Introduction
When Ahmed Nenshi, a Muslim merchant from Kutch (Gujarat), died in Bujumbura on 14 February 1930, his assets and liabilities had to be sorted out. Amidst a raging economic crisis, the liquidation of his legacy, amounting to about half a million francs, took more than a year and caused problems for the entire commercial sector in and around Bujumbura. All major commercial firms, whether Gujarati, Omani, Goanese, or Belgian, were involved either as creditor or as debtor. Despite a positive balance at the time of his death, bankruptcies of debtors and auctions generating less revenue than expected led to a slightly negative balance when his inheritance was finally closed. His father in Kutch, his African wife and their children in Bujumbura, his business partner, and the needy, whom he had designated by will as the beneficiaries of his inheritance, stayed behind empty-handed.Footnote 1
Based on bankruptcy and inheritance files like Nenshi's and on population and business registers, I reconstruct the commercial life of Arab and Indian traders operating from Bujumbura (then in Ruanda-Urundi) in the 1920s and 1930s.Footnote 2 The primary focus is on local commercial operations and on how these relate to long-distance networks, whereby the central position of first-comers, the role of the colonial state, cross-community interactions in Bujumbura, and the shifting significance of kin and community stand out.
Although this article decidedly plugs into the rich historiography about Indian and Arab communities in East Africa, my approach is also partly at odds with the dominant trends in the literature. Authors who made compelling cases for entanglements around the western Indian Ocean more often than not focused on coastal towns.Footnote 3 Those who did go further inland mostly limited their attention to communities from one area of origin in territories that once were part of the British Empire.Footnote 4 The few publications going beyond the bounds of the British Empire are mainly concerned with the precolonial period.Footnote 5 On top of that is a tendency to highlight the effectiveness of long-distance kin and community networks, thereby presupposing or at least privileging success.Footnote 6 There are good reasons for each of these traits in the literature. However, going beyond the coast, beyond the former British Empire, beyond the nineteenth century, beyond one diaspora community, beyond explaining presupposed success and focusing on a relatively small merchant community—in other words, going to the fringes of what the literature has done so far—allows us to reassess some of the received knowledge about Indian and Arab trade communities in East Africa—a region that cannot reasonably be reduced to the former British territories in East Africa.Footnote 7
To be clear, I do not want to contest the value and validity of the existing literature for the areas, times, and communities these writings deal with, but there are limitations to the reach of these findings. The authors coming closest to my case study are Colette Le Cour Grandmaison, who is the only one to have researched twentieth-century Omani traders in Bujumbura so far,Footnote 8 Gijsbert Oonk, who focused on the local operations of the Indian Lohana community in East Africa,Footnote 9 Thomas McDow, who traced Omani traders and their relations to Indian and African partners into the East Central African interior in the nineteenth century,Footnote 10 and Fahad Bishara, who scrutinized the role of contracts and credit in relations of trade and trust around the western Indian Ocean.Footnote 11 Le Cour Grandmaison has already shown that traditional patterns and hierarchies in the areas of origin were broken up in East Africa,Footnote 12 and Oonk pointed out that trading communities in East Africa have much closer (socioeconomic) ties there than they were accustomed to in their area of origin.Footnote 13 McDow came close to the area I am concerned with, but strictly speaking looked neither at Bujumbura nor at the colonial period.Footnote 14 Bishara's analysis includes the colonial period, but does not go beyond the coast. However, they both substantiated the importance of debt and credit relations, which is also the backbone of my analysis.
My analysis zooms in on Arab and Indian traders in the two decades around the Great Depression in a Belgian colonial context and based on archival sources derived from this context. The fact that I use bankruptcy and inheritance files in the colonial archives undoubtedly limits the reach of my analysis as well. In all cases leaving archival traces, there were European—usually Belgian—creditors involved. There is reason to assume that cases without European parties were dealt with differently and left no detailed traces in the colonial archives. Nonetheless, the retrieved sources are significant and relevant to address my research questions as they disclose a direct impact on all the leading Arab and Indian businesses in Bujumbura at the time and they give us snapshots of these merchants’ local and long-distance relations.
After this introduction, the paper consists of five parts. Firstly, I give a brief presentation of the Indian and Arab trading communities in Bujumbura during the interwar period, situating them in a regional East African context. Secondly, I sketch the Belgian colonial commercial policies directly affecting the Arab and Indian traders this article is concerned with. Thirdly, local relations inside Bujumbura are outlined based on inheritance and bankruptcy records, with particular attention to the period of the global economic depression in the early 1930s. In the fourth part, I address enduring yet modified long-distance influences and connections, with particular attention going to mobility and family relations. Finally, I conclude that the position of the Arab and Indian traders in Bujumbura was structurally vulnerable. Their kinship networks continued to play a vital role as mutual insurance at times of arrival and misfortune; however, having been relegated to an intermediary role in the long-distance trade, day-to-day business relations were increasingly a local, cross-community, and precarious affair.
Situating Indian and Arab Traders in Colonial Bujumbura
When Burundi came under Belgian rule as an outcome of the First World War, the main commodities for long-distance trade were hides and palm oil as export products, and salt and textiles as imports. The vast majority of the long-distance traders operating on Burundian soil were of Arab or Indian origin: of a total number of thirty internationally operating merchants in 1918 and fifty-four in 1920, between 60 and 75 percent were of Arab origin and between 20 and 30 percent of Indian origin. The residual fraction of approximately one-tenth was made up by a handful of Belgian and Greek companies.Footnote 15
At first sight, this is in line with what we know from the literature. In the late precolonial and the colonial period, merchants connecting small outlets in East Africa with long-distance trade more often than not came from the area around the Arabian Sea, mostly from the south of the Arabian Peninsula and the northwestern part of the Indian subcontinent. Their commercial predominance is often attributed to their effective networks, branched throughout East Africa and linked to Oman and Gujarat—and vice versa—relying on the long-time and long-distance maintenance of family or ethnic ties.Footnote 16 Religion—and particularly Islam—also formed a lubricant for trade, not the least because of the Islamic legal infrastructure buttressing contracts and credit relations.Footnote 17 Apart from the networks based on kinship, ethnicity, or religion, their business ethic—their willingness to work hard and long hours, to invest and to take risks, and to be content with minimal profit margins and an ascetic life—is also given as a reason for the “Asian” dominance in the East African commercial sector during the period under scrutiny.Footnote 18
These explanations have in common that they take the success story of Indian and Arab entrepreneurs in East Africa for granted. The reality, however, is less clear-cut. Not all attempts to start a business were met with success—far from it. Moreover, although the traders’ mobility was a prerequisite for commercial success, that same mobility meant that those who failed quickly disappeared from the scene—and, therefore, left scant traces in the archives. Oonk shows that around 1900 a large number of Indian traders (both Muslim and Hindu) went bankrupt within ten years of their arrival on the East African coast. Starting a business depended heavily on whether or not the (extended) family gave access to capital in the form of credit and information.Footnote 19
I observe something similar around 1920 in Burundi, far away from the Indian Ocean coast and beyond the bounds of British-ruled territories. Despite the clear preponderance of Arab and Indian merchants in the border-crossing trade, their presence proved fragile. Of the ten Indian trading companies that were active in Bujumbura between 1917 and 1920, not a single one appeared more than once in the annual lists. Permanence among Indian traders in the first postwar years was virtually nonexistent. Of the seventeen Arabs who had a shop in Bujumbura in the first years after the First World War, eight appear at least twice on the lists and five of them three years in a row. In Gitega, Rumonge, and Nyanza-Lac—after Bujumbura the secondary commercial centres in Ruanda-Urundi—the same pattern can be observed. Only from the 1920s onwards would Indian traders—some of them returnees—manage to remain active in Bujumbura for a longer time.Footnote 20 In other words, the seeming continuity of Indian or Arab long-distance trade is a continuity in numbers rather than a continuity in actual people.
However, although few businesses survived, the small group of successful merchants in the first postwar years remained the backbone of trade activities in and around Bujumbura throughout the 1920s and into the 1930s. It is worthwhile to identify this nucleus of roughly a dozen Arab and Indian merchants in Bujumbura around whom—together with even fewer Belgian, Greek, and Goanese merchants—Burundi's long-distance trade revolved.Footnote 21 Karmali Hansraj and Bandali Mohamed, the first Indian merchants to stay for many years, both arrived in Bujumbura in 1920. Alidina Ladak settled in the town in 1921. He and later his sons stayed in Bujumbura and were still commercially active there after the Second World War. Ahmed Nenshi, whom we encountered at the beginning of this article, returned to Bujumbura in 1921, having previously been there for a short time at the end of 1917. Janmohamed Gulamhussein Moledina also settled there in 1921 and ultimately died in Bujumbura in 1927.Footnote 22 His brother Jaffer, who traded in Gitega before Janmohamed's death, continued the business. Hasham Ladha settled in Bujumbura in 1921 and remained active throughout the 1920s. Ismail Fazel Shivji was registered in Bujumbura in 1921, but was mainly active in other trading centres along Lake Tanganyika (Kigoma, Rumonge, and Nyanza-Lac) throughout the 1920s. The majority of the twenty Indian traders who were active in Bujumbura between 1920 and 1922, however, left no further traces. It is not a coincidence that the seven traders who remained after 1923 (the year when heavy price cuts in the leather market caused difficulties for many traders) were among the eight Indian traders who had to pay more than four hundred francs in taxes in 1922. Only the biggest trading companies could survive. Taking into account that two of the seven most successful merchants died, in the end only five out of twenty-three Indians who tried to start a business in Bujumbura between 1917 and 1922 were still active in the same city ten years later.
In the early 1920s, Arab traders proved more successful and stable than their Indian counterparts. The eight Arab traders who had spent at least two years in Bujumbura before 1920 were all still registered as traders in 1922. Of the fourteen taxpaying Arabs in 1922, at least ten were still commercially active ten years after their establishment in Bujumbura. As with their Indian counterparts, the fact that ten out of fourteen Arab traders in 1922 had a firm big enough to accrue a tax load of four hundred francs or more proved to be a fairly reliable indicator of durability. The eight Indian and Arab traders paying six hundred francs or more in 1922 all survived their first ten years in town.
In the same period, the years 1920 and 1921, three Indian traders settled in the northeast Burundian town of Muyinga and surfaced in Bujumbura later in the 1920s and 1930s. Abdulgani Ali, Nathu Hassan, and Mohamedali Musaji had entered Burundi from the northeast and were connected to the world market via Bukoba on Lake Victoria and the railway to Mombasa. The predominant link with the Indian Ocean, however, went from Bujumbura over Lake Tanganyika to Kigoma and then by rail to Dar es Salaam.Footnote 23
By 1928, on the eve of a local and global economic crisis, the number of Arab businesses based in Bujumbura had increased from fourteen to thirty-three. The number of Indian firms dropped from eighteen to thirteen during the same period.Footnote 24 If we also include Arab and Indian staff employed in the trade sector in 1928, we count forty-one Arabs and sixteen Indians. This number would remain relatively stable among the Arab trading community until the Second World War. The number of commercially active Indians rose to thirty-six in 1939.Footnote 25 However, in taking a closer look at individually identified Arabs and Indians in Bujumbura in that period rather than restricting ourselves to anonymous numbers, we notice that the Arab and Indian trade communities were in fact almost completely swept away. Of the sixteen Indian traders in 1928, there were only four left in 1939. The nominal increase is accounted for by newcomers. Of the forty-one Arab merchants in 1928, only seven were still around in 1939. Six of these seven were already there in 1920 or had been working for Arabs who were already active in Bujumbura in 1920. The same applies to the Indian side, where three of the four survivors of the crisis of the 1930s were traders who were already around in 1920.
Table 1. Indian and Arab Merchants with Permanence in Bujumbura in the 1920s and 1930sFootnote 26
There are also some indirect survivors, who did not personally remain in town from the early 1920s to the late 1930s but were responsible for the survival of a business. Ali bin Salim bin Amer Tawqi and Masud bin Ali bin Amer both continued the business of their respective father (a5 and a3). Both fathers were already in Bujumbura in 1918. Amarsi Mauji survived the crisis in the service of Shahador bin Shahadad (b1). Hamed bin Mohamed Ma‘mri and Seliman bin Sef Ma‘mri traded in the name of Nassor bin Salim Sinawi (a6), who was already registered in Bujumbura in 1918 and had branches in Rumonge and Nyanza-Lac. Ali Rawji (i8) also intermittently appears in the files from 1919 until after the Second World War, and thus survived the crisis.Footnote 27
In the end, all traders who survived the crisis of the 1930s had either already started their commercial activities before 1920 or directly or indirectly perpetuated businesses that had been set up during the first postwar years. The businesses of Ali bin Amer Marhubi (a3),Footnote 28 Nassor bin Salim Sinawi (a6)Footnote 29 and Salim bin Amer Tawqi (a5)Footnote 30 even date back to before the First World War.Footnote 31
Colonial Trade Policy
The large number of Arab and Indian traders who did not succeed can be attributed to several factors that mutually reinforced each other in the years around 1930. We will look at the effect of deaths and bankruptcies at a time of economic crisis in the next section, but before we move on, the discriminatory colonial regulation of the commercial sector should also receive attention. The creeping effect of cumulative discriminatory measures profoundly altered the landscape of Indian and Arab trading life between the early 1920s and the mid-1930s. It was the explicit aim of the colonial government to direct the commercial sector to the benefit of Europeans, and that this would happen at the expense of the (mostly Muslim) Asiatiques (Asians) was a not unwelcome side effect for the Belgian colonial administration. The existing image of Indian and Arab merchants at the time of the Belgian takeover was straightforward:
This core group with small profits, if we except a few lucky ones, have the trading sector of this country in their hands. Deliberately starving, they criss-cross the different regions in search for traffic, of whatever kind it may be. This bunch of parasitic strangers tends to lose importance thanks to the measures taken after our occupation.Footnote 32
This view is obviously a biased sketch of the situation in which Indian and Arab traders found themselves, but gives a clear insight into colonial aspirations relating to commerce, and is also in line with the aforementioned hardworking and ascetic reputation of Arab and Indian merchants.
During and immediately after the First World War the first priority of the colonial government was to get the economy back up and running, and for that purpose the contribution of Arabs and Indians was still desirable, even indispensable, from a colonial point of view. After the regime change, Arab and to a lesser extent Indian traders were needed to revive regional and long-distance trade. Although many merchants failed and also left in those days, the small scale and diversity of their business model enabled some of them to diversify by temporarily switching to agricultural activities; returning to the leather and textile trade as soon as this became possible again. From the 1920s and even more so in the 1930s, however, this same small scale was no longer a good basis for confronting market fluctuations, recessions, and crises. For small enterprises it was increasingly difficult to take the blows of economic volatility. This mechanism is all too familiar in economic history, but colonial trade policy gave the so-called free market an all too visible helping hand.
The Belgian policy aim was “that Europeans should [. . .] centre their activities in commerce, industry, and transport.”Footnote 33 The downside of this policy was that Indians and Arabs were forced into retail and distribution in the interior. They had already taken care of retail and distribution previously, but now they gradually lost the international or long-distance part of the trade, and that was a new evolution. Their small-scale operations were now embedded in a system which as a whole functioned on a global scale and in which they no longer had much impact on how the different links of the chain fitted together. This also altered the role of long-distance kin connections in the practical operations of long-distance trade, in which they had been reduced to an intermediary role.
Specifically in the domain of exporting agricultural produce (hides, wax, honey, palm oil, coffee, cotton, etc.), most Indian and Arab traders were reduced to an intermediary position between farmers on one side and colonial enterprises on the other. They were, however, still unbeatable in this local intermediary role, because they were willing to accept very small margins—“deliberately starving,” as our colonial administrator phrased it in 1921.
For the European trading companies, Arab and Indian (as well as Greek) intermediary activities constituted an easy and inexpensive way to acquire and sell produce without having to hire expensive branch managers or employees:
[The Greek, Indian, and Arab traders] have only a very minimal capital at their disposal. They need moneylenders; [. . .] an agile lender will make a contract with his client; the repayment of the loan will be payable in specified native produce, at the cost of the day. This is a very simple means to drain produce from the secondary market [. . .] without everywhere having to create agencies headed by a European, which would lead to great cost.Footnote 34
Credit had always played a crucial role for Indian and Arab traders in East Africa. In the nineteenth-century caravan trade, merchants often went out with borrowed goods, which they intended to repay with the proceeds from the return cargo.Footnote 35 However, if the price of the return product collapsed, as was the case with the leather market in the early 1920s, or if government interventions put downward pressure on prices, as was the case with food in 1925, merchants depending on credit and small margins were immediately threatened with insolvency. In those circumstances European trading companies, which usually operated on a larger scale, could keep low-priced goods in stock until better times and, furthermore, bought those products at bargain prices. At first sight, there is a cynical but purely economic logic at play here, where the small players were pushed out of the market. Upon closer inspection, however, there were intentional distortions at work as well.
The colonial government, in addition to restrictions on immigration and residence,Footnote 36 repeatedly utilised economic circumstances to curb opportunities for Indian and Arab (primarily Muslim) merchants and to pin them down in a role as intermediaries. By making full use of the Indian and Arab contribution the colonial economy could prosper, and by giving Indians and Arabs minimal opportunities for growth the government strengthened the competitiveness of European companies. In this respect, it is telling that in 1921 one of the eighteen Indian and two of the fourteen Arab trading companies in Bujumbura had a larger turnover than the average among the twelve European trading companies.Footnote 37 European, mostly Belgian, businesses were on average indeed larger than the average Arab and Indian ones, but the largest Arab and Indian traders were operating on the same scale as the Europeans. To put an end to this more or less level playing field, and to impose racially segregated niches in the commercial sector, the colonial government had to ensure that Arab and Indian businesses would grow more slowly than the European ones. The subtlety of the sundry measures that would shape such policy ensured that Arab or Indian traders were never targeted explicitly or openly: rather, seemingly “neutral” measures would strike a blow to the Arab and Indian business model of the moment and safeguard the European company type.
As early as 1921, plummeting prices on the leather market were seized upon as an opportunity to impose administrative obligations on all traders without distinction. Every trader had to record his accounts and stock, ostensibly to help avoid future bankruptcies. The additional stipulation that the accounts had to be drawn up in Latin characters, however, in reality deterred non-European traders rather than avoiding bankruptcies.
By the end of the decade, in 1928 and especially 1929, a genuine famine raged in many parts of Burundi: allegedly in order to protect consumers, the colonial government prohibited sales on credit. This negatively affected the position of Arab and Indian intermediaries. Both the traders and the end consumer made transactions on credit. The colonial measure was meant to protect the latter. In reality the practice continued, albeit illegally. In many cases an ordinary peasant was not able to pay for delivered goods until after the harvest. In the case of bad harvests and inability to pay off debts, the trader who had acted clandestinely had no legal means to reclaim the debts. These intermediaries were mostly Arabs or Indians.Footnote 38 Rather than protecting the indigenous population, the prohibition on credit had in fact removed all formal and legal protection for petty merchants, often leading to insolvency. The smallest traders, who sold directly to the end consumer, were most affected by this measure, but as we will see, around 1930 the chain of credit relations escalated and brought some larger players into trouble as well.
Certainly, the small and relatively weak contingent of colonial civil servants was unable to enforce all the rules the government stipulated. Still, it was only possible to escape the official straitjacket in relative illicitness and by staying as inconspicuous as possible. In other words, both the official and unofficial scenarios favoured small-scale and low-profile operations and were complementary in creating separate commercial circuits for Europeans, Asians, and Africans. The crisis proved a further catalyst towards this partitioning of the regional trade sector.Footnote 39
Trading Life in Bujumbura before, during, and after the Great Depression
On the basis of six inheritance files and five documented bankruptcies from the years 1924 to 1934, I was able to reconstruct the dynamics and risks to which the Indian and Arab merchants were exposed. The files all involve European—usually Belgian—agents, which probably explains why the colonial archives contain these files in the first place. However, they do allow us to reconstruct the local and long-distance relations of the Arab and Indian traders, and to highlight both the central position of the postwar first-comers and the role of cross-community connections.
Predominance of the First-Comers
One of the two most successful Arab traders, according to the 1922 figures, was Salim bin Amer Tawqi (a5), and the second largest Indian trader, according to the same figures, was Ahmed Nenshi (i4). Both died in 1930.Footnote 40 These two men were not only among the biggest traders in terms of tax burden, but were also the traders with the longest track record in town.
Salim's commercial activity in Bujumbura began before the First World War. He lived there with his wife Alza binti Masud and in 1909 his son Ali was born in Bujumbura. Because of his status as a successful pioneer, he was the patron of many newcomers, especially from the Tawqi and ‘Isri branches of the Omani tribe al-Harthy.Footnote 41 He was also chairman of the Arab Committee (comité arabe) in Bujumbura, which represented the interests of the Arab community with the colonial administration. Despite his deep and long-standing roots in the town, he died in Muscat (Oman).Footnote 42 His relative Hamud bin Mohamed Tawqi, who had traded in Bujumbura since the early 1920s and was secretary of the Arab Committee at the time of Salim's death, became the new chairman of the committee.Footnote 43 A few years after the death of Salim, Hamud married his widow. Hamud's first marriage had been equally meaningful: he had married the daughter of another pioneering merchant, Adia binti Nassor bin Salim Sinawi (a6). This indicates how kin connections were maintained locally.
Ahmed Nenshi (i4) had come to Bujumbura in 1917, and in the Indian trading community of Bujumbura he was regarded as the man with experience. During the 1920s, several Indian traders who were absent for some time or had just arrived in Bujumbura appointed Ahmed Nenshi to represent or assist them.Footnote 44 On 14 February 1930 he died at the age of forty.Footnote 45 His death and the processing of his inheritance coincided with the severe economic crisis that was raging in the region as a result of crop failures in the late 1920s and the global depression. The commercial life was traditionally based on a chain of credit relations and deferred payment,Footnote 46 but if one link of that chain was removed, the whole cluster of debt and trust relations was affected and, at least when a European business partner was amongst the creditors, a legal procedure liquidating the estate was set in motion. In times of economic crisis this liquidation of one trader's outstanding debts and credits inevitably caused problems for other traders. The estate of Ahmed Nenshi had a value of about half a million francs, with assets and liabilities amounting to more or less the same amount. His debtors found it difficult to meet their obligations in these times of crisis. Auctions of his stock and other property generated less than the estimated value because of a clear lack of purchasing power at the time.Footnote 47 His creditors were also living through hard times, but due to difficulties in liquidating the estate of the deceased, they had to wait very long before they were paid their balances.
In May 1931, fifteen months after the death of Ahmed Nenshi, the inheritance file was still not closed. Because this specific case had a big impact on the whole trading sector in the region, the main creditors filed a request on 27 April 1931 asking that what was available at the time should be paid out. The person who had by testament been entrusted with the settlement of the estate was Polydore Edouard Van Hende, the Flemish manager of the export company Estaf, which had been founded in 1917 under the name Société Générale d'Exportation (Genex). Van Hende supported the request. The signatories of the petition were six Indian traders: Nazarali Alibhai, who represented the firm Allibhai Panju & Sons in Bujumbura; Gulamhussein Alidina Ladak (son of i3), who represented Bandali Mohamed (i2); Kanji Virji, who had started his trading activities as a representative of Hasham Ladha (i6) but was now working on his own behalf; Pirsab Hussein Ali, who represented Ali Rawji (i8); Karmali Hansraj (i1) in his own name; and Bhura Meghji, who represented Hasham Ladha (i6). Apart from Jaffer Gulamhussein, all major Indian traders who had survived the first postwar years either endorsed the request themselves or their representatives or successors did. Conversely Allibhai Panju is the only person in the list who had not already been active in or near Bujumbura in the early 1920s. His company became prominent only in the 1930s and was still active in the 1950s.Footnote 48 Pirsab Hussein Ali would become a representative of Allibhai Panju & Sons in the 1930s, which indicates that this company was part of the same circuit as Ali Rawji (i8). It shows that the Indian part of commercial life in Bujumbura was still in the hands of the pioneers and their entourage.
At about the same time, on 5 May 1931, the (acting) chairman and the secretary of the Arab Committee, Masud bin Habut Tawqi and Hamud bin Mohamed Tawqi (a1), together with the Indian trader Amarsi Mauji, who had been employed by the Baluchistani merchant Shahador bin Shahadad (b1) but was by then already working on his own account, and the Goanese merchant de Souza, who was active in Bujumbura from 1917, asked for a meeting with the Governor of Ruanda-Urundi seeking measures against the crisis in the commercial sector. They requested that payment problems be dealt with by mutual agreement without litigation and without formal charges, let alone bankruptcies. They also demanded that bankrupt traders should no longer be automatically expelled. The meeting took place on 6 May, but did not lead to any direct result. Notably, three of the four signatories were suing each other for belated payments or delivery of goods at that very moment. In March 1931 Hamud bin Mohamed demanded that the agreed delivery of ten tons of flour to Amarsi Mauji should go through, and around the same time there was a dispute between de Souza and Amarsi Mauji about the purchase of a lorry. The situation was apparently so serious that despite—or perhaps because of—their mutual business conflicts these traders acted together.Footnote 49
Within a period of just ten days, all the important trading communities in Bujumbura, across national or ethnic differences, cried out, and did so together. Regardless of their origin, these European and Asian, Muslim and Catholic, Indian and Arab traders had, directly or indirectly, been active in Bujumbura since the early years of the Belgian colonial administration. Trading life as a whole revolved around a small group of first-comers.
Entangled Communities
The description of these dynamics takes us to the heart of a second reason traders did not “survive,” namely bankruptcies. When someone could no longer cope with accumulated debts and obligations, he had to close the books, but this could in turn cause problems for the creditors as well. The same mechanism as for the handling of inheritances would then begin to operate. Already in the 1920s it had become clear how inheritance and bankruptcy cases were intertwined. What both procedures have in common is the fact that at a fixed time all assets and liabilities, debt and credit relation are mapped and disentangled and are expected to be paid in due course. Since trading was thoroughly based on credit relations, such reset operations were a threat to the system and could possibly undermine it. Especially when a file with a deficit had to be closed, and when simultaneously the creditors themselves were already going through difficult times, this could cause a vicious circle. Even if there was no deficit, as in the case of the estate of Ahmed Nenshi, the crisis situation led to an outcome where there was nothing left for the heirs, which from their point of view was undoubtedly equally dramatic.
Two cases from the 1920s, one of which dragged on until 1931, illustrate this entanglement between inheritance and bankruptcy, as well as the entanglement between traders of different origins. The bankruptcy of the Indian trader Kassamali Chandu in 1925 was triggered by a complaint filed by Mohmad Moorji in July 1923, demanding that he pay up an amount of 5,944 rupees (55,873 francs). In March 1923 Mohmad Moorji, then residing in Kabele, on the border between Uganda and Ruanda-Urundi, had raised a claim against an employee of the Indian trader Mohamedali Musaji (i11) in Muyinga. Probably both claims were entangled, if not one and the same, and constituted the reason for his coming to Bujumbura.Footnote 50 Mohmad Moorji called for the assistance of Ahmed Nenshi (i4) in this case. However, even before the bankruptcy was pronounced, Mohmad died in Bujumbura on 29 July 1924.Footnote 51 The Bujumbura part of his legacy showed a positive balance of 32,826.80 francs, but that included a credit of 33,375 francs from the bankruptcy of Kassamali Chandu.Footnote 52 Mohmad Moorji had mainly been active in Kabale and had outstanding debts in Kampala, Kabale, and Zanzibar, but these could not be paid as long as the bankruptcy of Kassamali Chandu had not been settled. Apart from Mohmad Moorji, the main creditors of Kassamali Chandu were the Belgian company Genex / Estaf and the Indian trader Janmohamed Gulamhussein Moledina (i5). The Arab traders Mselem bin Salim Shebi (a4) and Nassor bin Salim Sinawi (a6) had also been trading partners of Kassamali, but for smaller amounts. These debt relations illustrate how the three major trading communities in Bujumbura (Arab, Indian, and Belgian) did business with each other.
The second case study is the most extensive record of the period discussed here, namely the legacy of the Indian trader Dada Juma bin Juma Achim.Footnote 53 He died in July 1929 and left behind debts of 1,261,200.74 francs, assets of 835,804.97 francs, 140,000 francs in stock, and 123,132 francs in property. His principal associates were the already mentioned Abdulgani Ali (i9), who was appointed as the successor of Dada Juma in his will of 1928, and Nathu Hassan (i10), who would take care of the interests of the Indian traders during the settlement of the inheritance. Nathu Hassan and Abdulgani Ali had already done business together in the early 1920s in Muyinga, where they had gone bankrupt in 1924. Dada Juma apparently belonged to the same trade circuit, connected to the Indian Ocean world via Bukoba and Mombasa.
Almost all traders from Bujumbura appeared on the list of creditors and debtors, regardless of origin. The main creditors were in order of magnitude the Goanese de Souza (nearly 400,000 francs), the Indians Allibhai Panju & Sons (over 200,000 francs), the Belgians of Genex / Estaf (over 120,000 francs), the Indians Bandali Mohamed (i2) (nearly 100,000 francs) and Hasham Ladha (i6) (over 80,000 francs), the Arab Hamud bin Mohamed Tawqi (a1) (over 50,000 francs) and the Indian Ali Rawji (i8) (nearly 50,000 francs), accounting for about one million francs of Dada Juma's debts. Of these assets more than three-quarters were ultimately paid. Of the debtors, who also included Indians, Arabs, and Europeans, no less than six had an inheritance or bankruptcy pending at the same time. The local nesting of bankruptcy and inheritance files as well as the commercial entanglement of the different trading communities becomes clear.
Trade operations between merchants from different communities are not surprising, but their relationships were not limited to purely commercial transactions. As mentioned earlier, the Fleming Van Hende was designated by Ahmed Nenshi (i4) as the person to handle his inheritance, and four Arab, Indian, and Goanese traders jointly called for a meeting with the Governor of Ruanda-Urundi. We also find an example in the inheritance file of Dada Juma of intercommunity interaction that went beyond purely commercial transactions or credit relationships. When the Indian trader Dada Juma had his will composed in 1928, an Arab merchant was his witness. Likewise, the will of the Baluchistani Shahador bin Shahadad (b1) had been registered using two Arab witnesses.Footnote 54 There are also numerous examples of branches of Indian, Belgian, or Goanese trade firms that were managed by members of another community.Footnote 55 Admittedly, these examples that go beyond strictly commercial transactions are still directly or indirectly related to business. They do show, though, that the local trading communities were more intertwined than usually becomes visible in diaspora studies. They also show that local transactions and relations along Lake Tanganyika rather than long distance were the lifeblood of everyday business operations. The next section gives a glimpse of how these local operations relate to long-distance connections.
Communal or Commercial Connections with the Home Country?
In the interwar period Bujumbura was the most important commercial centre of Ruanda-Urundi. From Bujumbura, established traders opened branches in the rest of the mandate territory. Prominent traders from Bujumbura like Nassor bin Salim Sinawi (a6), Hamud bin Mohamed Tawqi (a1), and Ahmed Nenshi (i4) had representatives in Nyanza-Lac; Salim bin Amer Tawqi (a5) in Astrida; and Dada Juma, Abdulgani Ali (i9), and Nathu Hassan (i10) covered Bujumbura, Gitega, Muyinga, and Nyanza-Lac.Footnote 56 Regarding spatial hierarchies between trading cities within Ruanda-Urundi, Bujumbura stood clearly at the top. Branches were almost always in the interior.
In the wider African Great Lakes region, however, the situation was different. Seen from Kigoma, Bujumbura was the secondary trading post. Especially in the Indian trade community, the Bujumbura traders were often branch managers of companies from Kigoma. The estate of Dada Juma, for instance, showed that Allibhai Panju & Sons, Bandali Mohamed (i2), Hasham Ladha (i6), and Ali Rawji (i8) were the largest Indian traders active in Bujumbura around 1930. All had their headquarters in Kigoma.Footnote 57 This also explains why the petition of 27 April 1931 was signed by the representatives of these four traders, not by themselves. The main Indian traders who resided permanently in Bujumbura were Karmali Hansraj (i1) and Ahmed Nenshi (i4), but they were actually only second-tier merchants. The pecking order within the Indian trading community was so clear that the chairman of the Indian Committee in 1928 was Pirsab Hussein Ali, the representative of Ali Rawji (i8), and in 1929 Kanji Virji, the representative of Hasham Ladha (i6)—not one of the traders working on their own account in Bujumbura.Footnote 58
The hierarchy of trading cities on Lake Tanganyika is clear; much less clear are the commercial relations between Bujumbura and the respective homelands around the Indian Ocean. Almost all major Arab traders in Bujumbura were born in Muscat, and if not, they were born in Bujumbura to parents coming from Muscat. This direct link with Oman was especially important when arriving in the region or when handling inheritances, but there is no evidence of direct trade relations with “home” or with relatives on the east coast of Africa in the period under scrutiny. Upon arrival at Lake Tanganyika kinship was important for someone to get installed in the country and to make first contracts. I do not want to underplay the importance of these first steps, but once they were in place the local conditions—which indeed included local kin—mattered most. In the records of inheritance cases, only traders in the immediate vicinity of the lake appear as debtors, creditors, or staff. The link with Muscat only shows up in the list of heirs. This might have been different in earlier times, when Indian and Arab traders had not yet been pushed into intermediary roles and still controlled large stretches of the long-distance trade, but in the period under scrutiny, long-distance relations with kin and family were important for the arrival and the start-up phase, but not for business operations as such.
In general, this also applies to the Indian traders. Many Indian traders were born in Bombay and Kutch; a significant number also in East Africa: Zanzibar especially, but also Lamu or Kigoma.Footnote 59 Their origin suggests a link with the Indian Ocean world, but their trade relations nevertheless seem to be limited to the specific region where they were active. Only very rarely do we observe long-distance debt relations in inheritance or bankruptcy files, such as a trifle of 331 shillings in the huge file of Dada Juma, or the previously mentioned case of Mohmad Moorji, who at his death still had debts in the Ugandan cities of Kabale and Kampala as well as in Zanzibar. However, the latter died in 1924, which might mean that Indian merchants then still controlled part of the long-distance chain, which they would lose by the next decade.
Their actual trading activities may have been geographically restricted; the traders themselves were not. Mobility was an essential feature of the Arab and Indian trade communities in East Africa. The mere fact that they came from Oman, Gujarat, Zanzibar, or elsewhere to Lake Tanganyika already proves this, but movement remained part of their lives. Welcoming and employing relatives who came over to Bujumbura was an expression of adherence to the communal structure. Long-distance or family relations were important upon arrival or in times of need, but apart from that, social life mainly unfolded locally. This shows in the number of established families, reflected in an increase in the number of Indian and Arab women and children in Bujumbura. In 1928, the share of women in the adult Indian and Arab population was only 30 percent; in 1946 this had risen to 40 percent. In the same period the number of Indians and Arabs younger than fifteen years rose from less than 30 percent to more than half of the Asian population.Footnote 60 The population register also shows a growing number of elderly widow-mothers who spent their old age with their children. By 1946 at least eight Indian widows lived with their children in Bujumbura. There are also examples of a sick father who came to live with his son, and a seventy-year-old man who came to live with his younger brother.Footnote 61 These data indicate that Arabs and Indians were organising their life locally and in all likelihood were planning to remain in the area, perhaps permanently.
In everyday life, the family structure was less part of the commercial apparatus than it was a system of social security that could cater for work, trade and/or marriage partnerships, and care for old age. This informal security system worked in both directions, and the country of origin was not necessarily the heart of the whole system. Following the death of his son Ahmed Nenshi (i4), Nenshi Somji, who lived in Kutch (India), literally expressed his dependence on this informal social security:
The maintenance of my family and myself chiefly depended upon my [. . .] late son Ahmadbhai. [. / .] I humbly request your honour will be kind enough to make fair, just settlement of the estate, goods and ready cash of my late son Ahmadbhai Nenshi, between I the father and the children of Ahmadbhai Nenshi through the African lady whom he lawfully married, after clearing the accounts of debtors, partners etc., according his will.Footnote 62
The will to which this fragment refers made it clear which relationships had a claim on the estate: the children, the father, the business partner, the life partner, and the needy. The fact that Ahmed Nenshi had bequeathed a fifth of his estate to charity or poverty relief is not unusual for a Muslim. The will of Shahador bin Shahadad (b1) also expressly stipulated that alms should be given to the poor.Footnote 63 Since, in the end, the estate of Ahmed Nenshi was closed with a slightly negative balance, this distribution never actually took place, but that does not undo the principle and the intention of solidarity on a family, business, and community level.
The bond with the homeland was clearly important, but it was a long-term interest, which was intended as social security or a safety net. Trading life itself took place locally, in town, with the interior, and between Bujumbura and Kigoma. By 1930, the connections with “home” were communal rather than commercial.
Conclusion
Arab and Indian traders on Lake Tanganyika found themselves in a vulnerable position, which caused grave difficulties in times of economic crisis. Their characteristic small profit margins meant that many traders could not overcome a severe setback. Eventually it appeared that the merchants who survived the Great Depression of the early 1930s were either already present in the region in 1920 or were associated with those first-comers. Parallel to these happy few, there was a huge turnover among Arab and Indian merchants, suggesting that quite often their efforts were not successful. Their total number remained more or less stable and the continuity of trade itself was guaranteed, but apart from a core of long-standing merchants very few individual traders stayed for longer periods in Bujumbura.
A notable difference between the leading Arab and Indian traders was location: the main Indian traders in Bujumbura operated from Kigoma, whereas the leading Arab traders were located in Bujumbura. Within the Arab trading community, Bujumbura was clearly the focal point for all of Ruanda-Urundi. Arabs were also more active in the interior than Indians.Footnote 64 Within the Indian trading community, on the other hand, Bujumbura, although undoubtedly the main trading post within the Belgian mandate territory, remained subordinate to Kigoma. All commercial life, however, deployed against a background where the Arab and Indian traders gradually lost control over international and long-distance trade to European companies.
Long-distance relationships with family or home remained salient and were of great social and symbolic importance in the life of Arab and Indian merchants. Nevertheless, they played only a small role in their daily doings. Although trade in the Great Lakes region was obviously part of a much longer chain, individual business success was earned on the spot and the longer distances were beyond their control. Nonetheless, family ties had an undeniable importance at the beginning and at the end—both of business and of life. At the beginning of a career kinship could provide access to a trading network. At the end, when a business went bankrupt, at old age, or in case of death, family ties were important as a safety net. However, for Omani and Gujarati merchants in Bujumbura the commercial significance of ethnic and kin connections had been subordinated to colonial global capitalism by the 1930s at the latest. Forced into an intermediary position in the long-distance trade, the day-to-day business relations were increasingly a local, cross-community, and above all precarious affair.
Acknowledgements
Earlier versions of this paper were presented at the 55th Annual Meeting of the African Studies Association (Philadelphia, 1 December 2012), at the 18th Workshop on African History and Culture (Polenz, 2 October 2020), and at the 5th Annual Conference of the Collaborative Research Centre, “Processes of Spatialization under the Global Condition” (Leipzig, 10 October 2020). I am grateful to Adam Jones, Dirk van Laak, Nick Gestrich, Jochen Lingelbach, and the two anonymous reviewers for their useful comments on successive versions of this article. I also want to thank the Research Foundation–Flanders (FWO), Ghent University, and Jean-Marie Nduwayo for enabling the field and archival research that provided the invaluable data for this research. I wrote the article in the context of the Collaborative Research Centre “Processes of Spatialization under the Global Condition” (SFB 1199) at Leipzig University.
Competing interests
The author declares none.
Open access
