2010), the CUNY Graduate Center Victorian Seminar (6 December 2010), the Departments of English at Indiana University of Pennsylvania (1 March 2011) and New Mexico State University (28 January 2011), and the Department of Liberal Arts at Savannah State University (10 May 2011). Finally, the author is indebted to two anonymous readers for Scottish Geographical Journal for their thoughtful and detailed feedback.

Notes

1 All references to Livingstone's 1871 Field Diary ([1871a] 2011--2013) refer to the ‘documentary edition’ ([Online] Available at: http://livingstone.library.ucla.edu/1871diary/transcriptions.htm). A critically edited version is also available from this same source ([Online] Available at: http://livingstone.library.ucla.edu/1871diary/three_versions.htm).

2 The number of miles given here is only a rough estimation. The Distance Measurement Tool on Google Maps ([Online] Available at: https://maps.google.com/) provides a distance of about 238 miles between these two locations. Google's mapping of the modern-day locations of Nyangwe and Ujiji, however, is approximate, and Livingstone's travels, of course, would have taken him on a more indirect route. Livingstone himself ([1871f] 1872, p. 11) estimates the distance as ‘between 400 and 500 miles’, but, given the numbers above, this estimate seems exaggerated.

3 Scholars continue to debate the exact date of this meeting. See the ‘The Date of the Livingstone-Stanley Meeting’ in the 1871 Field Diary critical edition (Livingstone [1871a] 2011–2013) for the latest addition to this debate. Scholars have also questioned whether Stanley actually said ‘Dr. Livingstone, I presume? ’ in greeting Livingstone (Jeal 2007, p. 116ff.).

4 For more on Livingstone's representation of the Nyangwe massacre, see the 1871 Field Diary critical edition ([1871a] 2011--2013) and the extensive critical materials included, particularly the section on ‘Livingstone in 1871’.

5 The original inhabitants of Nyangwe were the Wagenya (Stanley 1878, vol. 2, p. 123). Livingstone (1872d) writes that ‘no trader or traveller ever ventured [into the region in which Nyangwe lies] till two years ago’. Cameron (1874a, 7 August 1874) states that ‘[The Arabs] have been 20 years in the country + settled at [Nyangwe] for 4 or 5’, while Stanley (1878, vol. 2, p. 117) notes that the quarters of Mwini Dugumbi in the northern section of Nyangwe represent ‘the first Arab arrival here (in 1868)’. Stanley's field notebook (1876–1877a, entries after 25 October 1876) reiterates this information, while also providing the most extensive early Arab history of the settlement. This history indicates that Dugumbi's slaves actually preceded him to Nyangwe. Most curiously, Stanley's field notebook suggests that the supposed practice of cannibalism among the African tribes settled in and around Nyangwe first developed as a strategy of resistance to one of the more violent Arab pioneers at Nyangwe, Mwini Makaya: ‘[The inhabitants] proclaimed themselves man eaters with ostentation. They slew one or two of [Moeni Makyeah's] people + ate them’. Despite his extended stay at Nyangwe, however, Livingstone himself failed to witness any cannibalism firsthand – ‘I offered a large reward in vain to any one who would call me to witness a cannibal feast’ – and concluded that the inhabitants ‘eat only enemies killed in war’ (Livingstone [1871e] 1872, p. 9). In an unusual turn of events, Livingstone attended Makaya during the latter's final illness, by which time Livingstone had apparently both forgiven Makaya for his past atrocities and become genuinely fond of him: ‘Muini Makaya the path finder to Manyema died to-day, and Livingstone showed genuine + unaffected sorrow’ (Stanley 1871–1872, 6 March 1872). The author is grateful to Mathilde Leduc-Grimaldi for her assistance with correcting a transcription of the relevant parts of Stanley's field notebook (1876–1877a). All subsequent references to this text refer to this transcription. For more on Victorian practices in the production of illustrations of African people and scenes such as those appearing in Figures 1, 3, and 5 of this article, see Koivunen (2009).

6 The Arab settlement of Kasongo, just upstream from Nyangwe, might also be added to this list, although Kasongo grew to prominence only after the period (the early 1870s) that is the focus of this article. Sheriff (1987, p. 191) provides an excellent map of the principal nineteenth-century East African trading routes.

7 There are a number of alternate spellings for this region. This article will invariably use ‘Manyema’, unless quoting a source, to avoid confusion.

8 Early initiatives west of Lake Tanganyika (in the 1830s) led by the Arabs and by the Nyamwezi in the form of Msiri's ‘Yeke’ focused on the regions of Kazembe and Katanga, both of which lay to the south-west. However, once ivory supplies in those regions began to dwindle, Arab interest shifted further north to Manyema (Birmingham 1981, pp. 130–133, 138–139; Sheriff 1987, pp. 186–190).

9 Hinde (1895, 1897) provides the most detailed accounts of the fall of the Arab forces at Nyangwe. His accounts include a photograph of Nyangwe (1895, p. 433) against the background of the Lualaba River; this is one of the few known nineteenth-century photographs of the village (Dhanis 1895, p. 35 also includes an illustration of the Nyangwe market based on a photograph by Lt. Lemery, but the present author has been unable to track down the original). Nyangwe's significance as a trading center continued despite the defeat of the Arabs; the Belgians, for instance, incorporated the village into their scheme of colonial development (Glave 1897a, p. 916). Ankei (1985, p. 94) reports that the market at Nyangwe was abandoned only in 1962.

10 Livingstone's objectives for his final expedition to Africa (1866–1873) defy simple elucidation (Bridges 1973).

11 Moorehead's books (1960, 1962) remain the best accounts of the European dimension of this quest, although Jeal (2011) has recently filled in some key details. Three recent articles by the present author (2008a, 2008b, 2010) elaborate the non-Western contexts of Burton's, Speke's, and Baker's expeditions, although it is also important to cite the earlier efforts of Bridges (1970) to link Speke's travels to the lives of the African groups he encountered. Finally, Kennedy (n.d.) provides the best short introduction to the topic, particularly as he places the Nile quest within the context of British global initiatives, sketches both the Western and non-Western dimensions of the quest, and outlines the impact of British celebrity culture in the shaping individual explorer legacies.

12 Jeal (1973, p. 324) provides excellent comparative maps of the Central African watershed as it is and as Livingstone believed it to be.

13 Livingstone stayed in Nyangwe from 30 March 1871 to 20 July 1871 and was present for some 24 markets. Cameron's visit lasted from 4 to 24 August 1874, thereby allowing him to witness six markets. Stanley stayed but a few days, arriving on 27 October and departing on 5 November 1876. From the available evidence it is impossible to determine how many markets Stanley saw, but it could not have been more than three.

14 Livingstone ([1871a] 2011--2013, [1871b] 1964, [1871c] 1964, [1871d] 1964, [1871e] 1872, [1871f] 1872, 1872a, 1872b, 1872c, 1872d, 1874), Cameron (1874a, 1874b, 1875, 1875–1876a, 1875–1876b, 1876, 1877a, 1877b, 1877c), and Stanley ([1876a] 1970, [1876b] 1970, [1876c] 1970, [1876d] 1970, 1876–1877a, [1877a] 1970, [1877b] 1970, 1877–1878, 1878). To this might also be added the records left by later European visitors such as, for instance, Wissmann (1890, 1891); Trivier (1891); Hinde (1895, 1897); Glave (1897a, 1897b); and Foà (1900). Finally, the autobiographical references of Tippu Tip (Whiteley 1958–1959), the renowned Arab slave trader, to Nyangwe represent a significant non-Western source.

15 As will be discussed below, the name of this group – if the group can even be conceptualized as a unified and coherent entity – takes a variety of forms in the relevant literature. This article will invariably refer to the group as the ‘Wagenya’, except when quoting other sources, in order to avoid confusion.

16 The distinction between British geographical discourse and (subsequent) imperial representation is needed to avoid anachronism. It is important to remember that – despite the discursive trajectory that will be mapped out here – the activities of British explorers and others during the ‘pre-prelude’ (1856–1876) and ‘prelude’ (1876–1884) to the partition of East Africa were not inexorably leading to the establishment of imperial control. Rather, these activities often took shape and direction because of ‘continuing reluctance on the part of the British governments to be drawn into direct responsibilities on the mainland’ (Bridges 2000, p. 66). So although this article will establish key discursive connections between British geographical discourse and imperial representation, particularly around the term ‘darkness’, the explorers did not necessarily see themselves as working toward the specific kind of European imperialist initiatives and strategies of representation that eventually came into being.

17 Kennedy's biography of Burton also takes this approach:[…] many aspects of Burton's life and thought make a good deal more sense when set in the context of historical scholarship on the various lands and peoples that served as the backdrops for his career and the objects of his curiosity. (Kennedy 2005, p. 279; cf. 2007, n.d.).

Hopkins's (2008) and Youngs's (2009) reviews of Jeal's recent biography of Stanley (Jeal 2007), conversely, censure Jeal for failing to take account of such non-Western regional histories, a failing that Jeal has worked to address in his most recent book on the exploration of the Nile (Jeal 2011).

18 This approach, in fact, has a long critical tradition. In the African context, relevant scholarship goes back to research on the role of Africans in the exploration of East and Central Africa by Simpson (1976) and to a series of 1970s and 1980s articles on ‘Voyageurs Africains en Afrique équatoriale’ in Zaïre-Afrique by François Bontinck.

19 As a result, the article will not be able to give the detailed attention merited by the British and broader European contexts most relevant to the discussion here. For more on these contexts, particularly the development of Britain's ‘unofficial empire’ in East Africa, see Bridges (2000).

20 This list might also be expanded to included older studies by Ceulemans (1959), Page (1974a, 1974b), and Bennett (1986).

21 In a despatch written during the 1871 search for Livingstone, Stanley ([1871] 1970, p. 115) describes Manyema as ‘the El Dorado of the Arabs and Wamrima tribes’ due to the region's ‘wealth of ivory and reports about the fabulous quantities found there’.

22 Renault (1987, p. 148) divides this network into a hierarchy of towns forming principal bases such as Kasongo (20,000 residents in 1889) and Nyangwe (10,000 residents in 1889); centers with thousands of people (e.g. Mtowa, Riba-Riba, Kirundu, and Isangi); posts held by smaller garrisons; and, finally, ‘delegates of the big traders, established in subject villages whose chief had to hand over several hostages as a guarantee of good behaviour’.

23 This network might be compared to a similar network developed in West Central Africa by the Chokwe, who from the 1840s onward ‘established their own long distance caravan system, with their own routes, carriers and leaders’. Eventually this network linked the western part of the Luba Empire to the Atlantic Ocean in a manner similar to the way in which the East African network linked the eastern and northern parts of the Luba Empire with the Indian Ocean (Birmingham 1981, pp. 122–124).

24 Tippu Tip, of course, rose to the greatest prominence among the Arab traders, making Kasongo his principal residence in about 1875, hence the rise of that settlement thereafter (Bennett 1986, p. 115). Ultimately, rule of Nyangwe would be divided among two Arabs, each of whom had his own area of influence in the settlement (see page 214). When Mtagamoyo assumed sole leadership of Nyangwe in the late 1880s, he became second in importance only to Tippu Tip in the general region (Renault 1987, p. 213).

25 Wangwana as distinct from watumua (‘slaves’) and washenzi (‘savages’ or local Africans) (Northrup 1988, p. 24).

26 Livingstone (1872b, p. 19) argued that Indian financiers from Zanzibar and the East African coast enabled the network to operate:The British Banian subjects have long been and are now the chief perpetrators of the Zanzibar slave trade; their money, and often their muskets, gunpowder, balls, flints, beads, brass wire, and calico, are annually advanced to the Arabs, at enormous interest, for the murderous work of slaving, of the nature of which every Banian is fully aware.

As a result, Livingstone (1872b, p. 20) also gestured in this direction when accounting for ultimate responsibility for the Nyangwe massacre: ‘the chief perpetrator, Tagamoio, received all his guns and gunpowder from Ludha Damji, the richest Banian and chief slave-trader of Zanzibar’.

27 Page (1974a) develops a detailed analysis of the role of the wangwana in the slave trade of East Africa, particularly in Manyema.

28 German explorer Hermann Wissmann, who visited Nyangwe in 1882, wryly observed that,It is unmistakable that, wherever they come to, the Arabs undertake a certain cultural improvement which is, however, purely egoistic and carried out so ruthlessly, that, although accomplished vigorously to their own benefit, it doesn't become a blessing for the natives. (1890, p. 179; translated from German by Sara Duana Meyer at the author's behest; all subsequent references to this text rely on this translation)

29 Stanley (1876–1877a, entries after 25 October 1876; cf. [1876a] 1970, p. 321) describes this process in no uncertain terms:

The fine open pastoral country between Mana Mamba and Nyangwe teemed with populations rich in large flocks of goats ^\sheep + hogs +/ [word illegible], banana plantations, each ridge, + eminence was crowned with a neat Manyema village. When Moeni Dugumbi, like a free hyaena lurked on the south bank of the Luama. Eight years later, this same Moeni Dugumbi + his Mtagamoyo are the recognized chiefs of populous Nyangwe, but behind them \lies/ the ^\once/ populated plain is ^\converted to a/ wilderness.

The traders, moreover, practiced such methods widely in the region. ‘In Manyema above all […] whole villages disappeared’, notes Renault (1989, pp. 151–152), a circumstance that became self-perpetuating because it ‘compelled the traders to advance still further to discover new [human and material resources]’ and so reignited the cycle of violence.

30 This was certainly the less desirable location. Cameron (1877a, vol. 2, pp. 1, 13), who visited the area on his departure from Nyangwe, describes it as a ‘about as pestilential a place as it is possible to imagine […] stagnant, muddy backwater, reeking under the sun's rays’, and adds that ‘The place was inhabited only in the dry season by the fever-proof Wagenya, owing to its being flooded for four or five months of the year’. Ironically, it was from this same location and with the assistance of the Wagenya, that the Belgians launched their successful assault on the Arab forces at Nyangwe on 4 March 1893 (Hinde 1895, p. 434).

31 One important aspect of the massacre that critics rarely, if ever, note is that Zanzibari violence on the opposite bank of the Lualaba River both preceded and succeeded the slaughter in the Nyangwe marketplace: e.g. ‘16th [July 1871 …] 12 AM Dugumbe's people shooting people on other side Lualaba = set fire to a village on bank = many captives caught on other side river’ (Livingstone [1871a] 2011–2013, p. 297b/148, cf. p. 297b/146). Stanley (1876–1877a, entries after 25 October 1876) indicates that such forays continued long after the massacre and represented a typical practice of the traders settled at Nyangwe: ‘26th [Oct. 1876] Another Arab Expedition left night of the 25th and next morning attacked the Wagenya on the left bank of the Lualaba, with a most bloody result, + a large booty of slaves’.

32 Livingstone ([1871a] 2011–2013, p. 297c/105) made a similar observation even prior to the massacre: ‘The Manyema have learned to distrust all strangers and think to buy means plunder and murder’.

33 However, Ceulemans (1959, p. 47) differentiates between areas where the Arabs had settled and established their leadership, and areas into which these Arabs launched expeditions in search of goods. Stanley's history of Nyangwe (1876–1877a, entries after 25 October 1876) suggests that the village and its surroundings were at the point of transitioning from the latter to the former when Livingstone arrived in 1871.

34 Dugumbi, a half-caste who hailed from Winde on the East African coast, was the first Arab to settle at Nyangwe (Stanley 1876–1877a, entries after 25 October 1876; Bontinck 1974, p. 75; cf. Wissmann 1890, p. 178, who states that Abed bin Salim first arrived in Nyangwe followed by Dugumbi). Stanley (1878, vol. 2, pp. 118, 120) characterizes Dugumbi as ‘a vain, frivolous old fellow, ignorant of everything but the art of collecting ivory’. Cameron (1877a, vol. 2, p. 2) describes Dugumbi's mental deterioration and suggests that the latter's energies in Nyangwe focused principally on taking drink and drugs and maintaining a harem: ‘He had collected round him over three hundred slave women, and the ill effects of this arrangement and his indulgence in bhang and pombé were plainly noticeable in his rapid decline into idiocy’. By the time Wissmann (1890, p. 177) visited Nyangwe, Dugumbi had passed away and Mtagamoyo had become ‘the guardian and custodian of [Dugumbi's] sons’.

35 Abed was known for his generosity to Livingstone, Cameron, Stanley, and other European visitors to Nyangwe. His full name was Abed bin Salim il-Khaduri (Bontinck 1974, p. 240). Wissmann (1890, p. 177) pronounced him to be ‘a slim, medium-sized, beautiful man of about 70 years’ in 1882. Abed passed away at Ujiji in March 1886, while en route to Zanzibar where he had been summoned by the Sultan to pay a long-standing debt to Indian financiers. Abed's generosity and kindness only extended so far. As Wissmann and his traveling companion Pogge prepared to depart Nyangwe, Abed had a group Wagenya beaten when they failed to provide a canoe for Pogge. Abed justified this behavior by stating that ‘the Wagenya were bold and deceitful and kept in obedience only by fear’ (Wissmann 1890, p. 188). To Stanley (1876–1877a, entries after 25 October 1876), Abed further elaborated this philosophy, noting that the attacks on the Wagenya were frequent[…] because we cannot teach these Pagans to be quiet, they are always misbehaving, killing some of our people whenever they get a chance. A small force of five or ten guns dare not set out to hunt game or shoot elephants but that they will be attacked + perhaps cut to pieces if God's hand is not interposed to save them. This state of things has long ago hardened our minds, so that upon the report of the murder of one of our people, we all join hands + set out to pay them.

Bontinck (1974, pp. 240–241) enumerates the principal sources of information on Abed.

36 Livingstone ([1871a] 2011–2013, p. 297b/149) identifies Mtagamoyo as the principal instigator of the Nyangwe massacre, while Stanley (1878, vol. 2, p. 119) calls him ‘the butcher of women and the fusillader of children’. He was born in Bagamoyo around 1820 (Bontinck 1974, p. 241), and his full name was Mwini Mtagamoyo bin Sultani Wakasine (Anon 1984, p. 18). When Stanley (1878, vol. 2, p. 120) met him, Mtagamoyo was

[…] about forty-four years of age, of middle stature and swarthy complexion, with a broad face, black beard just greying, and thin-lipped. He spoke but little, and that little courteously. He did not appear very formidable, but he might be deadly nevertheless. The Arabs of Nyangwé regard him as their best fighter.

His riches came to include 10,000 slaves and 5000 guns (Anon 1894, p. 18); his territories eventually stretched ‘north along the Lualaba as far as Riba-Riba and, over the midstream of the Lomani, from Faki to Musumba-Chadi’ (Renault 1989, p. 149). Mtagamoyo died in battle with the forces of the Congo Free State on 8 January 1893 (Hinde 1897, pp. 41–50). Bontinck (1974, p. 241) enumerates the principal sources of information on Mtagamoyo.

37 Cameron (1874b, 9 August 1874) marveled at the life these men led:

People here have their own work which they do in a lazy sort of way just drifting through it, time is nothing to them, they have hordes of slaves + women + food as much as they want + I believe many of them will never have the energy to leave here but stop here till they die.

38 Bennett (1986, pp. 112–118) provides a general overview of Arab initiatives during this period, particularly Tippu Tip's activities, in the regions west of Lake Tanganyika.

39 In other words, Nyangwe was situated at the northern edge of the savannah lands that stretched to the south, and the village was just outside of (or at the edge of or just inside of) the Central African rainforest (cf. Raucq 1952, p. 25).

40 Tippu Tip (Bontinck 1974, p. 75) adds further nuance to this settlement pattern, noting that of the two parts of Nyangwe, ‘one part was for Arabs […] and another for those from the Mrima [East African] coast’, a division that maps onto, respectively, Abed's and Dugumbi's quarters. Wissmann (1890, pp. 178–179) notes that,

The houses are mostly built of adobe, covered with grass and have a verandah. They are surrounded by gardens and banana-brush and cluster in random alignment, but always those of the dependants and slaves around the houses of their masters.

By 1882, Nyangwe had expanded to a third part, six kilometers to the south, where Juma bin Salim lived. Hinde (1895, p. 435) writes that Nyangwe was reduced to a single house after being sacked by the Belgian forces.

41 These two islands (along with two smaller ones) remain visible in the modern satellite representation of Nyangwe available from Google Maps (Figure 2; [Online] Available at: https://maps.google.com/?ll=\-4.279482,26.250887&spn=\0.069586,0.077162&t=\h&z=\14). This representation, incidentally, gives the approximate coordinates of Nyangwe as south latitude 4.28 and longitude 26.25 (cf. Stanley, 1878, vol. 2, p. 117, which gives an estimate of south latitude 4.25 and longitude of 26.67).

42 In a letter to the RGS, Cameron (1876) notes that,

Here the bed of the river has a very rapid fall, + its current is very fast, from 3 to 4 knots opposite Nyangwe – I measured the width of the river at this point with a sextant + found it to be 1,020 yards, in many places it is much wider.

43 Livingstone and other explorers alternately use the Swahili word for market (soko) or the Genya word (chitoka). Both words (including variants of the latter) remain in use on this stretch of the Lualaba, as Ankei (1985, pp. 94–95, cf. p. 92) demonstrates in detail.

44 In his comprehensive study of precolonial Africa's exchange economy, Sundström (1974, p. 66) defines ‘barter’ as ‘handing over one commodity directly for another, either according to established rates or by on-the-spot bargaining’. Such periodic barter markets were not unique to Nyangwe. Livingstone (1872c, p. 12A) states that they were held ‘about 10 to 15 miles apart’ in this region. Moreover, such markets have persisted in the area to more or less the present day. In the 1970s, for instance, anthropologist Yuji Ankei conducted an extensive survey of 42 such markets on the Lualaba, all in the vicinity of Nyangwe. In these markets, Ankei (1985, p. 90) observes,

Barter rates do not fluctuate according to the supply-and-demand market principle. […] Imbalances of supply and demand are settled with a combination of gift-giving, one week lending, or deferred barter, of excess foods for other kinds, and finally by the use of cash.

Any of these practices, except for the last, would have been viable in the nineteenth century. In fact, Ankei (1985, p. 91) argues for the general survival of the precolonial barter economy in this region, while his work offers the most detailed overview available of such markets on the upper Lualaba (see, especially, 1984, pp. 9–15).

45 In his field notebook, Stanley (1876–1877a, 19 Nov. 1876) writes as follows: ‘Wagenya • or Waenya • fond of markets • they hold them alternate days on each bank’. Livingstone ([1871f] 1872, p. 13) suggests that the massacre occurred on a day when the market was held at the third location: ‘[The market] was held on a long slope of land which, down at the river, ended in a creek capable of containing between fifty and sixty large canoes’. Cameron (1874b, 8 August 1874) observes that such markets had a short, but intense duration: ‘The soko lasted 3 or 4 hours + the noise + smell were overpowering’.

46 ‘The island people, and the back country people meet on the neutral ground that is the river bank + establish market places where they exchange their produce – palm wine, palm oil, plantains, Indian corn & Peanuts’ (Stanley 1876–1877a, notes after 3 December 1876). Livingstone ([1871a] 2011–2013, p. 297b/141) offers additional detail: ‘The Bagenya are fishermen by taste and sell the produce of their nets & weirs to the other tribes who cultivate the soil at the different markets’. The barter economy would have facilitated interchange between these two groups of people (Ankei 1985, p. 91). During this period, the occasional contact made between nomadic hunters and farmers in the rainforest itself, by contrast, did not lead to the development of markets in the rainforest (‘Legaland’), a circumstance that continued into the twentieth century (Biebuyck 1973, p. 32; Birmingham 1981, p. 140). Livingstone offers his most detailed descriptions of the market in the 24 May entry of his field diary ([1871a] 2011–2013, p. 297b/133) and in the corresponding entry of his ‘Unyanyembe Journal’ (1872a).

47 Stanley's description (1878, vol. 2, p. 121) makes this most clear:

From one thousand to three thousand natives of both sexes and of all ages gather here from across the Lualaba and from the Kunda banks, from the islands up the river and from the villages of the Mitamba or forest.

It is also important to underscore, first, the social dimension that such a market would have fulfilled – ‘The market is a place to meet friends, to hear news, to contract alliances, to initiate sexual adventures’ (Herskovits 1962, p. x) – and, second, the key role that women would have played in enabling and maintaining the majority of core market activities (Herskovits 1962, pp. xi–xii; for more on the social dimension, see Bohannan & Dalton 1962, pp. 15–19; for practices along the Lualaba in particular, see Ankei 1984, pp. 55–56 and 1985, p. 90). Livingstone ([1871a] 2011–2013, 1872a), for instance, rarely mentions male market goers and represents the Nyangwe massacre foremost as an attack on the female participants. Livingstone (1872a, 1 April 1871) also reveals the centrality of the market to social interaction: ‘As a rule all prefer to buy and sell in the market to doing business anywhere else. If someone says come sell me that fowl or cloth – the reply is come to “chitoka” or market place’.

48 The principal currencies in the market were cowries, goats, and slaves (Cameron 1877a, vol. 2, p. 3). These currencies had the following relative values: ‘each slave = 3 to 5 goats[,] each goat = 30 to 50 cowries’ (Cameron 1874a, 10 August 1874).

49 Raucq (1952, n.pag.) includes a map of the ethnic groups settled around Nyangwe. This map provides some very general sense of the different populations that would have met at Nyangwe's nineteenth-century market, although a variety of methodological and historical considerations (see, for instance, Turner 2007) make it unwise to rely too much on such twentieth-century colonial population maps.

50 For a detailed look at the use of fish in markets on the Lualaba, see Ankei (1984, pp. 58–60).

51 ‘The Arabs, wherever they settle throughout Africa, endeavour to introduce the seeds of the vegetables and fruit-trees which grow in their beloved island of Zanzibar’ (Stanley 1878, vol. 2, p. 123). For more on the local African agricultural and fishing practices used to produce goods for such markets, see Ankei (1984, pp. 4–8; 1985, pp. 89–90). Vansina (1965, pp. 106–108, 164–165) describes the overall economy of Manyema in this period. Finally, Livingstone's observation ([1871f] 1872, p. 13) regarding the exchange of items from the two sides of the Lualaba is also of interest here: ‘The people west of the river brought fish, salt, pepper, oil, grass-cloth, iron, fowls, goats, sheep, pigs, in great numbers to exchange with those east of the river for cassava, grain, potatoes, and other farinaceous products’. That said, based on the available evidence it is impossible to determine whether the market pre-dated the arrival of the Arabs at Nyangwe, although the market's complexity and geographical reach suggest a much longer history.

52 Cf. Roberts (1973, pp. 60–63) and MacLaren (1992, pp. 41–43). Koivunen (2009) elaborates this point in the context of the illustrations used for Victorian travel narratives about Africa.

53 Multi-authored in that many individuals contributed to the final published form of the text, although the actual by-line might only contain the explorer's name.

54 Casada (1975, p. 195) long ago argued that such scientific papers ‘were designed for select audiences of specialists’ and so ‘tend to be more reliable than the sometimes sensationalized accounts designed for the general reading public’.

55 The question of which of these is the closest to ‘objective reality’ and whether such an approach can even be made remains open to debate. See the discussion between Youngs (1994), Bridges (1998), and Wisnicki (2010).

56 Helly's (1987) work on the creation of the posthumously published Last Journals (Livingstone 1874) is exemplary in substantiating this point.

57 The Livingstone Spectral Imaging Project restored the full text of this faded and ragged diary in 2011 (Livingstone [1871a] 2011–2013). Livingstone ([1871a] 2011–2013, p. 297c/107) called the plant used to make the ink ‘Zingifure’, while Cameron (1874a, 10 August 1874) provides the most detailed contemporary description: ‘A tree here bears prickly husks full of little red seeds which give a beautiful carmine colour[,] they mark at once. I am told Livingstone made red ink of them when he was here’. Jo Anthony of the National Trust for Scotland has recently collaborated with staff at Kew's Centre for Economic Botany to identify this plant decisively as Bixa orellana (Mzingifuri in Swahili; [Online] Available at: http://www.prota4u.org/protav8.asp?h=\M4&t=\Bixa,orellana&p=\Bixa+orellana#Synonyms).

58 Bridges (1987, p. 180) identifies three types of primary explorer records:

There is the first-stage or “raw” record made as [the explorer] went along, the more considered and organised journal or perhaps letter written during intervals of greater leisure and finally the definitive account of the expedition, usually composed after his return to Europe with a view to publication.

However, Bridges (1987, p. 184) adds that access to the first-stage records does not necessarily eliminate the need for records of the latter stages, which often provide the only way to make sense of the first-stage records. The ideal then is to use all the records and to apply critical judgment in evaluating the evidence, as has been attempted in the reconstructed history of Nyangwe above.

59 Livingstone ([1871a] 2011–2013, p. 297b/146ff.; [1871f] 1872, pp. 13–14; 1872a, 15 July 1871ff.; 1872b, p. 20; 1872c, p. 12A; 1874, vol. 2, pp. 132ff.) and Stanley (1872, pp. 382–383; [1871] 1970, p. 117). There may be additional narratives of the massacre by Stanley and Livingstone elsewhere that escaped the present author's notice.

60 As critics have noted of late, these traders also paved the way for European explorers and missionaries to visit the region, while trading network practices significantly shaped the experiences of these same Europeans (Northrup 1988, p. 22; Vansina 1990, pp. 240–242; Kennedy 2007; Hopkins 2008). In some cases, the Europeans then opened new routes for the Zanzibari traders, as Stanley did for Tippu Tip (Stanley 1878; cf. Wissmann 1890, p. 191; Kennedy n.d.).

61 The map in Vansina (1990, p. 241) places Nyangwe just inside of the central southern perimeter of the Central African rainforest. The map opposite the frontispiece in Northrup (1988), conversely, places Nyangwe directly outside the southern edge of the rainforest. Likewise, Ankei (1985, pp. 90–91) writes, ‘The vegetation type at Kasongo is savanna, and it tends to be in transition to tropical rain forest just north of the village of Nyangwe’. Given these points and those noted above regarding Nyangwe's proximity to the rainforest, the village might best be said to lie in what Raucq (1952, p. 6) calls a ‘transition zone’ between savannah and rainforest.

62 British explorers refer to the Lega alternately as the ‘Rega’, ‘Waregga’, and ‘Balegga’, and call the region ‘Uregga’.

63 Vansina (1990, p. 43) also observes that, ‘Contrary to myth […] no population in the area was ever isolated, “lost in the woods”. In fact, ease of communication, especially within the Zaire basin, is a major geographical characteristic of the area’.

64 Biebuyck (1973, p. 2) notes that the Lega's ‘sense of historical and cultural unity is very strong’ and ‘is rooted in a common language, common historical experiences, common genealogical charts, in territoriality, and in a common set of basic institutions and values’. However, Biebuyck (1973, pp. 5–7) also describes the intermixing that occurs between the Lega community and those groups settled at its margins such as the Songola, Zimba, Bembe, Komo, and Nyindu. There are in effect ‘cultural buffer groups’, he argues, ‘which are either segments of Lega clans that have been subject to strong linguistic and cultural influences from neighbors, or segments of other tribes that have been acculturated by Lega groups’. Among such ‘Lega-ized’ groups, Biebuyck (1973, p. 17) includes ‘the so-called Genya riverains found here and there between Kisangani and Kindu’, which the Lega consider to be ‘a different, yet friendly, people’.

65 For the geographical extent of the Luba Empire, see the maps in Wilson (1972, p. 578, cf. the verbal description on p. 575), Reefe (1981, p. 116), and Roberts and Roberts (1996a, p. 25). Wilson, incidentally, places Nyangwe just north of the central northern perimeter of the Empire, while Reefe, conversely, places Nyangwe within this same northern perimeter. British explorers referred to Lubaland as ‘Rua’ or ‘Urua’, and called the people ‘Barua’ (but see below in note 66). Cameron (1877b, p. 172) provides a Victorian-era estimation of the extent of the Empire.

66 Livingstone and Stanley do reference the Empire in their writings of the early 1870s. Moreover, Livingstone was aware of the Luba by the mid-1850s at least, a point that has escaped critical notice until now. In particular Livingstone's hand-drawn map (1857) of his 1852–56 trans-African journey places ‘the Luba’ at the top left-center of the map and describes them as ‘a large tribe, very averse to civilization’. It is also worth noting that in this instance Livingstone uses the word ‘Luba’ (as is common today), whereas in the 1870s, Livingstone himself, Cameron, and Stanley all used the word ‘Rua’ (or ‘Warua’) when referring to this group.

67 Three successive Luba rulers – Illunga Sunga (c.1780–c.1810), Kumwimba (c.1810–c.1840), and Illunga Kalala (c.1840–c.1870) – held primary responsibility for the expansion of the Empire during this period. The second of these extended the Empire to ‘a new, north-eastern province in Manyema’, thereby bringing the Luba to the doorstep of Nyangwe just before the period under discussion (Birmingham 1981, p. 137; Reefe 1981, pp. 150–151 describes Luba settlement in the vicinity of Nyangwe). The arrival of well-armed Zanzibar traders to the general area, in turn, contributed to the decline of the Luba Empire (Birmingham 1981, pp. 137–139; cf. Wilson 1972, p. 588; Reefe 1981, p. 159ff.). Nyangwe and Kasongo, in fact, served as staging points for raids on villages along the northern frontier of the Luba Empire (Reefe 1981, p. 168). As a result, the Nyangwe massacre may have been linked to this broader conflict.

68 Vansina (1968, pp. 155–156) argues that the power structure of the Luba Empire, like those of the Lunda kingdom and the kingdom of Kazembe, ‘had a nucleus which was tightly controlled by the central government’, while in the ‘outlying provinces […] the authority and power of the central government faded away more and more the farther one went from the center toward the boundaries’. Roberts and Roberts (1996a, p. 20; 1996b, p. 152) add nuance to this argument by noting that ‘there is no credible evidence that Luba-ized peoples ever constituted an “empire” ruled by a supreme authority’. Rather, ‘the Luba state involved a far more flexible set of relationships that extended in a wide circle of influence rather than authority’. As a result, the Luba Empire, they conclude, ‘actually was and remains a complex ethnic mosaic, full of distinct groupings by lineage, clan, politics, and geography’. For a discussion of the meaning of ‘Luba’, see Roberts and Petit (1996, pp. 211, 242n.).

69 The overall importance of the Wagenya on the river facilitated this role: ‘The Wagenya possess the whole carrying trade of the river, carrying people to different markets, which take place in different localities nearly every day’ (Cameron 1877b, p. 171). Stanley (1878, vol. 2, p. 147; cf. [1876d] 1970, p. 339) notes that the word ‘Lualaba’ itself appears to be a corruption of the Wagenya name for the river ‘Lu-al-ow-wa’, a point that further underscores the close relationship between the Wagenya and the river.

70 Of particular relevance here is Biebuyck's observation (1973, p. 21) that classifications of tribes ‘are far from reflecting the exact cultural situation in and around Legaland because they treat these various “tribes” as homogenous groups and fail to take into account the many overlappings that exist’.

71 This included the taste and after-effects of eating human flesh, which Livingstone (1872c, p. 12A) duteously reported: ‘They say that human flesh is not equal to that of goats or pigs; it is saltish, and makes them dream of the dead’.

72 The Arabs at Nyangwe made a rather disastrous venture over these rapids while Livingstone was staying in the village, with five people killed when their canoe overturned. Livingstone ([1871a] 2011–2013, pp. 297b/139–140) believed that he had been saved from being part of this party only through the intervention of Providence.

73 Cameron, who like Livingstone failed to sail down the Lualaba, made a concerted effort to determine its identity and, using collected oral information, came to the conclusion that Livingstone had been wrong:

I have seen every one Livingstone saw when he was here + cannot make out how he supposed the river to go to the Nile. I made a map today of all that I had heard + then took Tanganyika over his journeys on it + corrected it by him. If Piaggias lake has any existence, which I doubt, its waters from the general direction of the rivers hereabouts would most likely come to the Congo. (Cameron 1874b, 20 August 1874; cf. Stanley [1876b] 1970, pp. 331–332)

74 The manuscripts and published texts of Livingstone, Cameron, and Stanley are full of calculations made at Nyangwe that relate to these questions. For instance, the main part of Stanley's principal dated entry from Nyangwe in his field notebook (1876–1877a, 2 November 1876) is largely taken up with measuring and comparing the levels of the Lualaba River at Nyangwe and Lake Tanganyika at Ujiji. (I am grateful to Roy Bridges for providing me with a detailed explanation of the calculations in this entry and to Sarah Strong for additional elucidation.)

75 Cf. Hinde (1895, p. 442), who notes that, ‘It used to be a common saying, in this part of Africa [the Upper Congo basin], that all roads lead to Nyangwe’.

76 Another despatch from Stanley ([1877a] 1970, p. 360), written just after the great river journey, makes this connection even more explicit:

The eastern half of Africa is generally known, but the western half was altogether unknown. To anyone arriving from the East Coast with the love of exploring unknown wilds, what a field lay extended before him! The largest half of Africa one wide enormous blank – a region of fable and mystery – a continent of dwarfs and cannibals and gorillas, through which the great river flowed on its unfulfilled mission to the Atlantic! Darkness and clouds of ignorance respecting its course everywhere! (emphasis added)

That said, this example and the one above show that at this early moment Stanley still uses words such as ‘darkness’ and ‘gloom’ interchangeably and that in his use here ‘darkness’ does not yet carry the broader ideological connotations – especially as linked to justifying imperialism – that it would assume later in the century.

77 A search for ‘dark continent’ (no quotation marks – as per Google instructions) in the ‘English corpus’ between 1750 and 2008 in Google's Ngram Viewer (which graphs word and phrase frequency in all the books digitized by Google) reveals that the use of this phrase dramatically escalated between the middle of the 1870s and the beginning of the 1890s ([Online] Available: http://books.google.com/ngrams/). Moreover, Ngram Viewer indicates that this surge began in 1874, the year which witnessed the publication of Livingstone's posthumously edited Last Journals (1874). A search for ‘dark continent’ (in quotation marks) in the ‘Entire Document’ search tab in Gale's NewsVault (which aggregates periodical sources) produces similar results ([Online] Available: (http://find.galegroup.com/dvnw/dispBasicSearch.do?prodId=\DVNW&userGroupName=\birkb). Prior to 1849, the NewsVault returns no hits (although the digitized documents go back to the seventeenth century), then for the years between 1849 and 1877 the NewsVault returns 50 hits. However, beginning in 1878 – the year in which Stanley published Through the Dark Continent (1878) – the numbers start to climb: 196 hits in 1878, 152 hits in 1879, then 1744 hits total in the 1880s and 3948 hits in the 1890s. Broad analysis of these trends lies outside the scope of the present article, but the correlation between, on the one hand, the Google and Gale data and, on the other, the argument being advanced here does suggest that the writings of Livingstone and Stanley (and to a lesser extent Cameron) must have exerted some degree of influence in spurring general use of the phrase ‘dark continent’, especially as these explorers had close and well publicized links to Africa at this historical moment. If so, then understanding the specific experiences of these explorers in Nyangwe – for all the reasons discussed in this article – becomes all the more important. Note: As available to the present author through Birkbeck College, Gale NewsVault aggregates the following six databases: 17th–18th Century Burney Collection Newspapers; 19th Century British Library Newspapers: Parts 1 and 2; 19th Century UK Periodicals: New Readerships; Illustrated London News Historical Archive; and Times Digital Archive. Scholars at institutions with subscriptions to additional Gale databases (for instance, 19th Century UK Periodicals: Empire or Sunday Times Digital Archive) have the potential for an even broader search.

78 Coupland (1945, pp. 77–99) develops an extensive analysis of these factors. Although the issue of Livingstone's deteriorating health might also be added here, he actually remained fairly healthy while in Nyangwe – as was not the case for much of his last expedition.

79 Stanley (1878, vol. 2, pp. 97–98) was the first to make the link between Livingstone and Cameron. Stanley summarized these impediments as follows ‘“want of canoes and hostility of the savages”, reluctance of the Arabs to permit him to proceed by river from an officious regard for his safety, and the “cowardice of his followers”’. Cameron's journal (1874b) provides the best insight into all these causes, as in his published narrative (1877a) he significantly compressed and revised some of the key issues, particularly his extended attempts to secure a canoe.

80 The map in Ankei (1985, p. 93) places these rapids in the vicinity of the settlement of Kibombo as does Google Maps ([Online] Available: https://maps.google.com/?ll=\-4.279482,26.250887&spn=\0.069586,0.077162&t=\h&z=\14). For more on the overall hydrography of the Lualaba River in this region, see Raucq (1952, p. 21).

81 Google Maps gives the distance between Nyangwe and Kibombo (where the rapids are located) as about 34 miles in a direct line. As a result, by traveling 41 miles north from Nyangwe through the rainforest before returning to the river, Stanley clearly sought to avoid these rapids, although he does not explicitly state this (see 1878, vol. 2, p. 147).

82 Cameron's journal (1874b) details his attempts to collect and organize information about these rapids.

83 Elsewhere Cameron (1874a, 12 August 1874) further underscores the close relationship between the Wagenya and their canoes: ‘[the Wagenya] say their canoes are their legs, their wives & their children, that they use them every day to go to sokos [markets], to fish +c, + that they could not live without them’.

84 For Livingstone's unabridged description of his attempts to secure a canoe from the Wagenya, see his 1871 Field Diary ([1871a] 2011–2013) and the present author's discussion of this point in the section of the critical edition of the diary called ‘Livingstone in 1871’. As noted above, Cameron's journal (1874b) includes the information on his negotiations to secure a canoe revised out of the published version of the narrative (1877a). Notably, when Stanley finally sailed down the Lualaba, he did so thanks mainly to the fact that he, Tippu Tip, and their followers were able to steal some 36 canoes from the riverine populations downriver from Nyangwe (Stanley 1878, vol. 2, p. 185). This combination of strategies, incidentally, enabled Stanley to take the Wagenya at Nyangwe out of the equation, a move that both Livingstone and Cameron failed to make.

85 The true extent of this knowledge, of course, can never be known nor can scholars know whether any African individuals, groups, or sets of groups in this region regularly integrated their geographical information to produce the kind of collective knowledge that would have most interested European explorers. The point is that the African populations, because of what they did know, had the potential to produce such collective knowledge. For discussion of a similar issue in a different African context, see Feierman (2005).

86 For details of this alliance and its consequences, see Livingstone ([1871a] 2011–2013, p. 297b/141).

87 In other words, Livingstone uses the ‘ba-’ prefix to indicate the plural, which reflects his familiarity with various Bantu languages in southern Africa. In Swahili, this particular word takes the ‘wa-’ prefix in the plural (see below in note 89).

88 Additionally, Stanley's relevant field notebook entry (1876–1877a, 19 November 1876) does not quite correspond to the published version in terms of tribal name: ‘Wagenya • or Waenya’.

89 It's also worth noting that wagenya, a Swahili word that means ‘strangers’, suggests that the apparent coherence of the tribe may have been invented by the first Arabs to arrive at Nyangwe.

90 For instance, Maes and Boone (1935); Moeller (1936); Bulck (1948); and Boone (1961).

91 The first of these two groups, Maes and Boone (1935, p. 331) argue, lives on both sides of the Lualaba River between Stanleyville (in the north) and the Lilu River (in the south); the second lives wholly on the right bank between the Lowa River (in the north) and the Lulindi River (in the south). Nyangwe, incidentally, lies on this second stretch of river. Like Moeller (1936), Maes and Boone (1935, pp. 328, 331) adopt the name ‘Wagenia’, but also offer a string of variants: ‘Baenya, Ouénya, Vouaghenia, Vouaghenya, Vouenya, Vuagenia, Waenya, Wagenia, Wagenja, Wagenya, Waggenia, Wainya, Wenja’.

92 Moeller (1936, p. 173, cf. p. 186) distinguishes between the Wagenya of Kindu and the Wagenya of Kasongo and indicates that the first group derives from the Bagengele tribe with elements of the Wazimba tribe, while the second is of Luba origin.

93 Moeller (1936, p. 170, cf. p. 186) places the Wagenya of Nyangwe, of whom he provides a short history, into the first of these subgroups, which he locates as living on the banks of the river north of Nyangwe.

94 Bulck (1948, p. 502) links these three groups to, respectively, Stanleyville, Kindu, and Kasongo. Bulck, incidentally, uses the name ‘WaGenya’ throughout, but one of the linguistic charts at the back of his book (1948, p. 715) also records that the Genya of Stanleyville and Kindu are called ‘WaGenya’ in the plural, while those of Kasongo are ‘BaGenya’ in the plural. Finally, both Bulck (1948, p. 502) and Boone (1961, p. 41) conclude by noting that proper documentation for the group is minimal or non-existent.

95 In an email to the author (30 May 2011), Allen Roberts describes the situation in this way:

[…] generally social identity in this part of central Africa […] is situationally determined. That is, people have many different loyalties and ways of feeling and asserting commonality, and in some circumstance it may be expedient to self-identify as X, whereas the next day that would be the first thing I would deny because I am so obviously Y! However, a full consideration of whether Wagenya ontology (in the nineteenth, twentieth, and/or present centuries) forms a unique case or, instead, provides a representative example of a broader, transhistorical social condition in this part of Africa lies outside the scope of the current study. Of course, there is no objective reason why the Wagenya or any other such group need by defined by a western classificatory system in the first place. Consequently, the analytical emphasis here falls on examining what happens when such an attempt is made.

96 Excepting a comment on the first page of his narrative (‘[Livingstone] had died by the shores of Lake Bemba, on the threshold of the dark region he had wished to explore’, Stanley 1878, vol. 1, p. 1), this is the first time in the narrative that Stanley uses the term ‘dark’ (or any variant thereof such as ‘darkness’) to characterize the area that roughly emcompasses western half of the Congo Basin. Otherwise, prior to this moment, Stanley limits his use of the term to general characterizations (‘dark interior’, 1878, vol. 1, p. 2; ‘dark continent’, passim) or highly concrete descriptions (‘dark green foliage’, 1878, vol. 1, p. 29; ‘dark lakes of Karagwé’ 1878, vol. 2, p. 18).