In the early 1960s, the global oncology world was abuzz with developments coming out of a small central African country. Denis Burkitt, a surgeon working at Mulago Hospital in Kampala, Uganda, had suggested that the aetiology for a severe childhood jaw tumour, later to be called Burkitt’s lymphoma (BL), was likely viral in origin. His findings implied that the tumour might be caused by an arthropod-vectored virus – an electrifying assertion that captured international attention and prompted a large-scale search for the first oncogenic virus.Footnote 1 Establishing the relationship between cancer and viral infection would not only be a great scientific advancement but also offer a more hopeful vision of the future, in which certain cancers could be prevented through vaccination.Footnote 2 As a new frontier in oncology, the stakes in ‘the hunt’ for an oncogenic virus were, therefore, very high.Footnote 3
The pursuit to find the virus responsible for BL spanned many countries, institutions, and laboratories around the world. However, the most vital terrain in this global search was neither a big metropolis nor a major research centre. Rather, it was an unassuming part of northern Uganda known as the West Nile District. In the 1960s and 1970s, this region played an instrumental role in cancer virus research – serving as the hub of a large-scale survey run by the International Agency for Research on Cancer (IARC). This study established a link between Burkitt’s lymphoma and Epstein-Barr Virus (EBV) and, ultimately, helped cement the relationship between vectored viruses and human tumour development.
The decision to base this vast international research project in the West Nile District was, in large part, due to the work of Ted (Edward) Williams, a missionary doctor who moved to the Kuluva area in 1941. During his 38 years there, Williams built up a carefully catalogued repository of patient records. It was the detail and reliability of these documents that led IARC to select the West Nile as its primary research site.
The value of Williams’ records lay not only in the data they contained but also in the methods they afforded and ideas they inspired. As Denis Burkitt noted in Williams’ obituary: ‘Ted was an innovator with an inquiring mind, and his meticulous records of local disease patterns were soon recognised by research workers as an important source of epidemiological information’.Footnote 4 Their content may have been what originally drew international researchers to the West Nile, but it was their completeness, format, and capacity for reuse that transformed them into valuable research resources, allowing health experts to approach the relationship between cancer and viral infection in clever and creative ways. The ever-expanding interest in and uses of Williams’ records and record-keeping practices came as a surprise to Williams himself. In his later life, Williams acknowledged that he ‘had never imagined that [he] would be “in on the kill” of the first [virus involved in human tumours]’.Footnote 5 However, it was the production and preservation of painstakingly accurate patient records that placed Williams and the staff at Kuluva Hospital at the forefront of important developments in tumour virology.
Williams’ records offer a unique insight into early practices of epidemiological knowledge production in East Africa, a region often left out of broader histories of the field. Home to Makerere Medical School and Mulago Hospital, Uganda emerged as a leading centre for cancer research in the mid-twentieth century.Footnote 6 Williams’ work was part of a wider transnational research landscape, traversing institutions such as the Uganda Cancer Institute and Kampala Lymphoma Treatment Centre. Yet Williams’ assiduous record-keeping and archived epidemiological work offer a valuable window into a pivotal period in African epidemiological knowledge-making. This paper draws on this rich repository to highlight the intertwinement of paper, persona, and place in scientific knowledge production. We show how Williams’ fastidious personality, his locatedness in the broader political landscape of colonial and independent Uganda, and the paper technologies he relied on, all weave together to produce distinctive forms of epidemiological knowledge.
Scholars have long taken an interest in the epistemological functions of pen and paper.Footnote 7 Figures such as Bruno Latour, John Law, and Hans-Jörg Rheinberger have demonstrated how science’s ‘scrips and scribbles’ ‘constitute one of the most critical steps in epistemic processes’, enabling new options of order and arrangement and attending the transition from ambiguous data to stable facts.Footnote 8 Historians have similarly elaborated on paper’s importance within the realm of medicine.Footnote 9 From the writing of reports and the collection of patient charts to papier-mâché anatomical models, historical studies have probed paper’s character, and shown its many uses as a collaborative and collective carrier of biomedical knowledge. However, these works concentrate primarily on the early modern and enlightenment period.Footnote 10 Few studies have looked at how paper-based practices have shaped medicine beyond the nineteenth century, and most accounts of medical record keeping in the twentieth century have tended to focus either on big data or the stories behind the many digital tools and technologies used to count, measure, and weigh our lives.Footnote 11
In this article, we argue that despite the creeping influence of computer technology, paper continued to serve as an important ‘bearer of meaning’ – a substance on which new epidemiological knowledge could be recorded, reworked, synthesised, and stored.Footnote 12 Working in a region where electricity was an infrequent luxury and digital technologies were scarcely accessible, Williams produced an extensive paper archive. Tracing Williams’ research trajectory and the data repository he helped create allows us to see the unique configurations of people, writing materials, storage systems, and socio-technical infrastructures that shaped how epidemiological data was made in the mid-late twentieth century.
Importantly, we argue that paper and person cannot be separated. We show how paper’s significance lies not only in its materiality, but also in the ways in which researchers’ minds, methods, and milieus shape what the paper comes to represent. Carla Bittel, Elaine Leong, and Christine von Oertzen have argued that the ‘making and keeping of knowledge was never just about words or writing, but inseparable from material and social practices embedded in daily life’.Footnote 13 In other words, it is not enough to focus only on written inscriptions; we also need to consider the hands that produced them.Footnote 14 In this article, we show how Williams’ disposition, and the colonial bureaucratic cultures that helped inform it, played an important role in how his records were made and later used. As Cal Biruk notes, ‘the lives of data and the lives of those who produce it […] are impossible to disentangle’.Footnote 15 Williams was a meticulous person, both in his professional and personal life. Functioning as a sort of ‘paper machine’, it was Williams’ attention to detail and commitment to routine record-keeping, rooted in colonial administrative practice, that allowed his work to maintain such a strong foothold in oncological virology and circulate within the field of cancer research for years to come.Footnote 16 In this sense, we locate meticulousness as a key modus for scientific knowledge production.
While paper and person(a) are inextricably interlinked, we argue that place is an equally important constituent of knowledge production. Williams’ locatedness in Uganda was a critical element of his record-keeping practices – both in terms of the possibilities afforded him as well as the limitations which pushed him towards particular modes of epistemic improvisation. In the early to mid-twentieth century, biomedical research in Uganda was ‘dominated’ by ‘a culture of British colonial medicine’ shaped by ‘World War II scarcity, colonial developmentalist policies, White paternalism, and the adventurous orientation of medical missionaries’.Footnote 17 As a missionary and agent of empire, Williams worked within systems of colonial governance that afforded him degrees of flexibility, jurisdiction, and sovereignty not enjoyed by doctors in the UK.Footnote 18 For instance, he could draw on a large pool of African skilled and unskilled labour and pursue clinical and research activities across an expansive domain with relatively limited ethical restrictions – conditions that were embedded within racialised hierarchies of authority and broader structures of colonial harm. Equally, Williams and his work had to contend with the rapidly shifting terrain of Ugandan postcolonial independence as well as Idi Amin’s violent and authoritarian rule. This upheaval, and the political turmoil that came with it, had a profound effect on the conditions under which research and data collection could take place, as well as the custodianship of the records they produced.
By following the ‘socio-material paper trail’ of Williams’ work, this article explores the role of person, paper technology, and place in the making of epidemiological knowledge, particularly in colonial and postcolonial contexts.Footnote 19 While self-effacing objectivity has long been a central feature of how we think about modern science, this history helps show how the ‘scientific self’ and a researcher’s lived environment have a fundamental impact on how written objects are handled, systematized, condensed, mobilised, and negotiated.Footnote 20 As we will see from the following sections, the unique combination of personality, colonial organizational practices, and politics enabled Williams’ records to advance creative reasoning and biomedical discovery, fuel large-scale epidemiological, serological, and virological research, reinforce knowledge/power hierarchies, and ultimately, take on a life of their own.
Making meticulous records
Edward (Ted) Hammond Williams was born in Nairobi, Kenya, to a British family with deep ties to both the church and colonial service. Trained in medicine at London’s St Bartholomew’s Hospital, Williams began his career at London’s Mildmay Mission Hospital before returning to East Africa as part of the Africa Inland Mission – an influential Protestant movement that fused evangelism and self-discipline with education, local religious organizing, and medical work.Footnote 21 In 1941, Williams and his wife Muriel, both devout Christians, set out for Uganda. Their voyage from England took nearly three months. Their first home was a small thatched house at the Mvara Mission station near Arua in northern Uganda.Footnote 22 There, Williams as doctor and Muriel as nurse, began providing basic medical services to patients ‘on [their] front verandah’.Footnote 23 Determined to offer more substantial medical support, in 1946 Williams purchased a plot of land a few miles south of Arua and began constructing a leprosarium and hospital complex from scratch. The chosen site, beneath Kuluva Hill, was accessible only by a dirt road from Arua.
Williams was the archetypal colonial ‘medical pioneer’ coming to an exocticised and romanticised ‘Africa’ to practice a form of idealised ‘heroic medicine’ that was ‘brimming with mundane safari adventure stories, clinical discoveries, and the challenges of scaling medical care’.Footnote 24 Since boyhood Williams had ‘felt called to serve as a missionary in Africa’.Footnote 25 From a young age, he had been firmly gripped by the ‘Livingstone effect’Footnote 26 seeing the ‘explorer-cum-healer’Footnote 27 as a personal hero in whose footsteps he wished to follow. In Uganda, he enjoyed frequent trips into the bushveld, going to game parks for family holidays, regularly ‘hunting for the pot’ to augment the family’s food supplies, and eventually becoming an honorary game warden.Footnote 28 As a colonial missionary, Williams was afforded the vast ‘privilege and status [White men] enjoyed in the colonial context’,Footnote 29 a status that they could ‘not have hoped to achieve back home’.Footnote 30 This enabled him to operate with marked ‘organisational self-determination’ and ‘autonomy of space’, allowing him to transform Kuluva into his own ‘regional fiefdom’.Footnote 31
Constructing Kuluva was a monumental task that was heavily dependent on charitable donations raised through church and missionary networks. Every structure, including wards and surgical theatres, had to be ‘patched together with whatever [Williams] could get his hands on’.Footnote 32 Casting himself as a kind of frontiersman, Williams set out to acquire the skills needed to build and run a hospital, becoming an adept jack-of-all-trades. His mother had come from a family of builders and his father was a civil engineer, so he possessed some rudimentary knowledge of construction and architectural design. Building on this foundation, Williams ordered books from the UK to teach himself surveying, construction techniques, water systems, tiling, car mechanics, generator repairs, and basic electric work, among other things. Williams did not, however, build Kuluva single-handedly. Rather, he was dependent on cheap and plentiful African labour, which he mobilized through his dual position as missionary and colonial agent.
Like most medical missions, Kuluva operated as a ‘frontline service,’ supplying medical care in rural regions.Footnote 33 Yet Williams distinguished it from mission institutions that combined clinical care with religious education by describing it as a ‘wholly medical station’.Footnote 34 This description may have reflected Williams’ own perspective and priorities more than everyday practice, though, as Williams prayed before every operation and discussed the gospel with patients and their families.Footnote 35 Over time, the Williams family expanded Kuluva Hospital into a vast complex of 14 buildings that served much of the West Nile community.Footnote 36 Despite its rural setting, the medical services offered were fairly comprehensive. Not only did the hospital treat ‘everything from alligator bites to malaria’, but, for some time, it was one of the few hospitals in Uganda with a functioning X-Ray machine.Footnote 37
Although he often portrayed himself as single-handedly running the mission station, Williams, assisted by his doctor-brother Peter, relied heavily on the ‘skilled work and expertise of the many African[s]’ who formed an integral part of the labour force that worked on and maintained Kuluva Hospital.Footnote 38 Despite their critical role in knowledge production, these intermediaries are often eclipsed in ‘account[s] of (post)colonial science that cast heroic scientists and Western experts as drivers of knowledge’.Footnote 39 As is so often the case, ‘the historical record […] is largely silent’ on who these people were and how Williams understood his relationship to them within the ‘racial, paternalistic, [and] imperial dimensions’ of his missionary and medical work.Footnote 40
The records system
Soon after his arrival in Uganda, Williams became concerned with keeping track of his patients and felt that a proper records system was sorely needed.Footnote 41 Missionary doctors were known for their careful ‘attention to record keeping’.Footnote 42 Most famously, British missionary and founder of Mulago and Mengo hospitals, Albert Cook kept ‘meticulous case notes’Footnote 43 between 1897 and 1934, creating important ‘sedimentary layers’ of data that became foundational to later Ugandan cancer registration efforts.Footnote 44 As Megan Vaughan has noted, in colonial Africa, medical missions ‘played an important part in constructing “the African” as an object of knowledge’.Footnote 45 Missionary record-keeping not only tabulated disease but also imposed biomedical authority and social hierarchies, elaborating classification systems and practices around ‘tribal identity’ and land ownership, which were intrinsic to the operation of colonial power.
Indeed, the British colonial world within which Williams operated was a world ‘obsessed with paperwork’.Footnote 46 Numerous scholars have drawn attention to the textual worlds of empire, and how letters, reports, and records served as core constituents of how the British Empire functioned, dominated, and ruled.Footnote 47 In this context, accurate record-keeping was central to the operation of colonial governance. Documentary practices played an important practical and conceptual role in empire-building, translating people, land, and resources into categories legible to imperial systems of knowledge and control. Often cuttingly violent, these paper-based practices robbed indigenous inhabitants of language, culture, land, and rights, and encoded them into the papered epistemologies of empire.Footnote 48 At the same time, the power of these paper tools was not uncontested: East Africans frequently asserted their preferences, refused participation, and forced colonial officials to adapt their plans.Footnote 49 Even so, documentation remained central to the performance of colonial authority, projecting an image of order and control even when on the ground realities were far more complex. From the earliest imperial incursions, missionaries, like Williams, played a crucial role in the production of these paper worlds.Footnote 50 Shaped by the administrative logics of missionary and colonial work, Williams channelled this emphasis on documentation into creating a ‘proper’ records system at Kuluva Hospital.
Before instituting his own system, Williams visited numerous regional hospitals to see how they were keeping records. In many of these centres, patients were given a record card of their medical history, which they were expected to keep and look after. Many facilities, particularly those run by the government, would only keep patient records for a year before issuing new ones, regularly burning them to make space for fresh files. These systems, Williams felt, were ‘unsatisfactory’ as records often ‘got lost, burned, eaten by goats or otherwise destroyed’.Footnote 51 He therefore devised his own system, where patients would be given their hospital number on a metal tag, a material deemed robust enough to withstand the vagaries of goats and termites.
These numbers would correspond to cards held at the hospital where patient information was stored. The hospital team devised their own shorthand so that cards could be filled with as much information as possible, and no card was discarded. To help maintain this records system, Williams’ brother, Peter, ordered an Adana printing press from the UK and taught himself, and later his parents, how to print records cards on it.Footnote 52 Unlike in other systems, at Kuluva patients retained the same hospital number for life, allowing patients and their disease patterns to be tracked over time. ‘From this basic records system’, Williams tells us, ‘a series of registers and a daily journal of outpatient attendances were developed’.Footnote 53 While reasonably unique in clinical settings, this mode of monitoring was very much in line with contemporaneous colonial practices of scrutinising and surveying populations, across and beyond Africa. As numerous scholars have shown, the assignment of numbered identification tags or disks functioned as key tools of colonial governance, allowing administrators to track and regulate populations and delimit their access to welfare or aid.Footnote 54 Though anchored in colonial bureaucratic logics, Williams’s record system generated a rare continuity of patient data. His use of unique patient identifiers also served as an important foundation for his later cancer work.Footnote 55
In addition to maintaining a meticulous patient records system, Williams created a wide array of other paper materials based on Kuluva patient data, including maps, graphs, tables, and other forms of record making and keeping. Beginning in the 1950s, long before the advent of routine computer use, Williams did his collecting, calculating, and tabulating by hand. In Figure 1, for example, we can see Williams employ a simple tallying method, marking lines and then striking them through, to record hospital patient numbers by place of origin. In Figure 2, we see him graphing trends of patients coming to Mvara hospital by month and week.
Patients by region of origin, WTI/EHW/B/8, Wellcome Collection, London.
Comparing patient intake numbers across different years by time of year, WTI/EHW/B/8, Wellcome Collection.
As John Ziegler, an American collaborator of Williams’ who founded the Lymphoma Treatment Center in KampalaFootnote 56 noted, ‘rural hospitals are filled with patients all the time…They are busy places. To take the time out to meticulously record…patient[s] and the[ir] address[es] was a special effort and he was remarkable in the way he did that’.Footnote 57 The careful effort and missionary zeal with which Williams maintained patient data would come in handy later on, when growing interest in Ugandan cancer rates provided him with the opportunity to use his filing system as a basis for an operational cancer registry.
Creating cancer records
In 1961, Williams was invited by Denis Burkitt on a 10-week ‘tumour safari’ across east and southern Africa. Burkitt invited Williams, in part, due to their shared Christian devotion and love for David Livingstone.Footnote 58 However, Burkitt also wanted Williams to join because of his skills as a car mechanic and general handyman. Famously, it was on this ‘tumour safari’ that Burkitt gathered the evidence needed to show that incidences of Burkitt’s lymphoma corresponded to geographical and environmental factors, and thus were likely to have a viral aetiology.
Burkitt was at the beginning of an epidemiological journey that would catapult him and Ugandan cancer research onto the global stage. During the 1950s, several doctors working at Makerere University and Mulago Hospital had been intrigued by a strange yet disfiguring jaw tumour that appeared in numerous children. As a surgeon, Burkitt became interested in the tumour and started to look for clues that could shed light on its aetiology.Footnote 59 At the time, much cancer research fell under the banner of geographical pathology – a research tradition that gained prominence in the mid-twentieth century as imperial networks of researchers increasingly mobilised comparative analysis to assess whether, and to what extent, nutritional, social, economic, and environmental factors accounted for disease patterns observed across populations.Footnote 60 Working within this deeply colonially-informed framework, Burkitt mapped cases of the jaw tumour across east and southern Africa and theorised that there was a correlation between altitude, temperature, rainfall, and disease occurrence. His ‘tumour safari’ was his pièce de resistance. To test his hypothesis, he and his team undertook what he described as a ‘geographical biopsy’, travelling across east and southern Africa to plot the exact border between areas that had cases of Burkitt’s lymphoma and areas that did not.Footnote 61 From this work, Burkitt produced a number of maps which revealed a perceived ‘lymphoma belt’. Alexander Haddow, head of the East African Virus Research Institute (EAVRI)Footnote 62 and Makerere pathologist Jack N. P. Davies, both suggested that these geographic correlations indicated a very high likelihood of a viral origin – probably one with a mosquito or insect vector. Burkitt’s mapping work produced a burst of research interest, both locally and internationally, driven by the excitement of potentially finding the first clear evidence of a human cancer with a viral origin. The challenge was that while Burkitt’s ‘powerfully observant’Footnote 63 work demonstrated a strong correlation between climate and cancer, it did not offer definitive proof of or identify the specific virus involved. In light of this tantalising yet unsolved medical mystery, the hunt was on to isolate the virus responsible for BL.Footnote 64
Inspired by his trip with Burkitt, Williams, upon his return to Kuluva, began keeping detailed records of all cases of cancer diagnosed at the hospital. His patient records were in such good order that Williams could retrieve cases dating back to Kuluva’s 1951 opening, allowing him to compile a comprehensive cancer registry of over 800 patients.Footnote 65 The ‘keeping of adequate records’ and ‘quality [patient] notes’ was an important part of mid-twentieth-century cancer registration and a central tenet of geographic pathology.Footnote 66 However, at the time, it was relatively uncommon for countries to maintain population-based, or even hospital-based, cancer registries. In sub-Saharan Africa, only a handful existed: in Ibadan, Nigeria; Lorenço Marques, Mozambique; Johannesburg and the Eastern Cape, South Africa; Nairobi, Kenya; Bulawayo, Zimbabwe; and in Mulago, where Davies used Albert Cook’s historic data as a base to establish Sub-Saharan Africa’s first registry in 1951. The Kuluva cancer registry was among the few operating across the continent, fuelling Williams’s interest in cancer and cancer registration – an interest that would define much of his scientific career.
In 1966 Williams published one of his first scientific articles on ‘A local cancer registry in a mission hospital in Uganda’. Here, he outlined the value of rural cancer registration in enabling researchers to ‘embark on an exciting exploration of […] local cancer patterns’, as well as the exact procedure for registry operations. In his view, effective registry maintenance relied on a suite of paper technologies, including: 1. record cards with an outline of the map of the area; space for a biopsy report; space for photographs (see Figure 3)Footnote 67; 2. A simple index of cases for them to be found easily; 3. A filing system for the record cards; 4. Outline maps of the district to plot distributions of cancers on; and 5. Outline maps of the district where features such as rainfall or elevation can be shown.Footnote 68 Williams had spent some time working with Davies in the Kampala Cancer RegistryFootnote 69 and undoubtedly drew methodological inspiration from this experience. However, the central role given to maps suggests that Williams was also strongly influenced by geographical pathology and the more general atmosphere of ‘map consciousness’ and ‘map thinking’ prevalent amongst Ugandan researchers at the time.Footnote 70 For these researchers, maps were key ‘artifacts of geographic imagination [that]…translate[d] local observations into global knowledge’.Footnote 71
Kuluva Cancer Registry Form, WTI/EHW/J/3, Wellcome Collection.
Building on Burkitt’s work, and relying heavily on the tools and methods of geographical pathology, Williams began experimenting with different ways of collating and presenting cancer data on paper. In Figure 4 we see how Williams tracked West Nile District rainfall patterns. Rainfall was considered an important indicator for insect and mosquito activity and formed a key part of trying to prove an insect-borne vector for the BL tumour. In Figure 5, we see how Williams experimented with different ways of visualising population density and location in relation to Burkitt’s lymphoma, based on districts, population sizes, and disease incidence. Williams included shaded areas to show elevation, exhibiting how few cases of BL were found at higher elevations.
Rainfall in inches by year compared to cases of BL in Kuluva. WTI/EHW/B/8, Wellcome Collection.
West Nile divided by BL cases in squares, WTI/EHW/B/8, Wellcome Collection.
At Kuluva, Williams systematically documented all cases of BL. In addition to noting detailed demographic information, he, like Burkitt before him, plotted patients’ home locations using star stickers (see Figure 6).
Marking cases of BL on a map of West Nile using star stickers, WTI/EHW/G/1, Wellcome Collection.
In 1965, Williams ‘drew another map showing the distribution of cases using different colours to indicate year of occurrence’ (Figure 7). Here, he noticed an ‘easterly drift of cases’, writing: ‘in 1965 in the course of doodling on a map of the West Nile plotting the homes of BL patients I noticed a “grouping” which shifted by the year, and the shift appeared to be in an easterly and north-easterly direction.’Footnote 72
Showing the drift of cases over time, WTI/EHW/G/2, Wellcome Collection.
On Burkitt’s insistence, he showed this map to Richard Doll, a leading cancer researcher at the time. Intrigued by the map, Doll sent Medical Research Council statistician Malcolm Pike to go look at Williams’ data. Pike ‘checked the accuracy’ of Williams’ records and they proceeded to publish numerous papers on this observed ‘time-space clustering’ pattern.Footnote 73 These papers argued that Williams’ records showed groupings of BL infections occurring within similar time frames in similar locations, essentially a cluster of BL cases, some of which shifted with time across different areas (see Figure 8). This time-space clustering implied that there might be an external or environmental factor responsible for the sudden increase in incidence, lending support to theories of viral genesis. According to Williams, it was the discovery of this ‘time-space clustering [that] resulted in the West Nile being chosen for a large [IARC] project to be undertaken to determine the cause of Burkitt’s lymphoma’.Footnote 74
Showing Aliba case clustering, WTI/EHW/G/2, Wellcome Collection.
Ursula Klein has suggested that paper tools allow for creative possibilities of notation.Footnote 75 With Williams’ work we see these imaginative possibilities being explored – and how his mobilization of paper technologies brought together many different ways of looking at data. In making his important time-space clustering observation, Williams noted that he was ‘doodling’ on a map – a free-form practice of scientific thinking that rests heavily on the unboundedness of paper technologies. Christoph Hoffman and Barbara Whitman have suggested that writing and drawing constitute one of the most critical steps in epistemic processes – the step from (potentially) ambiguous data to stable facts, and from provisional ideas to guiding concepts.Footnote 76 Pen and paper, therefore, serve as key modes of securing epistemic realities. Here, we can see how Williams’ experimentation with different forms of written mediums, supported by his missionary background as well as colonial infrastructures and systems of administration, helped produce and stabilise scientific data. Because he was trying to do epidemiological, rather than laboratory work, Williams took note of a wide variety of things – and tried to find ways of bringing them in conversation with each other – all through the limits, but also possibilities, of the paper technologies available to him.
Burkitt’s lymphoma project
By the late 1960s, London Middlesex Hospital researchers Anthony Epstein, Yvonne Barr, and Bert Achong, in conjunction with Gertrude and Werner Henle at the Children’s Hospital in Philadelphia, had isolated a virus that seemed to be present in tissue samples from African children with Burkitt’s lymphoma.Footnote 77 This new strand of herpes virus was given the moniker Epstein-Barr Virus (EBV), now most commonly known as the key infectious agent in mononucleosis. The isolation of EBV was a momentous moment in the hunt for an oncogenic virus, ushering in a new wave of research into East Africa. As Burkitt remembers, ‘people zoomed in from all over the world because they saw the opportunity of being the first to find a human cancer virus, so there was intense competition’.Footnote 78
Epstein and the Henles worked closely with Guy Blaudin de Thé, head of IARC’s Unit of Biological Carcinogenesis.Footnote 79 In collaboration with a large number of researchers, including many based in East Africa, de Thé prepared a proposal for a large-scale multi-year project which would test for the presence of EBV in children before and after developing BL. Functioning as a prospective cohort study, the basic outline of the project was as follows: a demarcated area would be chosen as the project site. All children between the ages of 0 and 8 within the area would be ‘bled’. Each ‘bled’ case would be carefully marked with a serial number, with home location, age, and gender noted. The sera would be stored in Lyon, at the Henles’ lab in Philadelphia, or in other facilities in Europe. Whenever a child from this demarcated locality developed BL and was admitted to hospital, they would be bled again, and comparisons in EBV antibody levels and patterns would be made between their first sera batch and their second.Footnote 80
A project of this size was a huge logistical and scientific undertaking. The number of children involved and length of follow-up time required were enormous: 40,000 newborns and babies followed up for 8 to 10 years. The estimated cost of around one million US dollars per year was so high that de Thé was counselled to abandon the study.Footnote 81 However, financial support from a contract between the American National Cancer Institute’s Special Virus Cancer Program and Makerere University allowed planning to proceed. In 1968, a meeting was held in Nairobi to set out the parameters for the project. The initial proposal – to conduct field studies in Uganda and Kenya – was deemed too costly. However, there was enough support to enable George Kafuko, the first Ugandan Director of the EAVRI, to carry out a pilot project to establish the feasibility of a larger study.Footnote 82 Williams was invited to a follow-up conference in Entebbe in January 1970. Not well versed in molecular medicine, he felt ‘a bit lost’,Footnote 83 however, he soon realized ‘that the discussion revolved around whether the project would be in Ghana, Tanzania or West Nile’. To his ‘astonishment West Nile was selected because [his] records of where BL cases lived were the most detailed and accurate’.Footnote 84 Williams’ records showed that the West Nile District had the highest incidence of BL in Uganda – a crucial factor for a cohort study reliant on a certain proportion of participants going on to eventually develop the disease. His files also contained detailed notes on the home and village locations of all tumour cases Williams had treated over the course of his career.Footnote 85 This spatial knowledge was essential to a project predicated on re-locating and re-bleeding cohort members who had donated their blood years earlier. Consequently, the West Nile region was chosen as the location for the project and Williams was brought on as medical consultant, working under the direction of George Kafuko, who served as the project’s inaugural principal investigator.Footnote 86
Accuracy and meticulousness have long been central to scientific knowledge making.Footnote 87 Scholars have noted that, over time, the ideal of precision in science has increasingly privileged painstaking, methodical labour over once-celebrated flashes of genius.Footnote 88 In this IARC study, a similar reverence for rigour emerged – one that tied the production of reliable scientific data to the disciplined, industrious persona of its recorder. Williams’ records not only signified scientific reliability, but his methodical disposition, bureaucratic diligence, and positionality as a colonial missionary and locally embedded authority also served as important project assets.
As a missionary, Williams had an ‘extended presence in the region’. This not only allowed for undisrupted record-keeping, but it also allowed him to ‘communicate with patients in a number of languages’,Footnote 89 offering a degree of integration ‘(as much as Europeans could claim) into the local community’.Footnote 90 Having lived in a colonial Kenyan household until he was fourteen, Williams spoke some Swahili and learned the local language of Lugbara at Kuluva. His long-term residence in the area also meant that he had an awareness of community norms and practices and had a lot of trust from the local population.Footnote 91 Long-term relationships of trust and locally situated knowledge were crucial to the production of cancer data in mid-twentieth-century Uganda,Footnote 92 and specifically important for the West Nile project, where ‘the ability [of local researchers] to translate knowledge of the social, biological, and physical environment of the West Nile into cartographic representations amenable to statistical analysis’ was critical to the project’s success.Footnote 93 In initial planning, Kafuko and Williams used their local knowledge to modify several proposed protocols. For instance, IARC initially proposed that the project should have four field teams, but Williams persuaded the group that a fifth team was necessary to offset time lost to staff illness and leave. Kafuko also made several important modifications to team composition, suggesting that the teams should also contain, in addition to a medical officer, a medical assistant, nurses, and trained medical staff to assist with bleeding and treatment.Footnote 94 The researchers’ situatedness, cultural sensitivity, language skills, and long-term residence in the area not only enabled systematic record-keeping but were also crucial in the collection and interpretation of new data.
A key example of this is when Williams learned that ‘experts’ wanted to take twenty millilitres of blood from each child enrolled in the study.Footnote 95 Aware of the local context, he argued that it would be impossible to persuade parents to give so much of their children’s blood and the project would ‘only succeed if they took just 3mls’.Footnote 96 For Williams, knowing and being attuned to local contexts was an important part of research and record creation, to the point that he was ‘adamant that the pathological, virological, and statistical studies could only be conducted and given meaning with a proper appreciation of the culture of the communities being studied’.Footnote 97 Williams was so embedded in the West Nile District that ‘he saw himself as more familiar with local ways than even the Entebbe-based Ugandans [at the EAVRI] with whom he collaborated’.Footnote 98 In a ‘remarkable’ instance in which a ‘British missionary doctor’ explained the intricate complexities of local cultural norms to a ‘Ugandan scientist’, Williams wrote a chastising letter to Kafuko about the unsatisfactory nature of the collection of blood samples in villages.Footnote 99 Williams was concerned that staff members collecting blood were ‘either not conscious of the traditional feelings involved or merely dismiss[ed] them as primitive’. To avoid the collection teams being ‘regarded with great suspicion by the local people who impute the worst possible motives on those [collecting blood]’, Williams set out six demands on how the team should go about the sample collection process, including that ‘no “white” man should be permitted to take blood’ and that there should be ‘no haste’ in sample collection.Footnote 100
Through this episode, we see Williams’ pedantic and sometimes controlling personality emerge, as well as the extent to which colonial hierarchies and contextualised research practice informed the types of data he produced. As one of his ex-colleagues recalled, ‘he sort of ran his own ship there with BL, we tried to get him involved in protocols, but he was very adamant to do it his way’.Footnote 101 While this, at times, led to the creation of more accurate data, it also sometimes compromised the quality of data produced. Williams’ insistence that only 3 milliliters of blood could be taken from cohort members was a consistent headache for laboratory workers, with many arguing that there was not enough blood for them to carry out important medical tests. As Werner Henle complained to de Thé in 1974: ‘Because of the small volumes and the extreme value of the pre-illness sera, it would be wise to disregard “statistical niceties” and test the sera with utmost care […] basically, this is not the way to run important tests, but we did the best we could’.Footnote 102 Here we see how the socio-political context of medical research influences not only the scientific methods employed but also the kinds of knowledge that emerge from it.
IARCs’ West Nile District BL project was formally launched in 1971. It was a massive undertaking – one of the ‘biggest sero-epidemiological project[s] ever done […] especially in a tropical region’.Footnote 103 An IARC/EAVRI field-station was set up in Arua under the leadership of Indian doctor and medical officer Dharm Beri. The ‘fieldwork involved…[was] considerable’,Footnote 104 employing sixty-five people in the West Nile District, including supervisory staff, laboratory technicians, and members of six field teams, each composed of a medical assistant, two nursing assistants, two registration clerks, and a malaria scout.Footnote 105 These workers, who were primarily Ugandan, led community information meetings, enrolled families and children into the project, obtained blood samples, and collected vital identifying information.Footnote 106 They also kept detailed records of environmental conditions – housing, dietary habits, food production, hygienic practices, and climate. The day after collection, sera was separated in the Arua field laboratory and stored at –20°C. Each month, batches of blood would be sent to IARC, and other facilities, via the EAVRI. The records, encoded on microfilm, would also be sent to the Agency and stored on their central computer for subsequent analysis.Footnote 107
As can be seen in Figure 9, much of the crucial epidemiological and scientific work for this project was preformed by Ugandan laboratory personnel and fieldworkers. Despite this, the contributions and labour of these ‘invisible technicians’ have gone largely unacknowledged – a pattern emblematic of the broader exclusion of African researchers from international recognition and histories of scientific discovery, more generally.Footnote 108
Ugandan laboratory personnel separating blood samples for serum collection at the project laboratories, Arua, Uganda. IARC, ‘IARC Annual Report 1975’ (Geneva, 1976): 60.
Unlike the Ugandan laboratory technicians and fieldworkers responsible for taking and freezing blood samples, Williams was stationed at Kuluva, where he continued to manage the day-to-day operations of his hospital. As scientific advisor, his role in the study was to persuade ‘local parents in a selected area to allow blood specimens to be taken from their smaller children’ and deal with any BL cases that materialized, taking ‘the necessary biopsies, blood specimens etc from these patients and give them treatment’.Footnote 109 Ideally, every child who developed BL in the study area was to be sent to Kuluva where Williams would re-bleed them, take biopsies, and record their demographic data. Here too, fastidiousness was a defining feature of his work. Throughout the study, Williams kept a close eye on how many BL cases were observed within each county each year. He also obsessed over duplicate entries and misplaced sera – errors that eventually inspired him to implement a blood serum labelling procedure, designed to minimise confusion when sorting out re-bled sera from BL case families.Footnote 110 According to collaborator Peter Smith, this ‘rigorous numbering system’ served as a sort of barcode, allowing researchers to identify particular samples through a series of written numbers on tubes.Footnote 111
Part of the BL work was also focused on trying to discover a chemotherapeutic cure for the tumour, spearheaded by Kampala’s newly built Lymphoma Treatment Centre (LTC).Footnote 112 At Kuluva, Williams engaged in experimental regimes of administering chemotherapy drugs of various types in different combinations to BL patients. He made careful notes of the drug cocktails that patients were receiving. His treatment notebooks contain complex tabulations of various amounts of different drugs given, related to survival factors. The longevity of each child, often brief, was also carefully noted and related to their treatment protocols, a painful linear representation of BL’s swiftness and deadliness and the experimental nature of Williams’ work. Even though, as Mika points out, in later years many children at the LTC ‘actually got better and stayed better’Footnote 113 very few of Williams’ patients survived beyond a few months. Experimentation on subjugated bodies was part of the biomedical violence of colonial regimes.Footnote 114 While Williams’ aim was to save lives, the lack of documentation around consent or patient experience and suffering suggests his work was not excluded from this.
Ultimately, the West Nile project obtained sera from 42,000 children aged 0 to 8 years, between 1972 and 1974.Footnote 115 Children suspected to have BL were subjected to further tests (including more blood drawing, numerous biopsy cuts, malaria testing) and experimental treatment regimes.Footnote 116 The project thus turned the people of the West Nile District into a hyper-researched population and ‘living laboratory’Footnote 117 for biomedicine that was ‘tethered to colonial experimentation and the extraction of bodily materials and knowledge’.Footnote 118
Epidemiological reasoning in a time of transitions
Williams was working at the cusp of computing, where complex mathematical formulae, graphs, and statistics were already being mobilised in epidemiological work. As scholars have noted, while before the 1960s essentially all public health observations were made by manually inscribing data onto paper, during the 1960s and 1970s, statistical calculations were increasingly carried out electronically, with paper records transferred onto computer files, where they could be stored and later analysed.Footnote 119
IARC had a computer, stationed at its central office in Lyon, France. However, sitting at a hospital he himself had built in Uganda’s Northern Province, Williams did not have access to digital technology. As Ziegler recalled, ‘there were no computers, no cell phones, no internet. We used IBM punch cards to do counting and the tools of epidemiology that were available…we had cases and we had maps and that was about it’.Footnote 120 Some years later, Williams noted in a letter, ‘it is of course incredible that when I started my registry, micro-computers had not been heard of, and yet now they are commonplace and make analyses of this sort very easy’.Footnote 121 Without a computer, and yet still wanting to systematically collect data and undertake thorough analyses of it, Williams had to collect, collate, and compute all his epidemiological data by hand.
Paper is a conspicuous tool for mathematical practice – a fact to which Williams was not oblivious.Footnote 122 As Alan Turing wrote in 1948, ‘it is possible to produce the effect of a computing machine by writing down a set of rules of procedure and asking a man to carry them out. Such a combination of a man with written instructions will be called a “Paper Machine”’.Footnote 123 Williams can be seen as a ‘paper machine’, spending inordinate amounts of time creating and collating his BL records, and subjecting them to manual computations. As Williams notes in his autobiography: ‘every spare moment and many evenings I was engaged in my cancer research, trying out new ideas and investigations’.Footnote 124 Williams’s role as a ‘paper machine’ may have been influenced by family ties to colonial administration – his father, served as ‘Chief Computer’ in the British government’s Survey Department in Kenya until 1929. In Figure 10, we see Williams making his own graphs to examine trends in disease occurrence amongst Kuluva patients. Figures 11 and 12 likewise show Williams trying to establish correlations between in-patient numbers and cancer incidence through regression analysis.
Disease case numbers by 5-year period, WTI/EHW/F/2, Wellcome Collection.
Regression analysis on patient data, WTI/EHW/B/2/15, Wellcome Collection.
Regression analysis comparing in-patient numbers to cancer cases, WTI/EHW/B/2/15, Wellcome Collection.
Williams’ cancer research took place during a period of many transitions. Not only was he working at a time when epidemiology was moving from pen and paper to computer technologies, but he was also witnessing a shift in epidemiological forms of reasoning. As outlined above, much of Burkitt and Williams’ work was informed by geographical pathology. In the 1970s, however, geographical pathology’s popularity was waning as cancer research was increasingly being pushed ‘towards specific forms of molecularization’.Footnote 125 The Henles in the US, and de Thé at IARC, were part of this shift. In their correspondence with Williams, they consistently emphasised the importance of laboratory studies and tissue sampling, frequently describing the work they were doing as ‘sero-epidemiology’ – an approach to epidemiology based on the laboratory analysis of materials (often collected in the field).
As Joanna Radin has noted, the rise of sero-epidemiology, which emerged in tandem with postcolonial hierarchies of human development and anxieties about disappearing biological diversity, was associated with a new brand of meticulousness. Molecular epidemiological studies, based on practices of accumulating and analysing blood, were often seen ‘as the gold standard for achieving epidemiological intelligence’.Footnote 126 Sero-epidemiology was often described as being ‘modern’, ‘integrated’, and ‘logical’ and was thought to endow epidemiology with a new form of methodological rigour, as ‘collections of standardized tissue samples came to gradually supplant previous systems of aggregating observational reports of patterns of morbidity and mortality’.Footnote 127
Unlike geographical pathology, which relied on paper-based practices, particularly the drawing of maps, sero-epidemiology focused more on microscopic, molecular, and lab-based equipment. In a context where these technologies were expensive and hard to come by, researchers like Burkitt and Williams remained committed to paper technologies, and continued to use these, and other geographical pathology practices, to solve aetiological puzzles. We see a clear example of this in a 1970 paper published by Kafuko and Burkitt. Building off the earlier work of Gilbert Dalldorf,Footnote 128 Kafuko and Burkitt used the paper-bound methods of geographical pathology, such as map notation and cross-national comparison, to argue that holoendemic malaria was a very likely co-factor to EBV in the emergence of Burkitt’s lymphoma. Despite decades of continued research into the disease, this original pen-and-paper-based theory continues to be upheld as the most likely cause of the high rates of BL in Africa to this day.
In the work of Williams and Burkitt, we can see how certain paper technologies allowed for a particular type of epidemiological reasoning – one based on the enduring legacies of empire, knowing contexts intimately, and doodling and drawing things on maps. These technologies facilitated a unique epidemiological gaze that helped researchers ‘see’ correlations that perhaps those more embedded in serological studies were not as attuned to. We see this from Burkitt’s first hunch that he had discovered a viral cancer and later malaria co-factor hypothesis, to Williams’ identification of a time-space clustering phenomenon taking place in the West Nile.
Williams’ cancer records were not only produced in a time of transitions in epidemiological reasoning and technologies. He was also working during a period of social and political upheaval in the wider context of Uganda. In 1962, Uganda gained independence from Britain. While this was a major change that had a profound impact on British government employees, like Burkitt, most of whom had to leave, its effect on missionaries like Williams was less noticeable. As Williams acknowledges in his writings, the far more seismic shift came in the 1970s. In 1971, just as IARC’s West Nile project was taking off, Idi Amin overthrew Milton Obote’s government in a coup that was to usher in a new epoch in Uganda’s history. While Amin came to power on the promise that he would put an end to the corruption that had crippled Obote’s government, his regime soon grew to be ever more dictatorial and despotic. One of the early indications of the violent nature of Amin’s rule came in 1972, when he expelled Uganda’s South Asian minority from the country, giving South Asian residents 90 days to leave.Footnote 129 This resulted in a severe loss of medical personnel, with the number of registered doctors in Uganda decreasing from around 1200 to 600.Footnote 130
Amin repeatedly threatened to expel European, and particularly British, inhabitants of Uganda, creating an atmosphere of uncertainty and anxiety among expatriates. As Williams himself notes: ‘once a year at least there would be a threat of removing the British from the country…it was becoming increasingly difficult under the Amin regime to carry on our work with any degree of equanimity’.Footnote 131 By the late 1970s, most foreigners had left. IARC’s West Nile study was also affected by the power transition. Although he managed to stay on longer than most in the wake of Amin’s South Asian expulsion, Dham Beri, the medical official charged with overseeing IARC’s field station in Arua, eventually left in 1977 (only to tragically die in a car accident in India a year later).Footnote 132 Peter Tukei, who succeeded Kafuko as head of the EAVRI and PI of the West Nile study in 1973, fled to Kenya in 1977.Footnote 133 Other project staff, including epidemiologist Peter Smith, found themselves frequently stopped and, at times, harassed by military personnel on the journey to and from Arua.Footnote 134 Nonetheless, despite the fear and insecurity the emergent political situation brought, the project continued to run for several years – in no small part due to the constant manoeuvring of people like Kafuko, Tukei, and Williams. It likely helped that Amin was born in the West Nile district and had Kakwa heritage, a group very well represented in the region. Williams and his work also seem to have been known to Amin. Indeed, when in 1978 Williams fell and fractured his hip and had to be flown to Kenya for medical care, Amin ordered a special dispensation to allow Williams’ aircraft to fly out of the country.Footnote 135
With time, however, the ‘near disintegration of the state’Footnote 136 and overall deterioration of basic infrastructure such as electricity, water, and transportation during the Amin period made it ‘impossible’Footnote 137 to carry out scientific research and medical work – issues compounded by the near impossibility of procuring even basic medical and laboratory supplies.Footnote 138 By 1975, Werner Henle felt that ‘the West Nile study has provided the answer at least to the question we originally asked’,Footnote 139 which was whether the presence of EBV virus could be determined in all BL cases and directly implicated as a key factor in the causation of those tumours. This, coupled with increasing political instability, meant that in 1979 IARC’s Uganda-based BL team ended their projects and withdrew from the country.Footnote 140 Ugandan physician George Olwit, who joined the Arua team as a medical officer in 1978 and was one of the last EAVRI researchers to leave Arua, called it a ‘get away operation’.Footnote 141
As the situation deteriorated, Williams increasingly felt that with the ‘uncertainty of the security situation’ and the escalating difficultly of running a rural hospital, he simply ‘could not carry on’.Footnote 142 He saw his hip injury and the resultant reduction in his mobility as the final sign that it was time for him to leave, and in February 1979, he and his wife departed Uganda for good. This international exodus gave rise to a landscape of crumbling foreign scientific infrastructure, discarded gear, and deserted piles of paper stacked high on dusty desks and workbenches. One of George Olwit’s last memories of the Arua field station was seeing the road outside covered with what looked like snow.Footnote 143 The ‘snow’ was shredded documents from the project – material remnants of a vast paper archive that had been used to record clinical observations, label antibody titers, count cases of cancer, and keep up a vast correspondence with researchers from the United States, Europe, and around the world.
Historians often point to the vulnerability and impermanence of paper. Rohan Deb Roy has shown how factors such as humid climates, tropical rains, and insect infestations could bring colonial administrations to their knees through damaging or destroying paper currencies, judicial records, and other crucial nodes of colonial state-making.Footnote 144 However, for all its fallibilities, paper can also be incredibly durable, maintaining a presence that can, when properly stored, persist across centuries and geographic borders. In this way, paper records occupy a space of duality – hovering between fragility and stability, perishability and permanence. When Williams left Uganda, he took his records with him, an act reflecting colonial assumptions about documentary ownership and the stewardship of scientific knowledge.Footnote 145 Fearing they would not survive Amin’s regime or the vagaries of a tropical climate, he packed his papers into ‘one small tin-trunk, a couple of suitcases and desk drawers’, which were transported from Kuluva Hospital to his new home in the UK. It is within these repatriated trunks, suitcases, and filing cabinets that Williams’ printed and handwritten texts took on new life, continuing to serve as valuable epidemiological resources well into the twenty-first century.
The meticulous archive
Williams’ records fuelled key discoveries about the relationship between cancer and viral infections. However, the relevance of and for his data endured long after IARC’s sero-epidemiological study of BL patients. Texts have ‘afterlives’; they travel and are read and interpreted in diverse contexts, far from those in which they were originally produced.Footnote 146 As Yonatan Gez has noted, ‘the remarkable persistence of the past…cannot be dissociated from materiality’, and the materiality of Williams’ records has continued to shape research structures and sensibilities, exerting a tangible influence on scientific practice.Footnote 147 In this section we trace the many ‘afterlives’ of Williams’ records, as his files continued to be used in various arenas: in the domain of global cancer registration, for additional etiological studies on the relationship between cancer and viral infections, and for historical research.
Williams’ cancer data witnessed a resurgence of interest in the early 1980s. When Williams left Kuluva he took 30,000 patient case-files with him for ‘safe keeping’ – a resource that comprised a valuable record of historical cancer knowledge.Footnote 148 In compiling Cancer Occurrence in Developing Countries, a monograph which sought to create a catalogue of cancer data from regions historically excluded from Agency publications, IARC officials reached out to Williams, asking if he had any data to contribute.Footnote 149 Upon receiving the files, IARC officials thanked Williams for ‘the care and thoroughness which has gone into compiling this material’.Footnote 150 His records made such a strong impression that IARC would continue to contact him throughout the 1980s. Significantly, in 1984, Guy de Thé wrote to Williams asking if he had seen anything resembling AIDS whilst working at Kuluva.Footnote 151 In response, Williams noted that he had observed the presence of various aggressive forms of Kaposi’s sarcoma – a purple skin lesion and opportunistic infection commonly associated with HIV patients. This observation spawned a new wave of sero-epidemiologcial studies in Uganda, pointing to yet another example of Williams’ records having relevance far beyond his BL work.
In 1986, Sue Barmley, an archivist at London’s Wellcome Library, wrote to Williams, asking if he would consider donating the files that he had ‘brought back from Uganda’, as she believed that his papers formed ‘an extremely valuable and rare record of medical work’.Footnote 152 Williams’ records ultimately did end up at the Wellcome. Here, his self-curated body of work is preserved in perpetuity, where it continues to be revisited, redeployed, and reactivated by researchers, including the authors of this very paper. In fact, it was the neatly constructed tallies of cancer case counts that originally attracted the lead author’s attention, prompting her to look at Williams’ records in more detail and try to unravel the complicated relationship between these precise records and the person who crafted them. What emerges from these materials is not only a record of medical practice but also a representation of how Williams wished to be seen – as a meticulous and methodical practitioner whose identity was deeply rooted in the ethos of missionary medicine – an image not only reflected in his medical records, but also in his diary entries and unpublished autobiography. These materials reveal a ‘deep intertwinement between space and self’ that not only underpins Williams’s own work, but also permeates the research of the many cancer scholars who followed – accruing layered, increasingly subjective meanings and carrying forward the colonial legacies and unspoken socio-political tensions embedded within them.Footnote 153
Conclusion
In recent years, most historical and social science investigations of global health quantification (particularly in Africa) have highlighted the messiness of data by emphasising the social lives of numbers and denaturalizing assumptions that health metrics are neutral, objective, or ‘clean’.Footnote 154 However, looking at how researchers try to enact order and neatness also yields important insights into modes of epidemiological reasoning. Williams’ work on BL shows the power, value, and politics of epidemiological insights generated through the triangulation of persona, paper, and place in mid-twentieth-century East Africa.
In unpacking Williams’ records, this paper highlights how precision, accuracy, and meticulousness served as key attributes for the quality and verity of scientific knowledge production – particularly at a time where the accuracy and speed of computer-based information processing systems were beginning to outpace the methodical process of recording medical knowledge on paper. With Williams we see how the continued utilisation of paper technologies in colonial and postcolonial contexts allowed for creative thinking, doodling, and experimentation, things that helped him, and those around him, develop new epidemiological ideas. Today, we increasingly view paper as a slowly disappearing and obsolete medium, rarely thinking about the important epistemic role it has played in public and population health. Williams’ work demonstrates the strength of paper, both in terms of advancing some of the most sigificnant oncological work of the mid-twentieth century, and enduring with striking persistence, despite ongoing environmental forces and socio-political threats.
More than this, we see the close relationship between personhood and data production. Williams’ methods and the data he produced were intimately linked to his persona and position as medical missionary, without which the data could not have existed. As Datson and Galison have noted, it is important to acknowledge ‘the integral involvement of the scientific self in the process of knowing’.Footnote 155 Data does not exist without a subject that produces it – there is no data that is not shaped by people and the political worlds they inhabit. The person of Williams was crucial to the production of his data – his colonial background, his pedantic, meticulous nature, his insistence that records must be kept, his fascination with data analysis, and his incessant efforts to develop accurate and sustainable methods of records collection and conservation, were all key to the production of his West Nile data. As M. Norton Wise has argued about all scientific knowledge, ‘reliability is never a matter simply of a reliable instrument [it] is also a matter of people judged to be reliable using methods that display their reliability’.Footnote 156
In this article, we have used the extensive paper records left by Williams to explore how scientific subjectivities are crucial in scientific data production, and show how paper and person come together to shape epidemiological imaginaries, even in domains where digitization and the rise of computational epidemiology are thought to have rendered older, more analogue forms of data collection obsolete. Equally, we have demonstrated how these practices and personas are inseparable from the political and social context within which they find themselves. Williams’ work was embedded within the dynamics of late colonial and early independence East Africa. It was deeply dependent on colonial hierarchies, exploitation, and inequities that allowed ‘heroic’ medical personas to emerge and for far-reaching, unchallenged experimentation to occur. While these dynamics continued to play out in the early independence context, we see how confronting an ever-increasing authoritarian regime that was not friendly to outside interference also shaped the trajectory of both Williams’ career and IARC’s broader project. Through a careful exploration of this singular moment of epidemiological knowledge production in a small part of Africa, this paper has sought to follow the ‘constitutive entanglement(s)’ of the social and the material revealing the connections ‘between sites of knowledge making, sets of knowers and different kinds of epismteic work’.Footnote 157 In doing so, we underscore that it is ‘neither possible nor fruitful to disentangle knowledge making from its ambient cultures’,Footnote 158 from its structuring technologies and practices and from the people who produce it.
Acknowledgements
The authors would like to thank David Reubi for his consistent support, reading, and advice throughout the writing process, as well as participants from the Epidemy Lab Data and Disease in Historical Perspective workshop and King’s College London’s Culture, Medicine and Power research group who shared their insights on earlier draft versions. We are grateful to the Special Issue editors, John Nott and Lukas Engelmann, for their careful reading of the paper and continued enthusiasm throughout the process. We would also like to thank our anonymous reviewers as well as Medical History for awarding the paper the 2024 William Bynum Prize. The research on which this paper is based on was supported by the Wellcome Trust.
Competing interests
No competing interests.
Open access
