BACKGROUND

DIABETES CLINICS AND EXPECTATIONS OF ‘COMPLIANCE’

Diagnosis as ‘diabetic’ (the named condition) is the province of biomedicine and the clinic. But clinical diagnosis and treatment only scratches the surface of the problem, in various ways which we hope will become apparent. Diabetes – both the word and the medical condition – remains unfamiliar to many in Cameroon and commonly goes unrecognized. Most diabetes is therefore undiagnosed. Nevertheless, as the urgency of tackling diabetes has grown, so has public recognition of both the term itself and the health problems it signifies. To the extent that it is recognized, popular understanding tends to regard diabetes as a new or ‘modern’ disease, an assumption which clinic staff in both settings reinforced in their own explanations to patients. But that ‘modern’ label conceals uncertainty as to whether the disease itself is a modern occurrence, or only recognized through modern medicine, or a symptom of modern living (cf. Ferzacca 2001).Footnote ⁵ Against this background, newly diagnosed patients could draw on little by way of common knowledge of what diabetes was and what it entailed, a problem compounded by a new and unfamiliar terminology.

Unsurprisingly, the term ‘diabetes’ does not straightforwardly match with the lexicon of ill-health in Cameroon's main indigenous languages (cf. Pool 1994a, 1994b). To illustrate this through the more linguistically straightforward case of Bafut, two words, the indigenous fumbgwuang and English shugar, are most frequently used, often prefixed with nighoni (sickness, disease). Nighoni-shugar thus denotes ‘sugar diseases’ and nighoni-fumbgwuang ‘disease that is sweet’. Yet, fumbgwuang also refers to salt, indicating a taste that goes wider than the sweetness associated with sugar. The terms given by people reflected how the language of ‘high sugar level’ (hyperglycemia) in the clinic has found its way into popular idioms of ‘sweetness’. In both settings, many patients acknowledged that ‘diabetes was caused by eating a lot of sweet things’, identifying the significance of the food they ate and what they drank, and the changes in eating habits which have occurred over the last generation or so (De-Graft Aikins 2005). As awareness of the significance of the disease has grown, one relatively common form of self-diagnosis has been by tasting one's own urine for the taste of fumbgwuang.Footnote ⁶ Indeed, individuals who were already being treated for diabetes sometimes used this method to control their blood glucose levels, especially when they could not or would not do so in a clinic setting. Urine offered another clue also, for the swift arrival of ants where a person has urinated was also taken as a possible indication, as it was well known that ants were attracted by deposits of sugar or salt.

Attending a clinic for diabetes introduced a series of expectations and assumptions which formed the basis for future misunderstandings: notably around the nature of the disease, its complications, and the kind of diet that was advised or discouraged. When a patient first attended one of the clinics chosen for this study, there began a process not only of tests but of education and instruction also. While the former was anticipated, the latter typically proved harder to accept. For the fundamental ‘message’ clinics sought to convey was that the symptoms labelled diabetes indicated a medical condition which could not be cured, and instead required careful monitoring and management for life. Alongside that, patients were made aware that diabetes had many associated problems: it increased the risk of high blood pressure, heart disease and stroke; it reduced libido; and without careful management kidney failure and blindness were possible complications. Furthermore, patients learned that diabetes aetiology combined both inherited and behavioural factors, in which what they ate and drank would have played a major contributory part, and that it was through regulating their diet in future that they had the best prospect of managing their diabetes successfully.

This process of instruction typically unfolded through successive consultations with clinic staff, usually nurses. The Etoug-Ebe Baptist Health Centre in Biyem-Assi was unusual in its more explicit procedure for underlining advice and encouraging group participation in question and discussion sessions. Both the Christian message and the primacy of compliance with clinic guidance were more forcefully stressed there than in the other clinics studied, and something of the flavour of the educational message provided is conveyed in the following extract from one of these talks by a senior nurse:

Many attending this clinic did not stay for these talks and advice sessions, and senior staff realized that, even among those who did, such information went over the heads of many. But in all of the clinics, even where the educational instruction was not so collectively organized nor so precisely articulated, an explicit task was seen as how to underline to patients that their ‘compliance’ was essential, if symptoms were to be controlled and complications avoided.Footnote ⁷ The importance of compliance thus recurred constantly in discussion.Footnote ⁸ What was not clarified so clearly, but would emerge over time, was that ‘compliance’ entailed several levels of conformity and self-discipline. First, it meant acceptance of a timetable of regular appointments at the clinic (though the Etoug-Ebe Health Centre was alone in emphasizing a consciously disciplining style by fining patients who arrived late). These enabled the regime of monitoring of blood sugar level, blood pressure and weight to take place. Second, it meant acceptance of the need for self-monitoring of diet and changes to food and alcohol consumption. And third, perhaps above all, it meant acceptance that the clinic itself, and the biomedical approach, should remain the sole source of treatment, with the understanding that this was a chronic and incurable condition. The need to adhere to a biomedical regime and no other, to ‘modern’ and not ‘traditional’ medicine, was a constant and explicit clinic refrain.Footnote ⁹ How complicated the seemingly simple notion of ‘compliance’ would prove to be was something which was rarely apparent to patients immediately, as will emerge in the next section.

INTERPRETING INSTRUCTIONS: PATIENT AND FAMILY AMBIVALENCE

The most immediate trigger for accusations of non-compliance concerned apparent failures by patients to follow guidance about necessary changes to eating and drinking. This could be willingly acknowledged by those with diabetes in the common admission, ‘I'm cheating the doctor.’ But while alcohol consumption was a particular source of clinic frustration, patients themselves were often genuinely confused about what was or was not safe to eat or drink. An understanding that diabetes involved excessive levels of sugar in the blood often encouraged those with the condition or their families deliberately to cut out what they saw as sugary food. Many knew also that beer contained sugar, and thus, like soft drinks, ought to be avoided. However, certain other alcoholic drinks were commonly seen as bitter, and the more bitter the taste the more they were construed as ‘safe’. In particular, the local palm wine was seen as bitter and therefore ‘safe’: as one Bafut man put it, ‘We are told not to drink this palm wine, but this palm wine is medicine to us.’ Guinness likewise was bitter and seen as ‘safe’. Further, because Guinness Cameroon promoted its value in building strength, Guinness was often seen as virtually medicinal, a means to regain the strength to combat both diabetes and lost libido (not always recognized as associated), and thus doubly effective. Bitterness thus stood for good health for those with diabetes, in contrast with the sweetness that signified ill-health. What better way to neutralize excess sugar in the body than the bitterness of palm wine or Guinness? These arguments were put forward by both men and women, but more commonly by the former, and one, a patient in Biyem-Assi, put it like this: ‘When I come to this bar, I drink my Guinness and it makes me to feel strong. I have been told that it is good for diabetes patients because it is bitter’ (Biyem-Assi, 22 January 2002).

Because modification to habits of eating and drinking was emphasized by clinics as so vital, adjusting to the onset and diagnosis of the condition was more than a matter solely for the individual concerned. Household eating habits were likely to be affected, and given the social significance of commensality, that was no small matter. Some households saw the adjustments necessary as being relatively minor and easily made: ‘When we want to cook food at home we cook her soup differently and ours differently … at times, we cook food in the same pot, remove hers, then add oil and salt in ours’ (family member, Biyem-Assi, 28 July 2001). But it was not always straightforward, and family tensions around dietary changes were distinctly gendered. Men with diabetes were more likely to resist the modifications their wives or sometimes children sought to introduce; while women with diabetes were more likely to feel that their requests for the consideration of special dietary needs were disregarded, with the response that there was insufficient money. The contrast highlights the gendered nature of ‘compliance’, but also the wider financial burden associated with such conformity (cf. De-Graft Aikins 2005). Men's command over household financial resources guaranteed diabetic men priority access to drugs, treatment and special food. A woman with the condition typically had a much harder task in asserting an equivalent claim on household finances. And even where she had support from her family initially, sustaining it over the long term was another matter. Sources of tension and misunderstanding abounded, therefore, not only between clinic and patients, but quite often between the latter and other members of their families.

Some of the commonest misgivings patients had about clinic guidance on diet concerned the desirability of weight loss and the significance of blood pressure. They and their families struggled to come to terms with the idea that diabetes patients should lose weight, when this was simultaneously seen as a kind of diminution or disfigurement. These concerns were compounded because of HIV/AIDS (De-Graft Aikins 2005).Footnote ¹⁰ For dramatic weight loss now has particular connotations, accentuating resistance to it, for no one wants to be thought to be losing weight because of AIDS. Losing weight was equated with ‘growing dry’, shrivelling up, the sexual connotations of such an eventuality being implicit in the larger misfortune. In the words of one man with diabetes, ‘I feel that I am a man when I am fat. When you start losing weight, people think that you have AIDS.’ A family member echoed this view, even to the extent of asserting that ‘It is good he should die resembling a man than dying looking like somebody with AIDS.’Footnote ¹¹ Blood, its strength and its pressure produced further mutual misunderstandings. Clinic monitoring entailed the routine checking of blood pressure, as we have seen, with constant advice about the dangers of high blood pressure. Yet the clinic's language of high and low blood pressure easily got elided with popular understandings about ‘strong’ and ‘weak’ blood, frequently alluded to by indigenous healers. Attempts by clinics to lower blood pressure in patients with diabetes got portrayed as tantamount to ‘weakening’ the blood, construed to be a highly undesirable state signifying greater bodily vulnerability to harm of many kinds.

But it was the duration of diabetes, its chronic character, which in the final analysis made ‘compliance’ so hard for patients. For a period, adhering to a clinic regime may be manageable and financially feasible, perhaps even the preferred option (cf. De-Graft Aikins 2005). Sustaining this course was the hard part, especially if financial hardship became greater, or cultural expectations among men about the conviviality of shared drinking proved increasingly hard to resist, or weight loss (for either sex) provoked unwelcome gossip and rumour. Moreover, over time, as complications of diabetes arose, many patients tended to deal with these as distinct illnesses, rather than as causally connected with the diabetes. The extract below illustrates the mutual frustration when a longstanding diabetic patient believed that the doctor was ignoring her ‘real’ health problems, which she reckoned were no longer to do with diabetes, while he could not get her to accept that the symptoms were a consequence of her poorly managed diabetes (cf. Kleinman 1997):

In such circumstances, as symptoms and episodes of illness became more manifold over time, there would always be relatives and neighbours to urge the possibility of dealing with the problem once and for all. That meant turning to one of the many specialists offering the possibility of a ‘cure’. Much of the rhetoric of the more senior clinic staff suggests an uphill struggle in the face of patients’ moral frailty and educational shortcomings: against ignorance and self-indulgence, but also against ‘superstition’ and the deceptions of indigenous healers, well illustrated in the frustrated remark of a nurse at one of the Biyem-Assi clinics:

But frustrations like this could also be reciprocated by those who took a jaundiced view of both the effectiveness and cost of clinic care. In the words of one Yaoundé man, ‘Doctors treat us like guinea pigs for pharmaceutical companies, whereas we can get better with traditional doctors’ (12 October 2001).

Even where an individual with diabetes acknowledged that the condition required exclusively biomedical treatment, kin or neighbours might well reach their own conclusions about possible causes (and thus take appropriate action in response). A case in point concerned the death during fieldwork of a prominent man who had been a strong and very public advocate of biomedicine as the way to manage his diabetes. When he had what proved to be a terminal stroke, there was gossip about whether this had been caused as a result of a quarrel. The possibility that the stroke might be a complication of his diabetes was scarcely mentioned. His death a few days later accentuated speculation as to the ‘true’ cause, all the more so occurring at the start of the dry season, a time of the year widely associated with misfortunes of disease and death and a heightened danger of witchcraft. Such gossip made one man reluctant to give a eulogy at the funeral, fearing that by doing so he might lead others to imagine he had been seeking the deceased's position. Nor did the problems end there. While the family sought to organize an appropriately lavish funeral feast, they also wanted to make sure that mourners with diabetes had their dietary preferences properly respected. But what exactly were these preferences? Conflicting advice from individuals with diabetes or their families confused the picture. The result was a series of social gaffes in offering food and drink. Initially, for instance, soft drinks and red wine were offered. The soft drinks were rejected, but then switched for beer, which in turn was rejected by some though not by others. Drink was at least offered. Indecision about appropriate food to place in front of diabetic mourners meant that in the end a number of them left early, still unfed and clearly upset. When another member of the family was diagnosed as having diabetes a short while later, there were some who surmised that this might be attributable to failures of hospitality at the funeral feast.

SEEKING DIFFERENT ANSWERS: INDIGENOUS HEALING

As with the opening vignette, this last case illuminates a number of tensions and difficulties surrounding the understanding and management of diabetes, and also draws us towards a closer consideration of indigenous healing. A word first about terminology. We use the general terms ‘indigenous healer’ and ‘traditional healer’ interchangeably. Anthropologists sometimes demur at the connotations of the pre-modern implied by the term ‘traditional’. Yet ‘indigenous’ is arguably too essentialized a term, as also is the ‘ethnomedical’ preferred by some writers (De-Graft Aikins 2005, for example). Whatever our misgivings, the term ‘traditional healer’ is widely recognized in Cameroon, and its French equivalent, guerisseur traditionnelle, is equally part of common currency in francophone circles. Moreover, the term tradi-practicien is also common, often seen in advertising of services on signboards. Perhaps most typical in local English idiom and in Pidgin as a generic term is ‘native doctor’. The multiplicity of local terms current within a multilingual city means that the more general English or French terms are widespread. These terms embrace a host of diverse specialists, from herbalists to occult practitioners, whose claims to healing skill may rest on the use of plants, animal or inorganic products like earth and stone particles, or divinatory power. Most, but by no means all, are men. While several writers have pointed out that the notion of ‘traditional healer’ in sub-Saharan Africa masks some wide variations of technique and therapeutic practice (Geschiere Reference Geschiere1997; Gufler Reference Gufler2003; Pool 1994a), some form of divination is common to many of the varieties of indigenous healing practice. Even a common distinction between herbalist and diviner obscures the fact that herbal treatments often accompany divination.

That said, in Biyem-Assi, most of those contacted in this research identified themselves as herbal specialists. Those who would say, for example, ‘I am a doctor of natural medicine, so I use only herbs in healing my patients’, were not only asserting specialist knowledge of plants; they were also often distancing themselves from any assumption that they were practitioners who relied on occult powers. Distinguishing one's diagnostic practice as based on herbalism could be all the more critical because diviners from certain ethnic groups in the city were widely assumed to be witches themselves (cf. Geschiere Reference Geschiere1997; Rowlands and Warnier 1988). As one practitioner affirmed, ‘to understand some of these things about diabetes, you also need to be a witch’. By emphasizing a herbalist identity a healer marked himself off from any such possible association or suspicion, and simultaneously asserted a claim as a serious medical practitioner, with a conscious appeal to a more educated clientele. Signboards advertising an indigenous healer played a part here in claims to status as well as claims to healing skills. Some signs carried graphic pictures, which the advertiser presumably judged would make an impact. Yet healers who belonged to one of several healers’ associations generally looked down on such self-promotion. The very fact of belonging to an indigenous healers’ association was enough of an indication, as they saw it, of trustworthiness, and a sober signboard or even the absence of advertising was a mark of respectability. It was practitioners of this sort who were the more likely to want to reassure their patients that they offered the specialist knowledge to enhance rather than entirely supplant the efforts of biomedicine, implying – and even advocating – a partnership of sorts (though few were under any illusions about the response they usually received from ‘hospital doctors’), as the following remark indicates:

Yet whatever the techniques of diagnosis used by such a healer, a damaged social relationship was invariably identified as the ultimate cause of diabetes, exemplified through the harmful intervention of either an ancestor or a witch (as an enormous literature on divination testifies).Footnote ¹² Without rituals to restore ancestral protection (a matter to be affirmed in public), or nullify the threat from witchcraft (a matter to pursue in private, indeed in secret), there could be no final restoration to health, no ‘cure’. In this respect, diabetes is seen as like many other diseases, however ‘new’ it may be: one of numerous possible vehicles for manifesting damaged relationships, and just as amenable to restorative treatment as any other. It is in this explanatory framework that the claims and demands of the diabetes clinic get set alongside alternative claims relating to the capabilities of affronted ancestors or witches, as the next comment by a traditional healer suggests:

The dismayed protestations of more senior clinic staff that biomedicine could not possibly be considered on a par with rites to placate ancestors or counter witchcraft was daily confronted by the evidence that many Cameroonians of all social strata (including on occasions biomedical practitioners themselves) did indeed feel justified in weighing up the pros and cons of biomedicine and traditional healing in treating diabetes, and would readily combine the perceived benefits of both. Clinic-based and traditional diagnoses and forms of treatment thus contended for credibility, as the following comment by a man with diabetes in Biyem-Assi shows:

In short, while diabetes is considered a ‘modern’ disease, when it is recognized at all, it is not typically considered something that ‘modern’ medicine alone can treat effectively. In the final section we examine in more detail both causal explanations for diabetes and the meanings attached to notions of the modern.

EXPLANATION, MODERNITY AND CREDIBILITY

The three previous sections have demonstrated that while clinics treating diabetes emphasize to their patients that biomedical treatment is totally incompatible with recourse to indigenous healers, their staff are confronted constantly by clues that their patients feel few inhibitions in moving pragmatically between the two therapeutic approaches. Privately and often surreptitiously, clinic staff themselves may follow suit. All this, moreover, was just as evident in urban Biyem-Assi as in Bafut. What are the local explanations for such a patent disjunction between clinical expectations and therapeutic realities? Two main explanations were offered by clinic staff. The first was moral and psychological, and dwelt on the resolve (or rather the lack of it) of patients. Christian connotations of sinners straying from the path of the righteous were never far away in such depictions – not surprisingly, as three of the four clinics in this study were Christian and evangelical. Predictably, this was not a frame of reference much favoured by those who themselves suffered from diabetes. Interestingly, however, it was shared by the wives of some diabetic men, frustrated with what they saw as their husbands’ unwillingness or waywardness in following a clinic's advice. A second local explanation was economic in character, and emphasized poverty and material hardship. The costliness of clinic-based treatment – specifically pharmaceutical drugs – was indeed something that patients and their families commonly bemoaned – and, as we mentioned, allocating funds to such medical costs also had a gendered dimension. To some degree clinics recognized this, and Awah was reminded that, particularly in Yaoundé, the poorest were unlikely to be registered with a clinic at all, the implication being that those individual patients included in this study almost by definition did not come from among the poorest.¹⁴

What this local reasoning does not touch on is the way in which contention around compliance draws on tacit notions of what is or is not modern (cf. Ferzacca 2001). This is to link the ethnography we present here with a larger literature in the anthropology of sub-Saharan Africa, focusing in particular on the ‘modernity’ of witchcraft today. Before discussing how our ethnography relates to this literature, it is worth clarifying what we mean by suggesting that these arguments concerning patients’ compliance with clinic strictures draw upon tacit notions about modernity and what it means to be modern. Although we use the word ‘tacit’ here, the idea of the modern is not some intangible abstraction (see Nyamnjoh et al. 2002; Page Reference Page2007; Rowlands 1995). On the contrary, clinic rhetoric – particularly in Biyem-Assi – repeatedly asserts an image of what ‘being modern’ entails. This is promoted by senior staff quite consciously, with both educative and moral intent. Moreover, the bureaucratic procedures and time discipline expected of patients, the routinized laboratory tests and measurements, the unfamiliar understanding of the body and its ailments, as well as a new terminology, all combine to express the distinctive character of ‘modern medicine’ in relation to the management of diabetes. Clinic rhetoric thus fosters a set of associations between scientific knowledge or procedures and the rejection of ‘tradition’ or ‘superstition’ as the foundation of modern thinking and the cornerstone of better health (see note 9). That message is reinforced, moreover, as the state becomes increasingly active in promoting public health responses to chronic diseases, such as diabetes, seen largely as diseases of modernity, and caused primarily by changes in diet and ‘lifestyle’. Thus, a further association is made in clinic rhetoric with a modern disease that is only knowable and treatable by modern medicine.

Such forceful rhetoric in the context of diabetes management seems to consign indigenous healing and all that it entails firmly to the category of the pre-modern or unmodern. Yet interestingly a number of anthropologists have argued over the last two decades that witchcraft (or sorcery) is anything but unmodern, in writing that is highly germane to our own focus on diabetes and indigenous healing. Cameroon indeed has figured prominently in theoretical debate about witchcraft or sorcery in recent anthropology, notably with Geschiere (1997) and Rowlands and Warnier (1988) arguing that the politics of the post-colonial state has fanned rather than dampened the conditions for witchcraft to flourish, to the extent that it can be seen as ‘a popular mode of political action’ (Rowlands and Warnier 1988: 129). More generally, Geschiere's The Modernity of Witchcraft (1997) argues a now well-supported view that the ubiquitous presence of witchcraft in public and private affirms its salience for discourses around modernity in many parts of sub-Saharan Africa (cf. Ashforth Reference Ashforth1996; Comaroff and Comaroff 1993, 2003; Moore and Sanders 2001). Far from being anachronistic and left behind by modernity, as the older rhetoric of progress and development would assume, witchcraft shows no signs of fading away as an explanatory idiom through which to make sense of changing circumstances. This is just as true of the urban scene as the rural (Nyamnjoh Reference Nyamnjoh, Moore and Sanders2001; Rowlands and Warnier 1988). Witchcraft, and the constellation of practices produced by such beliefs, is thus more than simply compatible with modernity; it may be revitalized by it. A revealing commentary on the prominence of witchcraft in Cameroon today is also provided by what Geschiere calls ‘a dramatic shift in jurisprudence’ (1997: 169). Where the colonial state (and for a while its successor) had sought to suppress witchcraft accusations, nowadays those suspected of witchcraft are much more likely to be charged and face heavy sentences. In this process, indigenous healers may have a newly recognized and highly significant role to play as prosecution witnesses.

Yet in what sense does this evident ubiquity imply modernity? In what way is the capacity of witchcraft – and by extension all indigenous healing practices – for revitalization an expression of modernity? Not all recent anthropological writing on witchcraft in Africa has in fact been persuaded by the claims made for its modernity (Rutherford Reference Rutherford1999; Green Reference Green2005; Sanders 2003). Green and Mesaki, for instance, stress its ‘explicitly unmodern associations’ in southern Tanzania (2005: 371). While contemporary cultural and political dynamics may well foster the prominent place of witchcraft beliefs in Cameroon (and other countries of the region), and as Geschiere suggests even give rise to new forms of witchcraft (1997: 166), does it follow that such a constellation of ideas about health, harm and cure are construed by Cameroonians as ‘modern’, an integral feature of modernity?

No unambiguous answer seems to us wholly credible, and we rehearse in these two final paragraphs the contrasting perspectives necessary, we judge, to a nuanced conclusion. The range of writing arguing for the ideological prominence of witchcraft in African public discourse is certainly persuasive. Yet using diabetes as a case study we take the view that the ‘modernity of witchcraft’ line of reasoning also underestimates the force of science and biomedicine in people's lives. By showing how people move between alternative systems of diagnosis and treatment it is true that we have emphasized the continuing fragility of biomedicine's impact and the accompanying frustrations of clinic staff. But equally we would stress that what clinics offer is not easily discounted. Moreover, the rhetoric which accompanies clinic practice promotes, as we have seen, an insistently emphasized dichotomy in which ‘modern medicine’ stands opposed to ‘traditional superstition’. Such language carries weight, and is buttressed in the case of many clinics by the additional authority of various Christian churches. It may frustrate the more senior clinic staff that patient compliance with their regime proves so hard to secure; but even so, patients first get their diabetes diagnosed in clinics, start their treatment there, and do return, albeit inconsistently (cf. De-Graft Aikins 2005). Moreover, those who survive over the long term with their diabetes (for many of course do not, as complications increase) tend to become confirmed advocates of the benefits of clinical medicine, increasing the currency of such ideas and influencing the local climate of opinion. What Geschiere and others may thus have overlooked is precisely this indigenous counter-argument to the association of witchcraft (and by extension traditional medicine) with modernity, deriving its authority through clinic practice but ultimately from the status of science as one of the universal benchmarks of modernity. And within this frame of reference, we contend (contrary to Ferzacca 2001) that while indigenous healing may be ubiquitous it can never quite escape from its positioning by default as biomedicine's unmodern ‘other’.

None of this, however, is intended to minimize the arguments on the other side. To start with, the evidence presented above points towards a number of traditional healers consciously resisting that ‘unmodern’ label, and repositioning themselves as potential partners with biomedicine, appropriating elements of the language as they do so. Clinics may (and do) deride such pretensions. Yet the World Health Organization's endorsement of the declaration by the African Union of 2001–10 as the ‘Decade of Traditional Medicine’ also undercuts such an unbending stance.Footnote ¹⁵ In any case, the attraction and the power in popular consciousness of the explanations and ritual solutions proffered by indigenous healers cannot be gainsaid, whatever view is taken of their efficacy. Despite the strictures of clinic staff, people carry on oscillating between – and blurring the boundaries between – the two frameworks of knowledge and practice, and they do so because the promise of the indigenous healer remains inherently plausible to the majority. Moreover, hard as it was for more senior clinic staff to recognize, the science on which biomedical knowledge and procedures rest was far from being taken as the axiomatic bedrock of reliable knowledge that its practitioners and advocates assumed to be self-evident. The clinic regime could thus become, for a good number, both costly and unconvincing. It was not the intentions of clinic staff that were doubted. Rather, it was the explanation for diabetes, and the prognosis, which met with scepticism. The result was that, in the weighing up of diagnostic and therapeutic alternatives, biomedicine could easily be seen as less plausible than ‘traditional’ knowledge. What was seen by clinics as a flight from reason was seen instead by those living with their diabetes as a sensible (one might say modern) testing of the claims and capabilities of alternative medical approaches. For those suffering from diabetes and their kin the unpredictable interventions of ancestors or witches were commonly taken as perfectly plausible eventualities in themselves, in Biyem-Assi and in Bafut. Thus, when such diagnoses were offered by traditional practitioners, the possibility was not an outlandish one for those consulting, notwithstanding the disbelief and dismay of clinic commentaries on hearing such stories. Instead, recourse to these traditions came with all the trust accorded to something that was taken as ‘second nature’.

Acknowledgements

We gratefully acknowledge the contribution of the research participants in making this study possible. Funding for the PhD on which this article is based came from the North-East Diabetes Trust, the International Diabetes Federation, and Newcastle University. We are grateful to all three. Particular thanks are due to Professors George Alberti and Jean-Claude Mbanya for assistance in obtaining funding; to Professor Nigel Unwin, co-supervisor of this PhD; and to Hazel Glass. We should also like to thank Kate Degnen and Bob Simpson for their comments on an earlier draft, as well as the editor and two anonymous readers.