The compassion of concealment: silence between older caregivers and dying patients in the AIDS era, northwest Tanzania

Abstract

In northwest Tanzania, where AIDS has been present for 25 years, AIDS-related illness is a trigger through which community members discuss personal experiences of loss and assess social relationships. The terminal phase of AIDS demands intimate social relations between patients and caretakers. In this final phase of illness, caretakers are scrutinised for their behaviour towards the patient. In the moral world in which caregiving takes place, the act of concealing is considered an intrinsic part of proper care. Current debates on morality, stigma and secrecy inform my argument that acts of concealment around dying are not so much related to the exclusion and ostracism of patients but to inclusion and compassionate care.

Dans le Nord-Ouest de la Tanzanie, où le sida est présent depuis vingt-cinq ans, la maladie associée au sida constitue une amorce de discussion sur l'expérience personnelle de la perte, et d'évaluation des relations sociales. La phase terminale du sida exige des relations sociales intimes entre les patients et les soignants. Durant cette phase finale de la maladie, les comportements des soignants vis-à-vis des patients sont étroitement observés. Dans le monde moral où les soins sont administrés, préserver le secret est considéré comme un acte intrinsèque à des soins appropriés. Les débats actuels sur la moralité, le stigmate et le secret m'amènent à soutenir que les actes de dissimulation à propos de la mort sont liés à l'inclusion des patients et à la compassion pour eux, plutôt qu'à leur exclusion et à l'ostracisme vis-à-vis d'eux.

En la zona noroeste de Tanzania, donde el sida ha estado presente durante los últimos veinticinco años, las enfermedades relacionadas con el sida sirven de motivo para que los miembros de la comunidad hablen de las experiencias personales de la pérdida y evalúen las relaciones sociales. La fase terminal del sida exige relaciones sociales íntimas entre los pacientes y sus cuidadores. En esta fase de la enfermedad, se examina a fondo qué comportamiento tienen los cuidadores hacia el paciente. En un mundo ético en el que tienen lugar los cuidados, el acto de ocultar se considera una parte intrínseca de un cuidado adecuado. Los debates actuales sobre la moralidad, el estigma y el secretismo respaldan mi argumento de que, cuando se va a morir, los actos de ocultación no están tan relacionados con la exclusión y el ostracismo de los pacientes sino más bien con la inclusión y el cuidado compasivo.

Keywords:

• stigma
• HIV/AIDS
• care
• Tanzania
• concealment

Introduction

Scene one: sleeping patients, exposed illness

It is a hot afternoon in July, four months after my arrival in the village. I have been feeling a bit uncertain lately. Though death was present in many conversations, AIDS has, so far, been invisible to me. It is this invisibility that triggers my question to my neighbour: ‘Do you know who has a patient in their house right now? I never met anybody; it seems as if AIDS is not present at all’. My neighbour points to a house close by: ‘In that house there is a patient, but she is not yet sleeping’. She uses the word ‘sleeping’ as many people do when they want to indicate a patient is close to dying. ‘And the daughter of Ta Stephen, she is in Rwantege [dispensary] now, they brought her this morning’. I am surprised; I have known Ta Stephen for four months and have seen him almost every day, but I had no idea he had a patient in his house. ‘He has not told me anything about it’, I say to my neighbour. She answers, ‘He can't, he can't. It is very difficult to say that and it is also very painful’. ‘But is it certain that it is AIDS?’ My neighbour is certain. ‘Yes, she was admitted in Rwantege dispensary and it is serious. The mother slept there last night’, implying the patient is close to death.

In northwest Tanzania, those who suspect they are infected with HIV and their close relatives usually keep their suspicion to themselves. Even when the terminal stage has been reached, silence around AIDS-related illness is generally maintained. When a villager meets someone in whose house a patient is dying, there will be respectful enquiries about the condition of the patient, without reference to what is wrong with the ill person. At funerals the cause of death is also hidden or phrased as ‘he/she was ill for a long time’. Due to such silence, AIDS is at first invisible. But for villagers, who is ill or who might be ill is a subject of public speculation.

When patients reach their terminal stage, both silence and speculation intensify. Intimate feelings and emotions, relations between caretakers and dying patients, the social status of the family, the behaviour of caregivers towards patients – all these private issues become the subject of public scrutiny when a patient becomes seriously ill (cf. Mutembei 2001). These discussions take place in bars, on the street, in houses, on the paths through the banana groves where villagers meet. Often these discussions do not have malicious intent; it is the usual way to discuss everyday events. Stories around the terminal stage of AIDS-related illness thus provide an entry for understanding local moralities around caregiving.

At the same time, experiences of caring for people dying of AIDS were not easy to articulate for those who had lived through the loss of close relatives. The more I got to know older caregivers, the more they preferred silence. Older caregivers talked about the difficulties of caregiving in a circuitous way. Stories about families giving care also reflected concealment – to whom the stories were told and in the wording and manner of speech used to tell them: soft whispers, heads close together, reflective pauses.

This article examines how tactical silence employed by older caregivers, patients and people outside the family over the course of the dying process relates to notions of ‘good care’. Caring for dying patients is often framed within discussions on whether the extended family is able to cope with the increased demands on its resources (Ankrah 1993; Seeley 1998). What remains unexamined is how care and ‘good care’ are enforced by the community and shape how caregivers and those around them deal with the demands.

Silence and care

To understand acts of concealment around AIDS-related illness, I propose a relational framework asking: What does concealment protect? How does it affect the interpersonal relationships of older caregivers? How does it relate to notions of good care?

Current analyses of stigma and how it is embedded within social relations shed light on how AIDS-related illness can distort interpersonal relationships. Posel (2004) argues that emotions surrounding sex, death and embodiment are central in any analysis of stigma (5). The terminal stage of AIDS-related illness implies a slow degeneration of the physical body (Henderson 2004), often leading to feelings of disgust and shame (Posel 2004, 8). AIDS illness thus often leads to a social death that precedes the actual biological death (Niehaus 2007). This social death, related to feelings of shame and disgust, can damage relations between people living with AIDS, their caregivers and others. Secrecy and silence are responses to this envisioned shame and disgust (Posel 2004, 8) and thus a mechanism to protect patients (Emlett 2008, 712).

This paper analyses how tactics of silence and concealment contribute to what is considered good care. My point of departure is not the patient or stigma as such, but the caregiver and their caregiving. Following Zigon (2007), I see illness events as moments that open up debates about morality and praxis; moments of severe illness shed light on local moral dispositions (140). Seen from this perspective, tactical silence and concealment can only be understood when situated within particular moral worlds (Kleinman 2006; Yang et al. 2007). Concealment of AIDS illness functions as a core praxis to protect the honour of individuals, an honour that frames what good care is (Iliffe 2005, 363). Concealment – framed in the public health literature as denial – is in my analysis central to practices of relating in a local moral world.

Social relationships within and between families are constantly renegotiated through living together. As care takes place within specific kinship relations (Dilger 2010, 117–8), moments of severe illness reveal the expectations of relatives, but also their shared histories that influence the quality of care. Not everyone should care, and not everyone can provide good care. Silence and concealment in response to AIDS-related illness can be seen as a mode of ‘indirect communication’ (Wood and Lambert 2008, 216), part of a cultural repertoire to explain and communicate socially sensitive topics. Proper ways of addressing socially sensitive subjects, depending on one's position in a situation, generate respect in the face of misfortune (Vu Song Ha 2008, 174; Whyte 1997, 217). In this sense, practices of concealment are moral ways to behave in specific social spaces.

Kinship relations in northwest Tanzania are informed by the ekibanja landholding system, which structures generational and gender relations and expectations (Setel 1999; Weiss 1996) as well as relations of care. Caregivers have different positions in the lengthy process of giving care, depending on their kinship relation with the patient. Relations between parents and children are particularly rife with expectations and (moral) notions of care. While AIDS-related illness can generate conflict between caregivers and patients, caregivers are also scrutinised to see whether they provide good care. Tactical silence in situations of care point to underlying tensions in relations of kin.

Setting and methods

The ethnographic fieldwork on which this article is based took place in a village in Muleba District in Kagera Region, northwest Tanzania, among older Haya2 mostly aged 60 and above. It was part of a larger study on the changing care relations of older people in the era of AIDS. Older people mainly live off the perennial cultivation of plantains and the sale of coffee. Some are also engaged in income-generating activities such as brewing and selling beer, mats and surplus crops. Rural-urban migration is common to the area. The monetisation of the economy in combination with the out-migration of adults has made it increasingly difficult for older people to receive care and support in their old age.

The fieldwork for the larger study was conducted in Kiswahili and Kihaya and relied on two core methods: the longitudinal following of several families headed by older men or women and the collection of stories around family events such as conflicts, funerals and illness episodes. The study was conducted together with Gussy, a 38-year-old female co-researcher from the area who interpreted conversations but also collected information independently. Semi-structured interviews were taped and transcribed from Kihaya to English; informal conversations were written down immediately after a visit.

As this paper deals with the unspoken and the concealed, I base it on a combination of sources: conversations with older people who had cared for and lost relatives dying of AIDS; observations of what was said and not said; four case studies of patients in their terminal phase; gossip about patients and caregivers; and responses to direct questions about AIDS and the issue of silence. Analysis of this material highlights the nature of intimate relations between older parents and their dying children as well as the paradoxically supportive and conflicting relations between families and the community.

The study ended seven months before the Tanzanian government commenced the national roll-out of antiretroviral therapy (ART). This is then a historical analysis of a time when AIDS automatically implied death. Nevertheless, two short follow-up visits in 2005 and 2008 indicated that no significant changes had taken place in communication around AIDS-related illness.

‘This disease of ours’: silence in an advanced epidemic

While the AIDS epidemic in northwest Tanzania began in the early-1980s, high prevalence areas3 remained limited to northern and urban areas. In Muleba District, to the south, prevalence rates at their peak reached 10% in rural areas, declining to 4.3% in 20004 (Kwesigabo 2001, 12). Of 47 older people in the sub-village, 22 had lost children to AIDS; two had lost all of their children to AIDS.

The long presence of AIDS in the village has affected how older people understand the illness and the death of its victims. They were extremely concerned over today's relationships; there was a general sense that AIDS was the inevitable consequence of contemporary daily life, framed within general responses to uncertainty (Whyte 1997, 204). In their references to AIDS, older people often mentioned moral punishment and gharika (Noah's flood). They had in mind the history of commercial sex work in the region, mobility and the often shady business deals with which the first AIDS ‘victims’ were associated. Other narratives voiced defeat. In stories about the illness of relatives, AIDS was referred to as ‘the man which came to finish us’ or rumala (the disaster). Or as a 78-year-old man put it: ‘watu wote tumedudulika’, literally meaning ‘all of us have become insectised’, using the metaphor of an insect to suggest the inevitability of infection. Nevertheless, older people mainly referred to AIDS as ‘this disease of ours’, indicating its acceptance as something belonging to Haya society rather than something caused by an outside agent (Mutembei 2001, 128–9).

Mutembei (2001) shows how understandings of AIDS are grounded in broader understandings of long-term illness, death and social danger and how these have changed with experience of the epidemic (62). As illness belongs inside the family, concealment is a normal way of dealing with chronic illness. AIDS, however, cannot remain hidden, exposing families to the scrutiny of outsiders. This visibility is most painful in cases of AIDS, where death is seen as final rather than a transition into another life (Mutembei 2001, 108). A Haya saying – ‘they have put words on me’ – conveys the social dangers present in being spoken about by others. In the context of AIDS, it means that gossip and public scrutiny raise suspicions about one's status, rendering the person familiar but somehow no longer themselves (Mutembei 2001, 71). Elesta, aged 70, describes how accusatory modes of communication – ‘talk’ – no longer serve a purpose due to the scale of the epidemic:

You can't talk about anything, nowadays you don't ask your friend anything. When you meet your friend s/he will tell you that s/he is in the same situation as you. S/he will tell you that maybe two children have already died. Others tell you that four children have already died, you find another one telling you that her children are sick. Nowadays we are no longer talking about it because we are all in trouble from this disease.

The fact that this is an advanced epidemic is central to the analysis of concealing practices. In the next two sections, I explore two ‘social spaces’ – relations between caretakers and patients and between caretakers and visitors – to understand what these instances of silence and concealment in the terminal phase of AIDS protect.

The morality of silence: concealing to create spaces for compassionate care

Caregiving for prolonged periods can strain caregivers' relationships within families. To understand concealment and what it protects in relationships between caregivers and patients, I begin by examining shifting power balances within families when a patient reaches the terminal stage of their illness.

A father's duty

On the same evening that we hear about the illness of our neighbour's daughter and her admittance to the hospital, we go to pay a visit. When we arrive at Ta Stephen's house around seven in the evening, we find him sitting in the living room; he has just returned from the hospital and his wife is still there. Ta Stephen is glad to see us, invites us to sit down, and a conversation ensues:

Ta Stephen: Thank you, it is really a problem. Right now, I have just come back, but her condition is not good.

Gussy:   We got the news, a woman who came to our house told us the news, but we did not know that her condition was this bad and my fellow [Author] was not there, she came yesterday. We were planning to go and visit her tomorrow.

Ta Stephen: Her condition changed yesterday noon. Her condition is not good at all, because her entire body is swollen. They say if you swell up, it is because of a problem with the kidneys. Therefore her condition is not good.

Gussy:   But is she conscious, can she talk?

Ta Stephen: Very little.

We:    And can she eat and drink a little?

Ta Stephen: Very little.

We:    Since when is she ill?

Ta Stephen: She started to be ill last year, the eighth month [August]. She used to live in Bukoba, but we took her to Rubya hospital. There she got some relief but she did not return to Bukoba. She started to live in a house behind this one. But her condition was so-so. Now she has become ill again. Her condition is not good at all and now you see that I cannot leave from here because we are awaiting news from there [dispensary]. Every three hours we receive news from the dispensary, a boy will come now to bring news from there.

We:    But have they checked what is wrong with her? What are they saying is wrong with her in Rwantege?

Ta Stephen: You know her problem is her kidneys. But if you look at her, if you examine her, you will see that together with the problem of the kidneys she has H I V [he pronounces each letter carefully, bending forward, looking at us].

We:    Pole sana, pole sana [We are very sorry].

Ta Stephen: Yes. HIV is dangerous, but what can you do?

We:    Since when have you known that it is HIV?

Ta Stephen: Since she returned the first time. That is the reason why I prevented her from going back to Bukoba.

This moment of illness somehow marks a change in our relationship. From this moment onwards, Ta Stephen is remarkably open. He explains how his daughter Elida, who is 36 years old, was married in another region. After her divorce from her husband due to ‘grave misunderstandings’, she went to Bukoba, the regional capital. In Bukoba she had a child from an ‘unknown father’. A year ago, in 2002, she returned to the village and was admitted to Rubya hospital. He immediately ‘knew’ she suffered from HIV, which he related to her stay in Bukoba:

You know, something else is very difficult to explain but it is obligatory [lazima] that I tell you, no other way [hakuna jinsi]: You know if a person leaves from here, and goes to Bukoba, and she goes to work. And at that time she does not have a specific job, it is a problem. It is dangerous.

It is dark now and Ta Stephen switches on the light. The atmosphere is intimate and what I hear Ta Stephen talk about, in subtle terms, are failed expectations: his daughter divorcing, leaving him with the burden of her child. We don't discuss AIDS, we discuss how expectations within care and support relations have changed. A relatively wealthy man and a clan elder, he relates how he keeps on caring for sick relatives, providing financial assistance, taking care of transport to the hospital, buying medicines and food, arranging for funerals and making decisions about the fostering of orphans. Ta Stephen has been doing this since his first son died in 1996:

You know my other child has died already, the one who was born from my first wife. He was born in 1963 and died in 1996 and his wife followed, and he left a child, who disturbed me very much. He died last year in the tenth month. Because he had no father and no mother and his grandmother was crazy, I was the one who cared. I brought him to Rubya, Ndolage, Rwantege [hospitals, between 3 and 30 km away]. In Rubya he stayed for two months and got TB treatment. And I was the only one looking after that child. No one else assisted me, sometimes I had to beg someone else to help care for the child.

Aside from caring for sick relatives, Ta Stephen is responsible for his elderly parents who are still alive at 86 and 92 years of age. They need full-time care. His burden of care differs from that of Elida's mother, who is present at her daughter's side, sleeping with her, cooking for her and cleaning her, providing all the physical care needed when a patient is confined to bed or admitted to hospital.

Concealing as part of ‘good parenting’

Relations between parents and children are relations of provision and authority (Whyte and Whyte 2004, 85). Fathers can endow their sons with clan land and thus exercise power over them. Sons build their houses and raise their families on plots of land adjacent to the house of their fathers, thereby creating an extended family, sometimes up to five generations. Women move to their husbands' land upon marriage where they have no access to clan land as they will always belong to the clan in their natal homes. Whereas parents are supposed to provide for their children by schooling them, providing them with land and opportunities and raising them to behave according to Haya norms and values, children are not ‘obliged’ (lazima) to provide for their parents in later life; this is considered wajibu, something done out of gratitude for mothers who suffered for their children. The presence of AIDS has profoundly changed the experience of being an older person, a parent.

In the stories of caregiving, it is striking how the relationship between parent and child is maintained as much as possible throughout the different stages of illness. Ta Stephen evokes a sense of duty towards his family; he provides and arranges for where his daughter will live and cares for his small grandson. There are implicit assumptions on how to act as a caregiver. In the eyes of older people, ‘compassion’ (huruma) is a core value within social relations. It is part of ustaraabu: proper, dignified, honourable behaviour.5 Huruma in inter-generational relations has to do with a sense of closeness, a good roho (‘nature’).

AIDS illness subtly changes the relationship between older parents and their adult children. The personal experiences of older parents and their dying children reflect ambivalence, the search for a redefinition of positions and emotions. Expectations of relations have, in essence, not changed; nor have practices of parenting and consolidating relations. Yet caregiving demands extraordinary efforts from older people. This dualism – the search for maintaining relationships as they ‘should be’ in new situations – is what creates pressure for older people.

The moral thing to do in response to an ill child is to protect them from gossip and from losing hope. Ta Stephen indicates that the only reason to talk about AIDS is to make practical arrangements. Concealment protects the relationship between the caregiver and the patient: open discussion can culminate in accusations between parents and their dying children:

I told my family members, I set up a small meeting and explained the problem and how we should handle it. And maybe I can tell a good friend. He will ask, ‘How is the condition?’ And I will say, ‘It is so and so’. But to say it is AIDS is not possible, for the following reasons. First of all, if you say she is suffering from AIDS, she now knows she is going to die, she is already dead and she can kill herself. And secondly, it is shameful, because it comes from prostitution. So to speak about it you are gossiping [about] your own child and when she hears she will leave the house and not speak to you.

Concealment here ensures that parents maintain the humanity of their dying children. Elida's father revealed her illness to us in an intimate setting, when we all expected the news of her death. But in later conversations, AIDS was never mentioned again. Elida knew she was dying but never spoke about it. Her mother did not speak about it either and only mentioned her loss after Elida had died. They were protecting the dignity of both the caregivers and the patient.

Concealing to create ‘loving’ care

Patients and caregivers attempt to maintain their relationships in everyday life. This relationship changes again when the patient becomes bed-ridden. It is extremely painful for caregivers to watch the symptoms of AIDS in its terminal phase. Henderson (2004) argues that the body's disintegration is often accompanied by disintegration of social relations (3). When the terminal stage is reached and the strains of caregiving increase, relationships change into ones of total dependency. The (looming) loss of a loved one and total dependency often create emotional tension.

There are clear gender roles in caregiving (cf Dilger 2010), in the division of labor and in what is considered ‘good’ care. For older women, good care involves intimacy and closeness, showing love to the patient. To show love is hard work in the terminal phase; the symptoms are terrible and the patient is unable to walk. As toilets are outside the house, the older person must exert enormous physical effort. To cook the food a patient needs, older women must collect firewood, buy eggs or milk and sometimes work on the land to buy necessities. And then there is the strain of living with a terminally ill patient who may die at any time. It involves sitting up at night and spoon feeding. As a woman of sixty-five said about the care of her daughter:

It is hard. You know where you are sitting now [on the sofa, with a cup of tea in front of me], if the patient is terminal she cannot even reach for that cup of tea. You have to place it in her hands. And then the patient wants an egg so you go to the neighbours and buy an egg and send a child to collect firewood and light the stove and cook the egg, and then she says, no, I am sorry, I am not hungry, maybe some milk, so you go to the neighbours to buy milk. It is very tiring …

The emotions of losing a loved one, the certainty of death and the intense physical experience of care lead to anxiety, stress, sleepless nights and to feelings of grief and anger. The association of AIDS with imminent death is often unbearable, especially in parent-child relationships such as the one Ta Stephen describes. Concealment serves an important function in maintaining a sense of closeness during an extremely emotional experience. Through concealment, emotions are contained and hidden and patients can continue living among family members as if the illness was not a reality. The cost is psychological and physical strain for caregivers.

Whispered morals: concealing to enforce compassionate care

Caregivers as well as patients have relationships with friends and neighbours, relationships that extend beyond the family. In these relationships, individuals constantly shift between concealment and sharing. Tactical silence plays a role in maintaining the dignity and position of the caregiver and the family.

A mother's lack of love?

The next morning Gussy and I decide to go to Rwantege to visit Elida. We are curious to see her as there are discrepancies between how Ta Stephen discusses care for his daughter and the stories Gussy has overheard in the village. She tells me:

They were speaking about it all the way down to Tophista. Tophista said that she had heard that Ta Stephen's daughter was seriously sick and I told her, ‘Yes, I got that information from her mother yesterday’. Tophista said that the people were laughing at the mother, because she did not care at all for her daughter, that she neglected her. They were speaking bad words about the mother. I asked why and Tophista said, ‘I do not know but she did not care for her at all. She went there yesterday when the condition worsened’.

This, I realise, is a serious allegation. In Kagera, a mother's care is the best and most important form of care. Without that care a patient will suffer. A mother not sleeping next to her daughter in the hospital implies a lack of love.

When we visit Elida in the dispensary we find Sofia, her mother and several siblings gathered around her bed. Elida is looking ill and is lying down, her mother is sitting on the bed drinking tea and eating bread. She relates how Elida was confused in the night and wanted to get out of bed, and laughs when she explains this. She also states that she has not slept the whole night and that she did not go on a planned trip to Dar es Salaam because of the patient. Outside the clinic, the nurse shares her doubts about how the family is caring: ‘The siblings are crying, but it seems they do not really show her love’.

I am intrigued by these words. What exactly is meant by ‘showing love’? I observe that Elida has hospital care, food, family around her and, later, when she returns home, a mattress, cups with milk porridge and mandazi (doughnuts) on a plate. When we leave Gussy remarks:

She will improve, but many times you see that people neglect the patient because they do not want to spend resources all the time, when they know they are going to die.

When I come to visit a week later, it is clear that Elida is dying. We talk about Saida Karoli, a popular singer who performed in Bukoba a week earlier. While I talk of drumming and dancing and songs, Elida lifts herself and asks me to tell her stories of Bukoba. We leave with her promise: ‘I will come to your house when I feel better’. She dies a few days later.

Concealing to protect the position of caregivers

In times of illness it is essential for caregivers to receive company – friends and neighbours who provide solace by sitting in the front room, even without saying much. Neighbours, visitors and friends also often provide assistance, bringing small gifts such as eggs. But amongst older people, there was a strong sense that talking about their problems was like asking for pity. Caregiving strictly takes place within families, especially concerning food and money – two areas that indicate a household's social standing – problems are not to be discussed with outsiders. Receiving company, then, is potentially dangerous for the social position of caregivers.

To prevent gossip, caregivers have different ways to conceal HIV. Sometimes patients are hidden (or hide themselves) in a room while visitors are received; the caregiver explains that the patient has ‘just gone to rest’. At other times, the patient is present but silence is maintained. Almost all older caregivers said something along the lines of: ‘The signs and symptoms were seen, so there was no need to say anything. The people could see for themselves’. In this way, older caregivers could reveal some of their problems without speaking about them and visitors could observe the state of the household.

Trust is an essential element in assessing what one call tell to whom. There was a general sense among older people that in contemporary society, trust had disappeared. They traced this heightened suspicion to poverty and to the struggle over resources, especially land, and increased fitina (envy) between people. Mistrust was also present within families and between close relations; some older women were accused of witchcraft or themselves accused their daughters-in-law of claiming their sons. Caregivers only tended to trust a select few individuals with their secrets, even within families.

Older caregivers therefore often remain alone in their caregiving. ‘In Kagera you die with your worries in your heart’ was an often heard end to our conversations.6 Concealment in care serves to maintain a family's dignity to the outside world (cf. Illiffe 2005); it protects the family's affairs, especially their social-economic status, from public scrutiny. Especially older mothers stated that silence in the face of gossip about their children was an act of dignity. To attribute an illness to AIDS when this is merely suspected is an act of disrespect, an act by which a person is turned into an already dead person (Whyte 1997, 217). For families, silence and concealment become a way of caring: one ‘speaks carefully and acts hopefully’ (Whyte 1997, 217) in situations where everyone knows what is going on.

These notions of dignity are related to a core orientation in Haya society: that of enclosure (Weiss 1996, 169). Weiss argues that this orientation is visible in everyday practice, from wrapping bunches of bananas to closing the doors when eating. Elida's mother was far from quiet. Even though she never directly mentioned her daughter's illness, she laughingly told me, her visitor, about her daughter's confused state. Visitors responded to this by doubting the extent of ‘real love’ Elida's mother had for her daughter. Concealment is a moral disposition in Kagera. How older caregivers approach AIDS and what they consider ‘good behaviour’ very much fits with what is considered virtuous for older women: being quiet and contained, suffering for one's children. Caregivers employ tactical silence to garner support from visitors and to protect their social position and the dignity of their families.

Silent discourse: the ambiguity of the ‘inside’

Conversations around caregiving in the village are tied to events, such as news that a patient is bedridden or hospitalised. Conversations are held in specific places, always away from the patient's family. Mainly women engage in the storytelling, especially stories about social relations and conflicts in other families (cf. Stadler 2003). By sharing such information, neighbours, nurses and community members evaluate how the illness is being handled by the family – not just the acts but the intention behind the acts as well.

Evaluations of care can strain relations between caregivers and visitors as the relationships between them are often close. Neighbours, for example, often know about the lives of the caretakers in detail. The Haya word for neighbour is omutaani, ‘those who enter with each other’ (Weiss 1996, 47).

Inside and outside should not be equated with any public-private dichotomy. Each relationship has an inside and an outside and contains concealment. Haya therefore actively reveal and conceal at every moment towards all people, including family members. Social relations are surrounded by secrecy; trust within relations is slowly built up by giving small pieces of unimportant information and checking how they are handled. Conversations concerning AIDS in the village are therefore often framed within an intricate social knowledge about the family history of the caregiver and the patient. Older men and women in the village can also play different roles at the same time. They can be active caregivers going to visit friends who are themselves involved in caregiving or who may have lost relatives not long ago.

Such constantly shifting positions shape the stories that are told. The stories often change slightly over time or are triggered and reframed by other events. It is therefore very difficult to maintain an inside, especially when family members have different relations of trust and there are conflicts in the family. It is difficult for caregivers to assess how visitors might frame information. In the intersection between caregivers and others, there is therefore a concealed discourse: people speak but there is no overt reference to the illness. Concealment is therefore never complete silence. Van de Port (2000) calls this the eloquence of silence; drawing on Taussig's (1987) concept of implicit social knowledge, he argues that silence is not necessarily the absence of speaking (229–30). References to AIDS are present. If a close friend asks, ‘How is the condition?’ and the reply is, ‘We were up all night’, she knows enough. The core of concealment between families and visitors lies in uncertainty over how the information will be used – whether visitors will associate the patient with promiscuity, adultery or prostitution or comment on the family's practices of care and how this could affect the family's position in the community.

The discussions around illness events thus have a moral purpose: to voice the core elements of compassionate care. These comprise closeness between patients and especially their mothers, as seen through physical presence and touch, kindness and the giving of hope and the hiding of emotions that could compromise the relationship. Visitors not only discuss but sometimes enforce norms around compassionate care, exposing to the community those caregivers who do not care properly.

Conclusion

This paper has explored how tactical silence enables ‘good care’ for patients dying of AIDS in northwest Tanzania. HIV is an advanced epidemic in this region. AIDS illness brings to the fore both local moral norms and relationships within families; the issue that occupies community members is not whether – but how – care is given. Illness events trigger questions that go beyond specific illnesses – compassion within social relations being one of them. Compassionate care is a moral construct that is continuously made and remade in response to illness events. Villagers have strong notions of what is considered ‘good care’. It includes tactfulness, closeness, love, duty and suffering – seen as reflections of the relationship between patient and caregiver.

A core element within compassionate care is concealment. Concealment is not denial; nor is it completely explicable by theories of stigma. It is a ‘core orientation’ – a ‘language’ in a society focused on enclosure, trust and social position. Concealment is ingrained in everyday practice; it is honourable not to expose internal affairs. AIDS illness reconfigures care relations within families and, because of its visibility, exposes families and their internal affairs to the community. Concealment prevents the potentially harmful exposure of conflicts and emotions between caregivers and patients and, hence, protects the social position of caregivers.

Concealment has a further side: it hides the interpersonal tensions that AIDS illness introduces into families. Illness heightens existing social conflicts, including witchcraft fears and allegations. Through concealment, anxiety, stress and anger are contained. Concealment hides power plays within families, maintaining family dignity.

Acknowledgements

I wish to thank Anita Hardon, Rijk van Dijk, Eileen Moyer and Sjaak van der Geest for commenting on earlier drafts of this paper. The research for this article was made possible by the Amsterdam Institute for Social Science Research at the University of Amsterdam.

Notes

^1. This article is based on a chapter from my dissertation entitled: 'Being old in times of AIDS: Aging, caring and relating in northwest Tanzania', Leiden: Africa Studies Centre (forthcoming).

^2. The dominant ethnic group in the region.

^3. In 1987, the Bukoba urban area had a prevalence rate of 24.2% (Lugalla et al. 2004, 186).

^4. These figures are derived from a population-based survey conducted by the Kagera AIDS Research Project in 1987. Follow-up studies in 1993 and 1996, using sentinel surveillance and population-based methodologies, found declining trends, attributed to the severity of the epidemic in the area and to AIDS-specific interventions (Lugalla et al. 2004, 185–6).

^5. See Moyer (this issue) ‘Faidha Gani? What's the point: HIV and the personal politics of (non)disclosure among activists in Zanzibar’.

^6. Many caregivers, suffer from stress, insomnia and anxiety over future life and the broader consequences of the impending death of their relative. In this sense, whereas the ‘shared understanding’ that is part of concealment leads to inclusion and dignity of both caregivers and patients, it also limits the social space in which caregivers can share the burdens of caregiving. This potential ‘downside’ of concealment for caregivers' psycho-social health would require an expansion of the argument that is outside the scope of this article but will be taken up in future studies.