publisher colophon

3Knowing a Fly

A chipukanana (insect) that threatened Rhodesia’s foremost economic sector—agriculture—had to be studied and taken seriously. In 1961, the entomologist Edward Bursell called it “an intelligence system of tsetse,” an infrastructure and system of knowledge production so thorough as to know mhesvi in the most intimate way. The entomologist was talking about an anthropomorphic kind of kuziva (knowing), getting into a mhesvi’s intimate life, “to live and breathe and think with it.” The goal was no longer simply studying mhesvi, but to engage in “a lifetime of affectionate study.”1

Such intimate knowing required placing mhesvi under surveillance, to know how much time it spent in different parts of the habitat at different times of the year, how much time it spent feeding, sleeping, or simply waiting to pounce on anything that moved.2 It also required mapping mhesvi (which vachena now called Glossina), establishing its boundaries, and knowing where to mount defenses and offenses against it.

That way of kuziva mhesvi had begun less surely and developed in a meandering way. Its journey is the subject of this chapter, starting with the institutional structure for kuziva mhesvi and then moving to knowing how to find it, what it eats, and what to do with such ruzivo (knowledge).

To know mhesvi in order to control it, hurumende established what V. Y. Mudimbe (1988) calls the “colonizing structure,” albeit with important caveats. Mudimbe meant by this concept “the procedures of acquiring, distributing, and exploiting lands in colonies; the policies of domesticating natives; and the manner of managing ancient organizations and implementing new modes of production” (3). That definition fits the establishment of an institution upgraded from a division to a branch to a department and downgraded as needs must. I replace “domesticating natives” with “destroying mhesvi,” and “new modes of production” with “the production of knowledge” while cautioning against Mudimbe’s readiness to see such production as anything “new.” The reason is simple: Mudimbe bought too readily into and overestimated the overwhelming powers of hurumende to define what became ruzivo rwevatema (the knowledge of black people) while grossly underestimating the role of vatema in that which he assumes to be ruzivo rwevachena (white science).

Mudimbe says that the colonizing structure effects “the domination of physical space, the reformation of “native” minds, and the integration of local economic histories into the Western perspective” (Mudimbe 1988, 3). The domination of physical space remains the core of the book’s focus, but all three elements of Mudimbe’s formulation are top-down readings of white rule and open to further explorations. The “reformation of natives’ minds” masked a vast appropriation—theft—of ruzivo rwevatema that profoundly defined the vachena’s (never quite successful) efforts to dominate space. As a case in point, the work of Mhoze Chikowero (2015, 19−79) illustrates how white missionaries with their very soporific acapella music drew little interest from would-be congregants, their fortunes changing only when they allowed vatema to bring their own musical inventions into the church. Vachena spent the first sixty to seventy years battling to destroy a deadly chipukanana that vatema had coexisted with for millennia; just as victory was on the horizon, vanhu vatema rose in rebellion to demand their independence back. The “integration of local economies into the Western perspective” is only half the story and depends on one’s analytical location: It is true when a person is looking from vachena’s perspective. However, seen from the rural village and vatema’s everyday lives, a person also sees vatema’s integration of Western economies into the local perspective, thus completely appending Mudimbe. Hence, his is only one part of the story; in fact, such a top-down approach masks a subtler reality: that of the colonizing structure being founded on ruzivo rwevatema and their means and ways of doing things.

Finally, Mudimbe’s colonizing structure “completely embraces the physical, human, and spiritual aspects of the colonizing experience” (Mudimbe 1988, 2). I add a fourth element that science, technology, and society (STS) can bring richly into Africa studies: attention to inanimate and animate things as actors. In this case, the focus is on the thing to be known and what it does to deserve being known, as well as the things deployed in order to know it. The thing to be known is also a thing that is known by its deadly mobilities. It becomes the venue, the meeting point, of those seeking to know, their knowledge traditions, and the means (tools) of kuziva (knowing), tools that are not only human fabricated, but also dehumanized people—vatema.

Why did hurumende become so interested in kuziva mhesvi? What infrastructures were needed to enable hurumende to know mhesvi, and how do we account for their evolution? What was to be known about mhesvi? How did vachena go about the process of kuziva/knowing it? And, what do we do with and learn from that way of doing things that made hudzvanyiriri (oppression) strong, that made mulumbeti (the devil in xitsonga) strong?

Why Did Hurumende Become Interested in Knowing Mhesvi?

The rationale for the authorities’ interest in mhesvi lay squarely in the chipukanana’s mobilities and transmission of hutachiwana (pathogens), the local varieties of which were deadly to mombe (cattle). The beef industry itself is an important example of how Southern Rhodesia’s economy was built upon ruzivo rwevatema—in this case, ruzivo rwemombe (cattle knowledge).

The roaring entrepreneurial successes white people from all over the world experienced in Gauteng (the Rand) during the gold rush of the 1880s convinced them that a second “rand” lay north of the Limpopo. When governments could not sponsor the Rhodesian occupation, individuals formed venture capital companies to do so themselves on the strength of mineral concessions fraudulently obtained from local leaders.

Cecil John Rhodes was one of those men. In 1890, his British South Africa Company (BSAC) sent a Pioneer Column to occupy the land east of the Ndebele kingdom, which now became, to the British Government, the settler “colony” of Mashonaland. The idea of the “northern goldfields” as a land with footpaths paved in gold turned out to be a huge disappointment. Pioneer farm rights, quoted at £100 each when the Pioneer Column was disbanded, were selling for £55 to £60 each by April 1893, while just three hundred vapambevhu were occupying farms.

The “failed settlers” of Mashonaland now sought reprieve in the West. Three years after the occupation of Mashonaland, after mounting tensions, the BSAC invaded and subdued the Ndebele and renamed their lands Matabeleland. Not surprisingly, the lure of a share of Lobengula’s izinkomo and fertile land in Matabeleland attracted 922 vachena as volunteers; the war ended relatively quickly. Mashonaland and Matabeleland, together constituting all the land between the Zambezi (north), Limpopo (south), Kgalagadi (west), and Nyangani Mountains (east), became Southern Rhodesia.

Here too is a clear illustration of vachena building upon local resources—albeit ones the Ndebele had mostly pillaged from their neighbors in raids into dzimbahwe, Tswana, and Tawana territory. Following the war, some 948 land rights were issued. Those who had taken part were given first preference to any claim until April 26, 1894, by which time many were struggling to sell and get out. Like Mashonaland, Matabeleland had no pavements of gold. The quest to appease the pioneers with Lobengula’s izinkomo also stemmed from knowledge of a well-developed trade in izinkomo that had existed with the Damara, Namaqua, and Ovambo since the early 1870s, fed in part by the wagon traffic from Cape Town and Kimberley. A man named Balane had made several two-year expeditions from 1869 to 1886 to the territories, buying eight hundred to one thousand beasts for £1 each on his first journey, which he sold for £7 to £8 each in Kimberley. Another man, Ericksen, at one point bought twelve hundred oxen from the Damaraland, which he pushed south to Kimberley via Lake Ngami (Stigger 1971, 16).

The Ndebele, meanwhile, sold their izinkomo (cattle), imvu (sheep), and imbuzi (goats) to local Jesuit Missionaries for handkerchiefs, blankets, and cotton cloth, even though prices fell as missionaries gained local ruzivo. Serowe also bought izinkomo from the Ndebele and delivered them south in December 1882, while in late 1889, a man named Dreyer took delivery of 750 Ndebele izinkomo, having supplied part of the Rudd Concession rifles promised. As Stigger shows, by July 1893, izinkomo trading was said to be “the most important business engaged in by traders among the Ndebele, pieces of calico forming the principal item demanded in exchange, although guns, powder in flasks, beads, blankets, coats, waistcoats and trousers were welcomed” (Stigger 1971, 16). In that year, izinkomo trade between Matabeleland and vapambevhu of Mashonaland was quite pronounced. For instance, in March, one H. J. Hill and two vachena returned to Salisbury with three hundred izinkomo and four hundred imvu and imbuzi. Hill was one of several Mashonaland settlers—including notably C. M. Acutt, Colenbrander, Dawson, and “Matabele” Wilson—who were also involved in this trade (16).

Vanhu vatema in Mashonaland were also prepared to trade izinkomo, with those going to look for work in town or going to boarding school often passing through mission stations to sell goods and get calico or gold (see figure 3.1). They usually rode on mikono (bulls), which upon arrival or in transit they sold to vachena. For instance, in 1891, others in Manyika (now Manicaland) and the Zezuru around Salisbury (Harare) freely traded mombe for salt, beads, brass wire, calico, and old uniforms “to itinerant traders visiting their kraals.” They also traded at the auction in Harare, some for around £6 per ox. Together, Shona and Ndebele cattle satisfied local demand: “When taxation was first introduced into Mashonaland in 1893, any tendency for vanhu vatema to part with stock is said to have been discouraged because of the trouble involved in selling their beasts” (Stigger 1971, 18). Vachena’s purchase of cattle was largely consumer-driven.


Figure 3.1 Men and boys riding on mikono/bulls. In addition to being wealth and being important ingredients in ancestral spiritual sacrifices, mombe were deployed as transport to work in the cities and deployed to mission schools, sold or slaughtered once there for upkeep and dried meat. This picture illustrates the multifaceted roles of mombe in dzimbahwe society.

Source: Zambezi Mission Record.

From 1894 on, vachena became more arbitrary and aggressive than before. After the defeat of the Ndebele, a Loot Committee was established to plunder Ndebele wealth. Initially, it set out to impound thirty thousand izinkomo for distribution to vapambepfumi, but by December 1895 it had seized sixty-five thousand. These are only the reported seizures; tens of thousands more were seized but not accounted for. A settler named John Meikle in particular built his Meikles Stores and Meikles Hotel empire out of looted izinkomo and thousands of hectares he held for speculative purposes (Stigger 1971, 18).

Still, even with so much pillage, Southern Rhodesia was anything but booming by 1896. The entire settler community was in a state of total farming stagnation (Stigger 1971). That year, a devastating cattle plague (rinderpest) that had started in British Somaliland in 1888 and followed the wagon trails of British East Africa (Kenya and Uganda) and British Central Africa (Malawi and Zambia) arrived. In a typical case of “ecological imperialism” (Crosby 1986), vachena’s ox wagons had introduced a devastating nyongororo against which vatema’s zvipfuyo and mhuka had no natural immunity. They died en masse (Spinage 2003). The extermination of mhuka denied mhesvi its most versatile means of transport and food source. Only those mhuka in the remote borderlands along the Zambezi, Limpopo, and Savé River Valleys survived. In the vast acres of the now tsetse-free land, vachena established their mapurazi (cattle ranches, farms) and mines (Mavhunga 2014).

The rinderpest (and vatema’s risings from 1896 to 1897) capped a depressed atmosphere in Mashonaland that had begun with the failure to find gold; most vapambepfumi had left for Bulawayo after 1893. When the risings broke out, they finished much earlier in Matabeleland and dragged on in Mashonaland to the end of 1897. Then, a commercial depression and unemployment set in among vachena now reduced to paupers and miners taking refuge in town. Meanwhile, the social depression caused by sending women and children to safety out of Mashonaland, combined with locust invasions and the drought, worsened. From October to November 1897 alone, there were eight suicides among vapambevhu (Kosmin 1971). While vachena were deep in depression, vatema dominated crop and mombe/cattle production in Southern Rhodesia (Machingaidze 1980, 282).

When vachena turned to ranching, they could only (initially) build upon mombe they had forcibly taken from vatema between 1893 and 1898 as spoils of war (Samasuwo 2003, 489; Phimister 1978). These were draught-oxen and disease-resistant cattle varieties, like those that vachena called the “Nkone,” “Mangwato,” “Matabele,” “Mashona” (also called the “Hard Mashona”), and “Tuli” breeds (Government of Southern Rhodesia 1924, 30). The settling vachena were cash-strapped and could not import any breeds from Europe (Machingaidze 1980, 285).

The foundations of settler society were not only built with means and ways (technologies) of vanhu vatema; in the early period, the settling vachena also played second fiddle to vatema in terms of productivity on the land. Before 1923, “beef production in the colony was severely hampered by the lack of capital on the part of most vapambevhu, crude ranching techniques, rampant cattle diseases, lack of transport facilities and … the lack of remunerative markets” (Samasuwo 2003, 490). World War I offered temporary respite, but by 1921, the prices of beef had slumped again; the local market was too small. The veterinarians had no choice but to slaughter mombe, keep carcasses in cold storage, and can beef.3

The process that would lead to a complete shift in fortunes started in 1923, when a commission of inquiry recommended the establishment of an abattoir to pave the way for producing frozen and chilled beef for overseas export. Negotiations began with the South Africa–based Imperial Cold Storage and Supply Company Ltd to set up abattoirs in Rhodesia.4 The Rhodesian Export and Cold Storage Company Ltd (RECSCO) was formed in 1927 as a subsidiary of the South Africa–based parent company.5

It was not a rosy beginning. RECSCO’s first ten years in business posted staggering losses, beef prices remained depressed, and the subsidy the Rhodesian government was paying only fed the parent company’s bottom line, instead of benefitting Rhodesian farmers (Mlambo 1996, 57). Faced with a choice to turn RECSCO into a parastatal or become completely private, the government chose the former.6 Arguments were made that “once a concern is under state control, efficiency goes by the board” and that the entire situation would “end in disaster.”7 They were waved away.

Out of RECSCO was born the Cold Storage Commission (CSC; also known as COSCO) in 1938. The parastatal was established as a guaranteed market for white ranchers and to rescue them from the grinding effects of the Great Depression through the forced acquisition of vatema’s mombe on the cheap.8 In 1941, CSC turned its fortunes around so much so that it upgraded its Bulawayo abattoir; built three more in Salisbury in 1943, Umtali (now Mutare) in 1946, and Fort Victoria (Masvingo) in 1951; and installed cold storage facilities in Que Que (Kwekwe) in 1946 and Gwelo (Gweru) in 1947.9 In addition to totally dominating the domestic market, CSC also expanded its exports into the United Kingdom, Northern Rhodesia, South Africa, and the Congo.10

Rhodesians (as vapambevhu came to define themselves) knew how to look out for each other—particularly in the 1930s, when white and black cattle owners had lost 200,000 and 250,000 head of mombe, respectively, due to Great Depression–induced hardship. CSC’s “guaranteed prices and markets” policy for whites could not have come at a better time in 1938. The prices were good. The floor price in 1950 was 70 shillings per one hundred pounds, cold-dressed weight, increasing to 97 shillings in 1955 and 113 shillings four years later. The national herd also increased in the postwar period, from below 3.6 million head in 1948 to 4.2 million in 1955 and 4.75 million twenty years later.11 In 1960, CSC extended its operations to Northern Rhodesia and Nyasaland.12

The CSC was created to guarantee white farmers’ success. It was a racially discriminatory affirmative action infrastructure, a structure to create an industry to buy mombe from whites and guarantee them good prices, a market, and a sustainable settler economy. It was a government mechanism to deliberately give vachena an unfair competitive advantage and force vatema to sell their cattle. Before World War II, hurumende did not bother to organize the marketing system for vatema; rather, individual white traders, white farmers, speculators, and agents of CSC and Liebig’s Ranch (property of the beef giant, Liebig’s Extract of Meat Company) went around buying mombe in villages.13 No weights were used; prices were negotiated on the spot (Mlambo 1996, 64).

The CSC not only became an apparatus for prospering vachena; it also took its place at the center of disenfranchising vanhu vatema—not through exclusion from its market, but by forcibly seizing their mombe and giving them to white ranchers. One thing had become very clear by 1941: Despite their “excessive herds,” vatema were not selling (enough), and Rhodesia was failing to supply fighting troops at the front with tinned beef. The view of the Native Department was that vatema should be forced to sell their mombe and that their herds should be capped and the rest sold or forfeited to hurumende.14 Under the Natural Resources Act No. 9 of 1941, Section 36, the government decreed that every household herd be limited to “prevent overstocking.” The Native Department was charged with organizing mombe sales at designated markets. The 1942–1951 period was the most intensive one for forced destocking, perhaps equivalent only to the looting of Ndebele izinkomo after 1893.15 To ensure that vatema had no alternative market, all potential buyers were supposed to show a Cattle Sales Permit Order (Government Notice No. 603), issued for two purposes only: buying for slaughter or farming purposes. The noose would further tighten in 1947 with the Native Cattle Marketing Act (Act No. 23/1947), under which the Minister of Native Affairs was granted power to prescribe methods of sale, venues, and who could or could not buy. Only the CSC, Liebig’s, and butchers the CSC did not supply could buy. No such restrictions applied to white mombe owners.

Evidence showing the existence of this discrimination was suppressed. According to Mlambo (1996, 67), “Africans in their areas were being forced to sell cattle against their own wishes, … the prices paid for the cattle were very low, … cattle belonging to Africans who were absent at the time of sale were sold without their knowledge or consent and … those Africans who refused to sell their cattle were punished or were threatened with punishment.” A subsequent commission of inquiry was a sham; quite contrary to its own evidence, including Native Commissioners assaulting vatema who refused to sell, the inquiry dismissed these charges as lies and disclosed just how happy vatema were to destock. In Gutu District, for example, vatema told the Native Commissioner they would rather much slaughter their beasts than sell them—whereupon the official ordered the brushes of the cattle’s tails cut as a sign they had been designated for slaughter and would be confiscated if seen again, incurring a huge fine for the owners.16 Having stolen vatema’s land in the 1890s and almost two hundred thousand of their cattle from 1893 to 1896, vapambepfumi continued their looting spree into the 1950s. The CSC bought the cattle at arbitrary, giveaway prices, then passed them on to white farmers to fatten and sell for slaughter to its abattoirs—at monstrous profit.17

The cattle plundered from vatema built CSC—and mapurazi more generally. The looters normally targeted drought time, when mombe were just matore (walking carcasses), their prices rock bottom, and bought them for less than a song. They struck when the market was oversupplied and prices depressed and selected mombe still not yet fully grown, the prices of which were a fraction of adult values. Munhu mutema would never get a better price than muchena; there was a price for vachena (8 shillings 4 pence per 100 lb. weight) and another for vatema (4 shillings 7 pence). In 1951, nhimura (the Native Land Husbandry Act) introduced even more destocking measures. Free competition was only introduced in 1956, by which time vachena had made their money.18

What Kind of Infrastructure Did Knowing Mhesvi Entail?

It was mhesvi’s threat to the emerging cattle industry that forced hurumende to build a branch dedicated to swatting this fly. The shape, character, and composition of this branch can only be understood through the pestiferous mobilities of the problem insect. The constant changes—from a branch, to a department, to a branch, and in personnel and job descriptions—are consistent with the status of the war against the indefatigable mhesvi and its unending ebbs and flows.

In 1909, as the insect (which vachena called “tsetse fly”) threatened to get out of control in the Chegutu-Kadoma districts, the government of Southern Rhodesia set up a Division of Entomology within the Department of Agriculture. There was no operational—let alone research—infrastructure when Rupert Jack (see figure 3.2a) was appointed government entomologist; in fact, only with the arrival of James Keswall Chorley (see figure 3.2b) to take the post of assistant entomologist was Jack’s title elevated to chief entomologist. By 1921, these men had performed so little work that the division’s annual report for that year was only half a page long. Throughout the 1920s, experimental work was conducted by the two men and a few farmers, and even that research focused only on trying to understand mhesvi breeding and the mechanical transmission of both flies and hutachiwana (which vachena now called “trypanosome”) through on-foot traversal. So thin-staffed was the division that when both Jack and Chorley went out for fieldwork, the headquarters virtually ground to a halt.19


Figure 3.2a, b Rupert Wellstood Jack (left) and James Keswall Chorley (right).

Source: S2A3 Bibliographical Database of Southern African Science (left image); Proceedings and Transactions of the Rhodesia Scientific Association 1967 (right image).

To share and better coordinate the burden of controlling mhesvi and give policy advice to government, the Southern Rhodesia Trypanosomiasis Committee (SRTC) was established in 1928.20 Its chair was Llewelyn Bevan, under whose stewardship SRTC worked tirelessly to pull together different government departments and white farmers’ associations to offer unified resistance to mhesvi.21 In the 1930s, Jack and Chorley worked with SRTC to build up an effective research and operational unit.22 As a starting point, a departmental committee of inquiry was established in 1931 to audit the state of the Division of Entomology and its tsetse operations.23 The recommendation to place operations under the Department of Agriculture’s administrative officer while restricting the entomologists to research mhesvi was a disaster in practice. By 1932, the entomological officer assumed the direct control of tsetse operations under the authority of the chief entomologist.24 The key lesson learned was that operations and research were inextricably bound together and should never be separated.

With Bevan’s death in 1938, the SRTC fell apart. Jack had also left by 1942, leaving Chorley—now chief entomologist—to reconstruct the committee and equip it to coordinate the many departments involved in the control of mhesvi and n’gana (Cockbill 1968).25 Throughout Chorley’s tenure, the work of combating mhesvi—let alone knowing it well enough to offer the right kind of response—suffered from serious logistical limitations. As he remarked in his 1951 annual report, “The only vehicles which are really suitable for the work in most of the areas are those with four-wheel drive. Of these the branch now has five, of which one is very old and constantly gives trouble. Apart from these the branch now has four 3-tonners and 12 smaller vehicles.”26 The division was, in summary, still a rickety, ragtag affair.

Two important reorganizations took place in 1951 that fundamentally redefined the control of knowledge production focusing on mhesvi and n’gana. The first was the reorganization of the SRTC into a smaller unit, composed of “only those officers directly concerned with tsetse fly and its effects.”27 Second, the Division of Entomology was transformed into a full-fledged Branch of Tsetse and Trypanosomiasis Control (BTTC) within the Ministry of Agriculture. It now had its own director answerable to the minister of agriculture, a senior entomologist, three entomologists, seventeen tsetse fly officers (TFOs) stationed in operational areas, and two clerical staff. The department hired Dr. Gerald Cockbill, a trained entomologist, zoologist, and botanist born in Cardiff, as a senior entomologist under Chorley, who now became the director.28 The TFO’s job was to supervise all aspects of field operations—initially game elimination, forest clearance, and fence construction, and later the spraying of OCPs. The entomologists were responsible for all research and were stationed at the central laboratory in Causeway, in Salisbury (now Harare), or at several field research stations—principally Rekomichi, Sengwa, and Lusulu, all located within the Zambezi’s mhesvi-infested belt.

In 1953, another commission of inquiry was conducted to review tsetse operations, this time zeroing in on the long-running program of indiscriminate game elimination. The commission’s report was submitted to the minister in 1954, recommending, among other things, the elevation of the BTTC into a full-fledged Department of Tsetse Fly Control and Reclamation (DTTRC).29 The new department was established on July 1, 1956. It was, at best, an enlarged BTTC, with a director, a senior entomologist, seven entomologists, five senior field officers, twenty-five TFOs, clerical staff, African field assistants, and twelve hundred to fifteen hundred magocha and general workers.30

Chorley had already retired by March, and Gerald Cockbill took over as acting director of DTTRC.31 He held the position until John Ford, previously director of the East African Trypanosomiasis Research Organization, took over as substantive director on September 27, 1957, with Cockbill as his deputy—designated senior entomologist initially, then chief entomologist.32

The idea of knowing mhesvi thoroughly as a prerequisite for killing it reached its climax during Ford’s tenure. To achieve this, he threw everything into improving the research capacity of the DTTCR, aggressively recruiting qualified personnel capable of initiating and sustaining research and operations even when resignations, retirements, and other loses of staff occurred.33 At the height of his powers as director in 1957, Ford’s complement of research and operational staff is shown in table 3.1 (my own tabulation).34

Table 3.1 The Department of Tsetse and Trypanosomiasis Control’s staff establishment in 1957

Present establishment Proposed establishment
HQ DirectorDirector
Senior entomologist (senior personnel officer [SPO I]) Deputy director
Administrative and executive officer Senior entomologist (SPO I)
Technical assistant
Administrative and executive officer
Field 1 entomologist (SPO II) 1 entomologist (SPO II)
6 entomologists (PO) 5 entomologists (PO)
4 senior field officers 1 survey ecologist (SPO II)
26 field officers 5 senior field officers
25 Africans, branch IV, grades I and III 18 field officers
50 Africans, branch IV, grades I and III

Two key appointments reflect Ford’s bias toward academically qualified laboratory personnel with hands-on, in-house training in field skills such as marksmanship, on one hand, and administrators with an understanding of entomological and operational work, on the other. He insisted, for example, that his deputy director must be a scientist with experience in running tsetse operations. To pay for the post, he would cut two entomologist positions.35 His senior entomologist would be a full-fledged field officer rather than simply an “Administrative officer sitting at Salisbury headquarters—which had led to failure to obtain data on operational progress, costing the Department money.”36 Finally, Ford also pushed for and got (in 1959) the appointment of a survey ecologist, which he viewed as a prerequisite to enabling the department to produce its own original maps of tsetse, game, and vegetation distribution, thus promoting greater accuracy and saving field officers time they would otherwise spend producing their own maps. The following year, the cartographic section was expanded with additional staff and provisions for their accommodation.37

The days of DTTRC as a department were numbered, however. On April 1, 1961, it was incorporated as a branch of the Department of Veterinary Services, with the assistant director of Veterinary Services (Tsetse and Trypanosomiasis Control) now the most senior official of the branch.38 Two years later, the Federation of Rhodesia and Nyasaland came to an end, and in 1964 the position of director was formally abolished. Ford chose to retire instead of staying on to work for Rhodesia (formerly Southern Rhodesia), because its white minority defied Britain and declared unilateral independence rather than grant independence to the black majority.39

Gerald Cockbill assumed the reins from Ford at a time of escalating mhesvi menace. He consulted with senior staff and concluded there was a need for a second review of the mhesvi and n’gana position, which the government accepted in late 1964. It involved the reintroduction of game elimination, “albeit in a much-modified form, now known as selective game elimination, supported by selective application of residual OCPs to dry season resting and refuge sites of mhesvi.”40 Cockbill continued the research momentum gained during Ford’s tenure. Even more emphasis was now placed on in-service staff training.41 Hence, it was “felt strongly that any Senior TFO, Entomologist, or Senior Animal Health Inspector being transferred to a district where trypanosomiasis is a problem—lacking specialist knowledge—should attend a special course at the Central Laboratory in Salisbury.”42 In the past, efficient control measures had suffered from the lack of professional staff, especially entomologists.43

Chief Veterinary Officer (Trypanosomiasis) William Boyt lauded the incorporation of BTTC into the Department of Veterinary Services as “an unqualified success.” Liaisons between the animal health inspectors and TFOs on the ground and among entomological staff had shown “a steady improvement.” District veterinary officers and entomologists now understood each other’s roles and opinions, thus eliminating past “danger of friction and delay.” The cooperation with other departments and ministries was still very poor, however: “Not all Provincial Administrators or officials of the Southern Rhodesia Department of Agriculture and the Lands Department appreciate the necessity for the restrictions and facilities requested.”44

Among the most important appointment was that of immunologist, based at the Salisbury lab, a Food and Agricultural Organization (FAO) position that was always occupied by a woman; when she was not there, the experiments most critical to knowing the trypanosome passenger of mhesvi—and, indeed, its bloodmeal—ground to a halt.45 This Immunology Research Section of BTTC continually experienced high staff turnover. First, M. A. Bolton terminated her appointment as Immunologist in June 1965 on getting married, having laid “a sound foundation for further immunological studies.”46 She had also developed techniques and procedures to uncover changes that occurred in the blood of mhuka in response to changes in trypanosome presence and trypanicidal drug treatment. Although the senior veterinary officer had acquired “some of the more specialized techniques,” the laboratory technician Bolton had trained had departed for 4.5 months of military training.47 The branch’s luck seemed to be turning in 1970, but the new immunologist, E. M. Steinberg, also quit her job to get married. That vacancy was filled by another woman, V. W. Emslie, “who within a few weeks successfully took over the intricate routine.” By the end of September, she had “almost eliminated the backlog of samples awaiting identification.”48 Then she too left.

On April 17, 1972, Dr. Gerald F. Cockbill retired as assistant director of Veterinary Services (Tsetse and Trypanosomiasis Control), and Desmond F. Lovemore took over. Other significant appointments that year included R. D. Pilson as chief glossinologist, Glyn Vale as regional glossinologist (research stations and field investigations), and A. Marks, replacing T. J. Casewell as a regional glossinologist.49

Cockbill seems to have left a vibrant department still capable of attracting qualified staff despite the raging liberation war. Branch staff were particularly active in publishing, with Boyt, Davison, Hursey, Lovemore, Pilson, Robertson, and Vale contributing papers to local science journals such as the Rhodesia Science News and international ones such as the Bulletin of Entomological Research.50 Meanwhile, in 1971, Dr. R. J. Phelps, an experienced University of Rhodesia glossinologist who was on leave, agreed to assist the branch in research. Two glossinologists and twenty-one TFOs were also hired, and one TFO was promoted to a vacant senior field officer position. Eleven new posts were also authorized, effective July 1, 1973: two TFOs, eight learner/tsetse field assistants, and a clerical assistant. A twelfth position, senior TFO, was approved for an August 1, 1973, hire.51 Another high point was Regional Glossinologist (Research Stations and Field Investigations) Vale’s completion of his doctoral thesis, entitled “The Responses of Tsetse Flies to Their Host Animals” in 1973.52

Then the war escalated, control became increasingly difficult, and an infrastructure that had begun to take shape started unraveling. When we rejoin the archival record again, it is 1982, two years after independence. In the 1981–1982 operational year, the BTTC experienced further losses of experienced personnel who were either retiring or resigning, not least of which was Assistant Director Peter Napier Bax, who quit rather suddenly in April 1982. Gerald Davison took over as assistant director, with two deputies—Vale as chief glossinologist (research) and Hursey as chief glossinologist (operations).53 More white officers quit in the 1983-1984 operational year, citing “the more attractive salaries and conditions offered by the private and commercial sectors.”54 It was also common in this period for whites to leave because of the uncertainty of transition to, and being ruled by, a black government. Many whites were struck by a fear of the unknown. They had never been in this position before.

Putting a Finger on the Pulse of the Fly

How do you gather data about a highly mobile chipukanana? And, assuming you successfully reduce mhesvi’s mobilities to “glossinological data,” how do you make it actionable intelligence of mhesvi? This section discusses two aspects of the production of knowledge about mhesvi, focusing first on the flyround and then on experiments at three field stations.

The Flyround

Earlier, we discussed an old-timer named Mabuzana, south of Gogoi, who told Swynnerton in 1918 that ndedzi was attracted to him because of the black coat he was wearing. By 1928, a method of inspecting the presence or absence of tsetse in an area had emerged, in large measure thanks to Swynnerton’s publication of the 1918 research from Muzvirizvi and his subsequent experiments at Shinyanga, Tanzania. This method involved walking a black “bait ox” along the footpaths and tiko (villages) located near traffic routes. Out of these beginnings, a systematic method of detecting ndedzi presence was born.

The flyround was a path created through the bush that a party traversed periodically, catching any ndedzi they encountered and recording data about them. The word round referred to the early practice of visiting selected areas to make “timed catches” of tsetse at “stations” (markings) along the path. The round was divided into sectors fifty yards long or corresponding with vegetation communities to be traversed (EATTRRO 1953, 30; Potts 1930).

In the rural areas of Chibwedziva (Savé-Rundé area) and Nembudziya (Gokwe), the men who traversed these flyrounds catching ndedzi were known in the misha/tiko (village) as mafrayi (the fly people). “Kwaiva navanhu vaifamba nemombe (There were people who moved about with black oxen),” recalls Willias Chabata, who was born in 1936. “They carried zvihuka zviya izvi (these nets). They hooked the mhesvi trying to bite the jon’osi (ox) with the net. They used to walk here, following behind the ox.”55 Said one man: “They would set off from here—one person with a net plus his companion moving ahead, holding the rope tied around the black oxen’s horns. Right now [around 2 pm] they would still not yet have returned. They would walk and walk, then stop, for a while, because the mhesvi would know that there is food there, I better head that way if I’m to survive. And surely you’d see it land on the ox. Then the mufrayi (singular of mafrayi) following behind would swing his net and catch it.”56

Dense thicket meant a path had to be cut through or hoed, but the normal practice was to follow existing paths wherever possible. The path was marked using a wheel, a chain and post, or trees marked at designated intervals (ten to fifteen yards) or “stations,” the “sectors” in between clearly numbered to act as guides. The start of a round was marked 0, and subsequent sectors were marked with ascending numbers (EATTRRO 1953, 30; Ford et al. 1959).

The catching team was composed of two catchers—one recording, the other leading the ox (EATTRRO 1953, 9, 32). On approaching the start point, the catching party stopped about one hundred yards before the first station; every member had to ensure that no mhesvi was on him. On reaching the path, the mufrayi in charge wrote at the top of his notebook page the flyround name, date, and other details, such as colors for marking mhesvi for release, time of day, and so forth. Normally, the sectors were fifty to one hundred yards long; mafrayi walked at normal pace to the first station, stopped, checked for following mhesvi, caught any in pursuit or engorged on them or the jon’osi, and recorded them vis-à-vis the station post (32). Once details for each mhesvi were recorded, it was either killed and thrown away or marked and released. These catches were entered according to the number of the station; a weather note for that sector (J for Jua [sunny]; M for Mawingo [dull]; V for Mvua [raining]); the number assigned to mhesvi during marking if previously marked; male or female; whether teneral (soft, immature, and pale) or mature; the hunger stage if a nonteneral male; the location caught; flies in copula (entangled, mating); mhuka and fresh spoor seen; and other changes (grass burning, flooding, damage by nzou, etc.).

The coding system was based on dzimbahwe modes of meteorology and time, the actual words Zuva and Mvura being corrupted into fanakalo—that is, jua and mvua. Mafrayi also searched for zvikukwa, putting them into boxes and recording the number collected per searcher per standard working day and sorting them into whole or empty shells. Finally, flyrounds occurred at the same time every day, starting and ending at the time “when the length of a man’s shadow equals his height” in the morning and in the afternoon (EATTRRO 1953, 33–34). This is yet another example of the appropriation of dzimbahwe ways of calculating time, as opposed to using clocks, which vatema did not have.

The role of black juveniles in the collection of zvikukwa cannot be understated. “A team of African juveniles varying from eleven to fifteen in number and working under an African field assistant,” disclosed Desmond Lovemore, entomologist for Sebungwe District, “have been making regular collections of G. morsitans puparia since May in the Gwababa-Mbelele river area.” By September 30, 1958, they had collected 14,786 “apparently live puparia,” the majority going to the Salisbury laboratory “with a view to building up a laboratory population,” with a small proportion handed over to Dr. Robert Barrass of the University College of Rhodesia for “work on parasitism.”57 This was no isolated incident; another team of five juveniles under the supervision of a black survey assistant had been collecting zvikukwa and shells in the same area since early August. They collected 947 “apparently live” zvikukwa and 4,397 shells.58

It is not an accident that juveniles were sent to collect zvikukwa. In communities in which fishing exists, the adults always dispatch the boys in advance to the alluvial riverbanks to dig up or scrape the ground for nyongorosi (earthworms) or zvikukwa to use as hook bait. I performed this chore as a boy; I sent my nephews as an adult. One acquires a certain hunyanzvi and exact ruzivo of where to look, what signs to look for.

Locals in Nembudziya still have memories of hufrayi (the work of flycatching). Reuben Mavenge still remembers mafrayi using “chimumbure (net), the same one they used at the gate.”59 Another local resident still remembers mafrayi at the business center, “going out to look for mhesvi” early in the morning, returning at dusk. He used to hear his own father looking forward to catching more mhesvi so that his month’s wage would increase. Many mafrayi “did sensible things with the mhesvi money”—or simply spent it on beer and prostitutes at Gokwe, arriving home empty-handed. After pay, they would first go to Gokwe Center to surrender equipment, then go home. These mafrayi were not just anonymous; they were physical and social people with names—like Fema Ngonda, African assistant, flycatcher, and mugocha.60 He is no longer the faceless, nameless, and invisible “fly-boy” of the entomologist’s or TFO’s field report. He and other mafrayi like him were makers of knowledge of mhesvi that vachena called “glossinology” (glossina knowledge), and they created this knowledge in a physical and intellectual way.


Figure 3.3a, b The flyround (left); searching for zvikukwa (right).

Source: (left image); National Archives of Zimbabwe (right image).

Sengwa Field Research Station

One of the recommendations coming out of the 1954 review was for a need to redouble efforts to understand mhesvi and n’gana. To do this entailed establishing more research facilities to enable the collection of more reliable data. Yet it was only after Unilateral Declaration of Independence (UDI) from Britain, with Southern Rhodesia becoming Rhodesia and international sanctions kicking in, that the establishment of these stations and localization of laboratory testing really took off.

Established in 1965, the Sengwa Field Research Station focused principally on the mhesvi-mhuka relationship, seeking to understand the habits, distribution, and dynamics of njiri the warthog, dzoma the bushbuck, nhoro the kudu (see figure 3.4a, b, c), and nguruve yemusango the bushpig—the favored hosts of the mhesvi that vachena called G. morsitans. The investigation—a joint undertaking between research officers from Lusulu and the Department of National Parks and Wild Life Management (DNPWLM) area in Gokwe—was to be conducted in the Sengwa area.61 It complemented a similar investigation in East Africa on two mhesvirupani that vachena now called G. swynnertoni (after Swynnerton) and G. pallidipes (pale-footed mhesvi) as distinct from mhesvirutondo (now G. morsitans; Isherwood et al. 1961).


Figure 3.4a, b, c Njiri the warthog (left), dzoma the bushbuck (center), and nhoro the kudu (right), northern plains of Gorongosa National Park, Mozambique, in October 2011.

Source: Author.

At its inception, Sengwa was given a mandate to address four broad points. First, it needed to determine the host preference of mhesvi, through the collection of blood meals from pairs of plots within each of five distinct vegetation types numbered from R1 to R10. Monthly samples of male mhesvirutondo were taken from each plot using flyround parties to determine seasonal variations in the fat content, size, and age of mhesvi populations. The goal was to “provide a basis for the interpretation of data on the game and mhesvi populations and composition of tsetse blood-meals of each plot.”62 The research complemented that in East and West Africa that dissected both mhesvi and their zvikukwa for clues (Bursell 1959; Potts 1933; Jackson 1949).

The second point was to determine how changes in the population of mhuka may influence that of mhesvi. Sampling in the two blocks, four square miles each, showed that numbers, distribution, nutrition level, and diet of mhesvi were similar in both. In August 1969, the DNPWLM research unit conducted a trial within Lutope to determine the feasibility of removing njiri the warthog by capturing them as they came out of their burrows. The mhesvi were marked with distinctive colors and released “so that a rough estimate of the effectiveness of the capture technique could be obtained from the relative numbers of marked and unmarked warthogs seen subsequently. … The capture technique would be a feasible means of removing warthogs provided that two catching teams were used, and that the four-square mile area was surrounded by a pig-mesh fence to prevent re-invasion.”63 The routine sampling in Lutope and Sengwa intensified in October as entomologists sought to establish any changes in tsetse population associated with capturing njiri. Preliminary results indicated that the disturbance had no obvious influence on tsetse quantities apprehended on flyrounds and coming to bait oxen.

The third point, to determine the means by which host preferences operate, entailed two types of work: (1) distribution and behavior studies to detect any special relationships between distribution and behavior of mhesvi and its preferred host and (2) studies to determine characteristics that rendered preferred hosts attractive to mhesvi. Samples of mhesvi from flyrounds and of engorged mhesvi from stationary bait oxen were used in experiments focused on developing alternative sampling.

The fourth point involved the development of devices to track mhuka and thus map their movement patterns. Two self-supporting antenna masts—thirty-five feet and seventy feet, respectively—were installed to act as base stations for the radio tracking of njiri the warthog. Priority was placed on tracking radio-tagged mhuka and detailed observation of the daily activities of tame, free-ranging njiri, especially in the Sengwa Gorge camp area. The successful development of radio-tracking techniques in 1968 marked a major breakthrough in the study of njiri dispersion and behavior. Dr. A. E. Rogers was the point man in this experiment, and with his team of black field assistants he deployed equipment with “ranges of four to five miles when using fixed receiving stations, two to three miles with a portable Yagi antenna, and about three-quarters of a mile with a hand-held loop-antenna.” These devices made it possible to “home-in rapidly on tagged animals in order to observe them, and also to plot their movements from a distance by taking successive bearings on them from two or more points.”64

Five njiri were tagged with transmitters, and thirty-eight were tagged with plastic collars. The plastic collar was made from a PVC tennis court marker, on which symbols were painted with PVC paint. They served as valuable temporary tags for immature mhuka, “lasting until the animal overgrows them, when they drop off.” In all, 163 njiri were tagged by the end of 1968, with recaptures showing that ear tags easily shed off and resulted in “fairly high” loss of information. Ear-notching was suggested instead.65

Encouraged, the research station extended this approach to dzoma the bushbuck at the end of 1968. In 1969, however, the radio collar experiment encountered a setback due to the short transmitter battery life, which was limited to about six weeks. Only toward the end of the year did the project resume, following the availability of a more suitable, longer-lasting battery.66 The field collars attached to njiri continued to transmit beyond six months; the laboratory ones continued transmitting after fourteen months.67 Battery life was critical, because these devices would be deployed in remote areas.

Finally, it is worth pointing out that up until the declaration of UDI in 1965, all blood meal lab tests were performed at Lister Institute in London, through the good office of Dr. Bernard Weitz. Sera were usually frozen at and transported in frozen state from Rekomichi, Lusulu, Sengwa, and other field stations to the airport, then flown overseas for testing. The difficulty of this process spurred Rhodesia to build a blood meal identification laboratory at Salisbury Experimental Station. By 1969, the lab had experimented with three kinds of hosts: vanhu, zvipfuyo, and mhuka.68


The chapter has shown that economics drove hurumende yevadzvanyiriri’s need to know mhesvi—an acknowledgment of the danger this hutachiwana-carrying chipukanana’s mobilities posed to the cattle (specifically, beef) industry. The industry itself is another example of ruzivo rwevatema as the foundation of the hutongi hwavachena (white rule); it must not be seen in isolation, but within a larger context of pfungwa dzavatema (idioms of black people) that vachena appropriated to survive and thrive as settlers. As the white occupation began, white writers submerged these appropriations under the propaganda that any credible knowledge they found vatema possessing and practicing (especially of stone architecture, metallurgy, and agriculture) had been brought in or built by an intellectually superior, more civilized people (Bent 1992; Selous 1893).

Some of these myths have been decisively debunked elsewhere on the African continent. For example, between 1500 and 1850, blacksmiths’ strong, high-carbon iron outcompeted vachena’s inferior product, which had high sulfur content and was brittle (Thornton 1990–1991; Goucher 1981). The Portuguese (1470s–1600s) intended to bring their own engineers to revolutionize how metallurgy was performed and set about establishing a foundry in the part of the Kongo kingdom that is now northern Angola—but it was not to be. So inferior and uncompetitive was their steel that they ended up employing Kongolese smiths in their foundry in Angola and buying locally made iron for resale along the western coastline. Black merchants and middlemen also extended their indigenous entrepreneurial traditions to exert control over trade between the white merchants and the hinterland (Diouf 2000; Dumett 1983, 663; Henige 1977). Similar examples include the decisive role of locally domesticated crops (Brooks 1975; Bowman 1987; Clarence-Smith 1994), mining and metallurgy (Garlake 1978; Phimister 1978), and medicinal plants (Osseo-Asare 2014) in the development of the United States’ and Europe’s industry and science from the sixteenth century on.

We can now return to Abiola Irele’s controversial statement discussed in the introduction, in which he once said that the only future for Africa lies in turning and following Western culture and civilization. “It is of no practical significance now to us,” Irele said, “to be told that our forefathers constructed the Pyramids if today we can’t build and maintain by ourselves the roads and bridges we require to facilitate communication between ourselves, if we still have to depend on the alien to provide us for necessities of modern civilization, and if we can’t bring the required level of efficiency and imagination to the management of our environment” (Irele 1983, 3; echoing Towa 1971, 1979).

This view has its critics, who reject a total sublimation of tsika dzevatema or chivanhu (black culture) into an imitation of tsika dzevachena (white culture), because nobody can predict the identity or desirability of the outcome. They instead urge vatema to take all the positives they can get from outside, while maximizing the strengths of their own innovations (e.g., Gyekye 1997; Falola 2008). The example of mombe has shown how vachena, by force, expropriated one product of ruzivo rwevatemamombe—to anchor their enterprise. The answer to Irele also lies in a tour of vachena’s early iron, gold, and copper mines throughout southern Africa, if not the entire continent, which were established precisely where vatema used to mine, smelt, and process such ores.

The rationale for the mombe discussion was to put into relief why vachena had to study mhesvi so thoroughly and intrusively. The rest of the chapter explored how it did this, taking the reader first through the establishment of a “colonizing structure” for the control and eradication of mhesvi and gopé/n’gana. The critique of Mudimbe—and by extension Africa scholarship—lies in the absence of a serious consideration of things (other than people) as actors in need of space in the historical narrative. When Mudimbe talks about the physical, people, and the spiritual elements, he is serving notice for the impossibility of merely talking of “the thing in itself,” be it means and ways of doing things or zvinhu (things; singular chinhu) outside the political economy of contextual meaning.

Increasingly since the 1980s, some Western scholars of science and technology have insisted on treating humans and nonhuman things as actors or agents, what one such scholar calls a “parliament of things” (Latour 1993). This utopian vision takes for granted the status of being munhu (human), not considering the possibility that one minute, someone is munhu, and the next that person is designated chinhu and subject to treatment as an instrument for eliminating other zvinhu. Read from dzimbahwe, Harry Collins and Steven Yearley’s (1992) dismissal of Callon and Latour’s formulation of human/nonhuman agency as a game of epistemological chicken that purports to respect the latter’s agency even as such nonhumanness is assigned from “a human-centered universe” (311) is both reasonable and problematic—reasonable because Latour/Callon’s descriptions of such agency were far over the top, absent a thorough engagement with how those designated zvinhu speak or not; problematic because there are less controversial, more empirically grounded, context-specific ways of accounting for the agency of things designated zvinhu.

Mhesvi is one of those less controversial instantiations of insect agency. It posed a real danger through its pestiferous actions and caused the establishment of an entire structure and set of procedures dedicated to putting a finger on its pulse, with the ultimate goal of destroying it. This is how its location became a site of work, or a laboratory. The flyround shows mhesvi’s location as a site of mobile work. Where mhuka that mhesvi most fed on roamed, field stations (static infrastructures) were established. The major work, however, was at transient sites of work and involved transient work, with researchers tracking mhesvi ever alert and lurching onto anything that moved for food, while hutachiwana moved with and inside mhuka like njiri, nguruve dzemusango, and mhesvi itself.

I started from chidzimbahwe with culturally grounded understandings of zvipukanana and mhesvi. I have entered vachena’s language and ways of producing knowledge, in which mhesvi became Glossina. I started from traditions that deployed organic, environmentally friendly means and ways of managing and coexisting with mhesvi and hutunga; I am now taking the reader into an environmentally destructive, pollution-intensive moment in mhesvi control.


Additional Information

Related ISBN
MARC Record
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.