Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Poetics (Amst). Author manuscript; available in PMC 2010 April 1.
Published in final edited form as:
Poetics (Amst). 2009 April; 37(2): 162–184.
doi:  10.1016/j.poetic.2009.03.002
PMCID: PMC2791322

Hearsay Ethnography: Conversational Journals as a Method for Studying Culture in Action


Social scientists have long struggled to develop methods adequate to their theoretical understanding of meaning as collective and dynamic. While culture is widely understood as an emergent property of collectivities, the methods we use keep pulling us back towards interview-situated accounts and an image of culture as located in individual experience. Scholars who seek to access supra-individual semiotic structures by studying public rituals and other collectively-produced texts then have difficulty capturing the dynamic processes through which such meanings are created and changed in situ. To try to capture more effectively the way meaning is produced and re-produced in everyday life, we focus here on conversational interactions—the voices and actions that constitute the relational space among actors. Conversational journals provide us with a method: the analysis of texts produced by cultural insiders who keep journals of who-said-what-to-whom in conversations they overhear or events they participate in during the course of their daily lives. We describe the method, distinguishing it from other approaches and noting its drawbacks. We then illustrate the methodological advantages of conversational journals with examples from our texts. We end with a discussion of the method’s potential in our setting as well as in other places and times.

1. Introduction

It is a truism of social analysis that theory and methods are intertwined. Yet our methods for studying cultural meanings have not kept pace with our theoretical sophistication. Contemporary theories of culture foreground meaning-making (Sewell, 2005; see also Lamont 2000; Eliasoph and Lichterman 2003; Fine 2003). But how can we document empirically the practices by which meaning is actually composed and revised in the course of mundane, every-day life? Here we propose a practical method: conversational journals kept by cultural insiders that permit access to the multiple and ongoing discourses through which meaning is made in situ. We demonstrate the value of our approach by analyzing texts in which rural Malawians interpret and re-interpret the meaning of the AIDS epidemic for themselves and their communities. Although the setting is particular, it illustrates the potential of conversational journals for capturing meaning-making in action (Swidler 1986) in other, more familiar, relational spaces.

The work of Clifford Geertz (1973a) is a useful starting place for a brief summary of recent developments in culture theory. Geertz, influenced by earlier semiotic and literary theorists, led cultural analysts away from conceptualizing culture in terms of individual-level, subjective meanings (e.g. “ideas” or “values”) to conceptualizing it in terms of publicly available symbolic activities and the codes that make them meaningful (Keesing 1984; Wuthnow 1987). Foucault and Bourdieu, albeit in very different ways, contributed a focus on discourses and practices (Ortner 1984). More recently, Sewell (1992) has proposed that cultural schemas are constitutive of the “structures” of social life, while also insisting on the active agency of human beings who “read” the structures around them for the schemas they instantiate, and who, since schemas are polysemic and transposable, can create new meanings by transposing, recombining, and adapting existing ones.

Some of the most promising empirical attempts to grasp the collective and emergent constitution of cultural meanings have come from studies of public ritual (Bellah 1967; Geertz 1973b; Alexander 1993; Sewell 1996; Spillman 1997; Olick 1999). Conversational journals share the virtues of this focus on culture as public and collective, while maintaining a methodological focus on meaning making as interactive and dynamic. The texts produced by local journalists recall the words and often the gestures of people as they are walking to an event, recounting it afterward, or simply talking about their ordinary lives. These texts capture the public deployment of meanings and the emergent collective realities those meanings create.1

To document practices of meaning-making, analysts must at some point convert experience into texts, either composing their own texts as ethnographers do (Geertz 1973a:19; Marcus and Clifford, 1985; Clifford 1986) or relying on available texts, such as a fugitive report of a court case or story (Darnton 1984), or “texts” created by intellectuals (Foucault 1978), institutions (Mohr and Duquenne 1997), newspapers (Jacobs 1996), reviews (Griswold 1987), etc. A major advantage of conversational journals, in contrast to most ethnography, is that the first level of transformation from lived experience to text—the transformation of public conversation into a written account—is carried out by a local immersed in the situation. The freezing of social experience in texts researchers can analyze happens not as an ethnographer records his or her fieldnotes, but as a member of the community being studied records what she or he has heard.

To study cultural meanings in contemporary settings, social scientists typically rely on data generated in response to questions addressed to individual respondents or small groups (surveys, semi-structured interviews or focus groups), or on the immersion of a cultural outsider in a community. The former make it impossible to capture the dynamic discursive practices involved in ordinary meaning-making, as people actively pursue their own agendas in natural settings. Individual ethnographers, like those who rely on historical texts, can, and do, come closer to meaning-making in action, but they are constrained by what is visible in their particular research site and audible within their own, necessarily limited, social network.

The conversational journals we describe below offer access in a new way to the active practices by which meaning is socially composed as people filter rumors and gossip about events, or exercise their collective creativity (Setel 1999) as they interpret what they have observed or heard about.2 In what follows, we attempt to demonstrate several advantages of conversational journals as an addition to researchers’ tool kit of methods. First, the texts of the journals provide a rare opportunity for researchers to overhear the ebb and flow of everyday conversations of ordinary people in natural settings that range from casual conversations in the market, to women’s group meetings to barroom brawls (Eliasoph and Lichterman 2003). Second, the corpus of texts, created by multiple diarists over a period of nearly a decade, give us greater access to the varied contexts in which local actors, pursuing their own agendas in a diversity of settings, actively reproduce the structures and meanings of their social world. There are also other, more pragmatic advantages to our approach. It does not depend on the presence of a cultural outsider, either interviewer or ethnographer; indeed, the texts themselves are written by insiders. Conversational journals, at least in our context, are also relatively cheap. And because the method is cheap, it allows the possibility of systematic comparisons across time and social settings in a way that would be unimaginable for traditional ethnography. Importing thirty ethnographers to spend years in rural areas would not be practical. Because a multiplicity of conversational journalists can sample a wide variety of settings, contexts and speakers, the texts hearsay ethnography produces permit systematic analysis of the relations among speakers, contexts, and meanings—and of changes over time.

In what follows, we first introduce the unfamiliar material in our texts through an extended example that provides a flavor of their style and richness. We then provide a description of our method along with an evaluation of its limitations and of the quality of our data. We follow with excerpts from the texts to exemplify the varieties of natural contexts in which rural Malawians collectively and publicly address the AIDS epidemic. We then suggest the potential of our method for analysts of culture in other settings.

2. An Extended Example

Our example, one of several discussions of a notorious local character, is occasioned by the death of a woman, Abiti [Miss] Baidon, long known in her community as a prostitute. These (much shortened) excerpts illustrate the ongoing and contingent conversations that frame the collective narrative of AIDS. All proper names have been changed, and journal excerpts are cited using the pseudonym of the journalist and the date of the journal in year, month and day format.

The journalist meets two women, one of whom she knows from their school days, the other a stranger, walking to the funeral of Abiti Baidon. The conversation begins with gossip about Miss Baidon. After invoking the common trope in the community that death from AIDS is a “salary” or just repayment for bad behavior, one woman confides to the others her own urgent concern, her fear that her own husband’s infidelities will kill her and their children:

After we greeted each other, Mrs. Bruce said to me that I did well to come and mourn for Miss Baidon because she would have killed all the people of V___ and other men from the outside areas. Miss Baidon has died of AIDS as a salary or gift for what she was doing.

Her friend who [wore] a traditional suit, green in colour and the white sandals answered. She said that the death of Miss Baidon has concerned her very much because she was sleeping with her husband. She had been quarreling with her husband for a long time because of her and her marriage was near to an end. Her husband was challenging her that he could divorce her and marry Miss Baidon.

The conversation then segues from infidelity to AIDS, with a meandering discussion that alludes to longstanding collective constructions of the significance of infidelity in the epidemic, social structures (gender, community elders), a version of the epidemiology of HIV, and a request for moral support:

About the AIDS disease, the woman said that her husband was advised at the church by the church elders and the Nkhoswes [traditional marriage counselors] that these days life is difficult because there is no time for enjoyment since there is the AIDS disease. Having one woman and depend on her is a very good thing because you can save your life and your children’s life…. Though her husband changed his behaviour but the woman was still worrying…. Now if [Miss Baidon] was infected during that time that she was moving with [a colloquial term for promiscuity] her husband, it is openly that her husband was also infected and if he was infected it also means that she is also infected with that AIDS therefore she will just die for nothing. Her husband will kill an innocent woman like her. The woman was very worried a lot because she has children and she was saying that her children will suffer a lot if their parents will die because they are very young….

The conversation ends, perhaps only temporarily, with advice:

She began crying and I told her that she must stop crying because [she] had no evidence that she was also infected or not.

At the funeral, the diarist witnesses a public ritual and hears spontaneous “AIDS education.” The Village Headman says that his people should learn from Miss Baidon’s death, for she was a prostitute and suffered a long time before she died of AIDS. Then the head of the mosque, Sheikh Abel,

…stood up and talked to people about the behavior of prostituting. He said that he was very shy [ashamed] because she [Miss Baidon] cheated God. In the Quran, prostituting is forbidden but every Muslim should get married. As for men, they are very lucky that they are allowed to marry more than one wife but they should make sure that they are faithful to them and he is loving all of them equally.… AIDS has no medicine. If you are infected, just wait for the death but it is also the painful disease because takes a long time for one to get recovered and it also takes time for one to die and leave the world…. He lastly asked all the Muslims to stop prostitution and the men were also asked to stop having other partners who are not their spouse to reduce the number of people who die of AIDS. (Alice 030523)

Miss Baidon, known more familiarly by her first name, Stella, appeared in several earlier journals, by three different journalists. At the time she first appears in the journals, a topic of gossip among young men hanging out at the local trading center, she had already “retired” from her profession and returned to her village, providing an opportunity for people to note that her body had changed, conventionally interpreted as a sign of the onset of AIDS.

Stella, a veteran prostitute at Vingula trading Centre commonly known as (STELLA MPAMBA LITAKO, Yao Phrase meaning that Stella only relies on her buttocks for daily living) has finally moved from Vingula Trading Centre to her home Chaphuka about 8–9 kms away from the Centre because she said that she is now grown up and just wanted to go home to retire. But according to what Harvey Saidi said, I learnt that she just decided to go home because men were no longer interested in her the way they used to do in the past when she was fat and attractive. Therefore after seeing that she was not making business, that’s why she decided to go home.

The conversation then turns to AIDS and to the likelihood that Stella—and presumably other prostitutes—was killing men like themselves.

And Harvey further continued that, before Stella left, he had been saying, who ever drunk water from her well should know that he drunk poisoned water and that sooner or later will die. He said that he meant that whoever had sex with her should know that he is infected. Harvey said he came to know her very well like the way he did during the time when he came from South Africa with his friend Medson. Medson was the one who was going out with her, and that’s why they knew each other. (Michael 011228)

Each time that Stella was invoked in our texts, there was a “cloud of commentary” (Hammel 1990:467) that combined facts with anxious imaginings. The occasions for discourse were quite various: in one it was Stella’s physical presence; in another she was not physically present but took center stage because she was sleeping with the husband of one of the conversational partners; in yet another, not quoted here, Stella appears as an actor in a scandal. The participants were also various: men at a bus stop and at a trading center, women at a borehole and walking to a funeral; some conversations were among participants who knew each other well, whereas others included strangers. And we have undoubtedly missed many similarly spontaneous conversations about her: the community that knew Stella extends well beyond her village. Moreover, conversations about her may not have ended with her funeral—she might well have been invoked later when other men and their innocent wives died. And, if not Stella, then other prostitutes and other deaths would replace her, calling forth abstract public health advice, intense speculation, rumor, moral lessons, and sympathy for victims as well as critique of those who should know better than to have sex with prostitutes. These successive interpretations illustrate the extraordinary richness of these texts and the complex collective meaning-making activity the journals capture.3

3. Turning Conversations into Texts

3.1 The Production of the Conversational Journals

The development of hearsay ethnography was stimulated by practical concerns. In 1997 Watkins and several colleagues began to study the role of social networks in influencing responses to the AIDS epidemic in rural Malawi. Because the focus of the Malawi Diffusion and Ideational Change Project (MDICP) was demographic, the primary data would come from multiple waves of a survey, supplemented by semi-structured interviews.4 After the first round of the survey the researchers had a great deal of data about the composition and structure of the social networks in which rural Malawians talked about AIDS They had not however, learned much about the content of the social interactions—what people said to each other, rather than to interviewers, about AIDS or their strategies for avoiding infection and death—and even less about the wider everyday interactions that shaped responses to the epidemic.

Thus, the researchers improvised by asking several high school graduates who had worked for the MDICP and who were living in or near the study sites to be participant observers as they went about their daily routines. If they overheard anything concerning AIDS, they were to make mental notes of what people said and did, and then write their recollections word-for-word in commonplace school notebooks that evening or soon thereafter. The notebooks were given to a local intermediary who mailed them to the researchers.

More than 700 journals have been written since 1999, each approximately 12 single-spaced typed pages, and each usually covering several different conversations or incidents. Since there are frequently several people conversing, we overhear, at second hand, several thousand people. Twenty-two journalists (9 females, 13 males) have contributed to our corpus of texts, with three (two males, one female) contributing very frequently, 13 frequently, and six only occasionally. The diarists wrote in English, a language learned in school, and used parentheses or carets (< >) to set off their explanatory comments or untranslatable expressions in the local language. The handwriting and repetitions suggest they often wrote rapidly. We have retained locutions that reflect local adaptations of English. English is taught in Malawian public schools starting in Standard 5, equivalent to U.S. fifth grade, and has become somewhat indigenized. For example, to be sexually promiscuous is to be “movious” and one who has multiple partners is said to be “moving around,” an Anglicization of a Chichewa expression, woyendayenda, derived from the earlier association of multiple partners with migrant labor. The naturalness with which the journalists adapt English to Chichewa, chiYao, or chiTumbuka linguistic forms means that their English is somewhat closer to local languages than is the standard English in which a Canadian, British or American ethnographer might translate local languages. We have retained most of the idiosyncrasies in grammar and spelling, although on occasion we insert obviously missing words in brackets and make minor grammatical changes in the interest of legibility.

Our method, conversational journals, provides the texts that we analyze below. These texts record hearsay evidence: we hear only secondhand, from the journalists’ ears—and their memories—to our eyes.5 As we have noted, social analysts frequently use texts created by others. Our method “artificially” produces an enormous number of texts that, we believe, give unparalleled access to the meanings that circulate in a given society.

We are not the first to focus on the dynamics of social interactions (cf. Mehan 1979; Fine 1987; 2003; Katz 1999), to recruit “insiders” as research collaborators (Williams and Kornblum 1985; Newman 1999; Price and Hawkins 2002), nor to recognize the advantages of qualitative methods for studying AIDS (Parker and Ehrhardt 2000; Nguyen 2004). However, the pragmatic constraints and “openings” presented by conversational diaries give unusual empirical breadth to what is usually limited to the intensive study of a single interactional context.

3.2 The Journals and the Journalists: How Good Are the Data?

Just as the texts analyzed by others reflect the biases and interests of those who recorded them, so also do our texts. Here we describe the creators of the texts we analyze, and ask how their interests might shape those texts. Although in Malawian terms these high-school graduates are relatively well-educated, many such people are unable to find jobs in the urban formal sector. Rather, they remain in villages, living side-by-side with those who have little or no schooling, and engaging in the same activities as others—small-scale trading, tending their maize fields, attending their church, going to neighbors’ funerals, and so forth.

The journalists’ close networks, the ones in which they spend most time, are homophilous, as are close networks elsewhere (McPherson et al. 2001), but many of the conversations they overhear have a very diverse cast of characters. For example, the most prolific of the female journalists is on many committees in her community and sometimes attends regional or national meetings of these groups; and many women, but also men, come to her for advice. The male journalists spend much of their leisure time hanging out with friends in the nearby trading center, the bus depot, or at a bar, where there may also be friends of friends or strangers. The male journalists write primarily about men’s conversations, the women about women’s, reflecting the gendered interaction typical of sub-Saharan Africa.6

The journalists were paid US$30 for an 80-page school notebook, an amount that was deliberately set high relative to incomes in rural Malawi, as an incentive to continue with the project. Incentives raise the possibility of fakery. But here we are in the same position as any researchers, from those who conduct surveys to classical ethnographers: neither we nor they know with absolute certainty whether reports of informants are accurate. We can, however, take advantage of information both external and internal to the journals to evaluate our texts. We draw on other data collected by the research project (e.g. the surveys and the semi-structured interviews) as well as many months in rural Malawi participating in MDICP data collection. In addition, the large body of texts provide opportunities for systematic analysis of internal evidence: because some of the more notorious characters in the area, such as the prostitute Miss Baidon, appear in the journals of more than one journalist, and some actors reappear in multiple journals of the same journalist, we can examine consistency of representation across journalists and over time. Most convincing, however, are the internal qualities of the journals. It is evident as one reads them that only a gifted novelist could have manufactured such a variety of voices, situations, incidents, and viewpoints. As Kaler (2003) observes, it would probably have been much more work to invent these situations and voices than simply to record them (some of the journals are available at, along with other data collected by the larger project).

There is no doubt that the pay motivates the journalists to seek out situations in which AIDS is likely to be discussed. Initially, the journalists produced one or two journals a month, but their productivity increased, first after the poor harvest of 2001 and then dramatically with the famine of 2002 when grain prices rose by approximately 500% (Malawi National Vulnerability Assessment Committee 2002). Journalists also sometimes pad their journals, for example by reporting at numbing length on a village AIDS committee’s informational meeting or reproducing nearly verbatim a pastor’s sermon. We have not discouraged such tactics, feeling that it is better not to censor what the journalists write. But this increased output does not undermine the value of the evidence they give about where and when discussions of AIDS take place. Indeed, that the journalists are motivated to write at length means that they often provide an extraordinarily vivid sense of the context and setting of their conversations, as well as of the rhetoric itself.

Despite the verisimilitude and the external and internal consistency of the journals, a journalist is not a mechanical amanuensis.7 These are texts of recalled conversations, not recordings, and their authors surely did not remember perfectly everything that was said, and by whom. But for those interested in culture, rather than in conversational analysis of the sort Emmanuel Schegloff, Harvey Sacks (Schegloff and Sacks 1973; Sacks, Schegloff, and Jefferson 1974) and others have pioneered, there are advantages (beyond the logistical impossibility of placing recorders wherever people gather to chat) to this reliance on local recall. Much as in ethnographers' notes, what can be gleaned is not the exact language, but the way meaning unfolds within a specific social situation. In this sense, the local journalists’ selective memory is not unlike the selectivity that shapes what ideas get picked up and passed on naturally, to become part of the larger culture (Varenne 1987; Wuthnow 1987).8

4. What Conversational Journals Capture and What They Miss

In this section we use excerpts from the journals to illustrate the advantages of these methods for capturing aspects of the dynamic processes of collective meaning-making that other methods typically miss.

4.1 Meaning-Making as a Collective Process

The example below is one of many, often lengthy, conversations among young, underemployed men hanging out at a trading center, talking in a bar, or playing bawo, that capture meaning-making as a collective process. Just as Miss Baidon’s death led an innocent wife to seek advice for her personal problem, here a man asks his friends for help with an urgent matter—the venereal symptoms he has been experiencing. This excerpt illustrates several characteristics of other journals: the raunchy terms in which the men, razzing their friend, talk about the symptoms of gonorrhea; peers urging a friend to be faithful or to use condoms; and the widespread misapprehension that sex with someone with AIDS means inevitable infection.

When we were chatting there one of my friends there began saying that now he is wondering as to what is happening to him. I asked to him what is happening to him which he is wondering of? He answered that he has only one sexual partner in his village but he had been sleeping with her for a quite long time and he had never noticed what he is observing rather feeling nowadays. One of the friends asked, what do you feel? He answered saying that he feels pain at the front of his penis more especially at the foreskin. I laughed and then his friend said that possibly the foreskin has a crack. His friend said that he doesn’t see any crack at all and he had clearly observed it and seen that the foreskin is okay no sores at all.

The men discuss possible diagnoses—perhaps the problem occurred because the sufferer was not circumcised or wore underwear that was too tight. They ask for more symptoms, which the sufferer provides. One of the men says that the disease is chizonono [gonorrhea]; he establishes his authority by regaling the others with a vivid description of the disease’s likely course if untreated, based on his own experience:

[Eventually] the person becomes rather produces bad smell due to the pus he produces and even green flies follows him where ever he is which is the very insult to him and a great problem. I laughed and friend laughed too. Then friend went on saying that chizonono is a very bad and dangerous disease which requires fast treatment to avoid one becomes burren [barren, sterile] and producing the bad smell which results from the pus which he may be producing.

After a discussion of possible treatment options—whether it is better to go to a traditional healer or to the hospital—the men begin in earnest trying to work out whether the young man’s partner has endangered him by being unfaithful; note how they swiftly move from epidemiological logic to gender:

Then I asked my friend why [if] he had been sleeping with her since November last year (2003) up to now in January 2004 that he had to face rather experience the pains he feels now and not from the first few months he started sleeping with her? One friend said that it could mean that the girl has another sexual partner apart from him. I agreed to what friend said. Then friend continued saying that these girls are very dangerous and when they are with you, sleeping with you they pretend to love you much there at that moment and when she is with someone else and also she pretends to love him very much and like she doesn’t have anyone else but him alone.

The man’s friends urge him to “divorce” his girlfriend, but the young man says he can’t because he “loves her so much” and because of the gifts he has bought her (he lists the gifts and their cost). Then AIDS enters the conversation:

His friend said that he made the great mistake falling in love with the school girl. His friend asked why a mistake? He answered saying that school girls are very unfaithful ones, they don’t trust one sexual partner but go for many and the end result being spreading the virus…. He said that indeed for sure his sexual partner has no AIDS. His friend laughed and said that anyone who goes for more than one sexual partner nowadays of AIDS that one has a high possibility of having the virus which causes AIDS…. Friend said that even the radio says that those having sexual transmitted infections are more likely to also have this virus which causes AIDS. We were just listening, and friend [who was] criticised said that after recovery or after being healed certainly he will not drop the sexual partner…. He went on speaking that since he had been sleeping with her for a long time and moreover plain sex [“plain” is widely used to mean without a condom], then there is no need that he can divorce her for if it is the matter of AIDS disease then he had already contracted it and how can he avoid AIDS and if she has it it means he had it…. Chatting really proceeded to the extent that we came out of the topic. (Simon 040130)

One of the great advantages of conversational journals is that they demonstrate the process by which meaning is socially composed. Even supplementing short-response survey questionnaires with semi-structured interviews or focus groups provides only a very partial glimpse of what goes on in social interaction (for examples taken from transcripts of semi-structured interviews, see Schatz 2002; Tawfik 2003; and Zulu and Chepngeno 2003).9 In addition, these methods are inadequate for vividly capturing the drama, the joking, the contradictions and disagreements of everyday talk (Billig 1987; 1992; Swidler 2001).

4.2 Agency and Action

Why is there such a difference between the texts produced by individual interviews and focus groups and the texts produced by our method? As Wendy Griswold (1987b) has argued in a paper on methods for studying culture, cultural artifacts are produced by agents implementing their agendas in contexts that constrain what they can accomplish. But the standard methods available to students of culture make that agency hard to observe. Researchers’ own questions and interests typically structure the research interaction; importantly, their agenda almost inevitably differs from or even disrupts ordinary interactions.10 In contrast, when researchers are in a position to overhear people’s natural conversations, as ethnographers are, they can begin to understand how cultural issues are raised when people are pursuing their own agendas in their own social worlds.

Classical ethnography does provide access to the everyday conversations in which participants implement their own agendas, though even in excellent ethnographies one almost never finds the back and forth of everyday talk.11 Village talk may also take a different turn when the anthropologist joins the conversation. Philip Salzman (1999:96) observes that ethnography “gives us a good idea of what people will say to anthropologists, what pronouncements it pleases them to make, which self-image they wish to present to us.” Salzman is too dismissive. An ethnographer who spends extended time in the field, observes and participates in local life and who has the skill to retain and record the details of conversational exchanges, would be able to capture some of the dynamics of everyday chatting, though obviously not in the number or variety of settings in which our several collaborators find themselves.12

A practical disadvantage of entering into the midst of an ongoing conversational universe is that we see some ordinary practices, such as attending funerals, or proposals of sex, but we miss others that need no explanation for those in the conversations (Power 1994; Elliott et al. 2002). A traditional ethnographer who stays for years in his or her research site will find out what people actually do as well as what they say, what people’s longer-term strategies and interests are, and how local institutions operate. The journals portray only indirectly the institutional background that frames ordinary social striving: we will not find out from the journals how land is inherited (though we can and do ask informants, just as an ethnographer might), which has implications when divorce is considered as a strategy for avoiding AIDS, or what it means in this society for someone to stop schooling at Standard 2, as a journalist’s momentary love interest did.13 Classical ethnographers would, as a matter of course, learn about social institutions, either by extended observation or by asking informants. Researchers who use conversational journals would have to do the same.

We should note here that while we speak of actors’ agendas, of the contexts and aspirations that lead to particular sorts of conversations, we are perfectly aware that not only the conversations, but also the actors themselves and the social contexts in which they participate are socially—and thus culturally—constructed. Nonetheless, it is an advantage of hearsay ethnography that rather than engaging local people in conversations constructed around researchers’ assumptions and agendas, the conversational journals give us access to contexts and agendas that come from participants themselves.

5. Conversations and Contexts: How and Where Meanings are Mobilized

The conversational journals capture a central feature of collective meaning making: the different ways cultural tropes are used and shaped in different contexts. More important than the physical location of conversations are the social locations that elicit varying uses of cultural meanings: scandalous stories provide entertainment; a chance meeting at the borehole offers an opportunity to seek advice for a deeply personal concern; gossip about other villagers provides narratives of moral instruction; and a chat at a funeral may turn into a philosophical discussion. Below we illustrate how participants in conversational interactions make meaning in a variety of different ways in different settings.

5.1 Scandals and Gossip

A surprising aspect of the diaries is the frequency of dramatic public incidents in which AIDS is invoked. In one journal, the diarist is at the market when he hears a commotion and hurries to follow the people rushing to one end of the market. He finds two men fighting. An onlooker explains that the men had been friends until the wife of one showed her husband a letter from the other, proposing a sexual relationship. Enraged, the husband confronts the seducer. The letter is read aloud, to much laughter from the audience. When the husband says he still loves his wife, members of the audience approve, calling out that it is right for him to stay with such a faithful wife, she will save his life! The audience debates whether the bloodied seducer should be killed. One man says suppose the wife had agreed and the seducer had AIDS. Another member of the crowd picks up on this and shouts “look his blood is black, that means he does have AIDS.” Someone else says the seducer was lucky this occurred in the market; if it had been in a less public spot he might have been killed.

The fight provoked laughter, but it served other purposes as well. For the women in the audience, the wife becomes, at least for the moment, a model of a faithful wife; for the men, the physical dangers of proposals are vividly enacted. Such dramatic incidents are evidently relished; in several instances the sole purpose of a visit to a neighbor is to relate a scandal, which—each time it is retold—offers opportunities for further, and perhaps different, evaluations of moral and immoral behavior.

Gossip about others is a pleasurable exercise of imagination, but it is also an important way in which people expand the boundaries of their understanding (Hannerz 1967; Sabini and Silver 1982; Dunbar 1992).14 In a world where even radio dramas are a rarity and almost no one can afford books or magazines (nor electricity by which to read during long evenings), conversation provides some of the same delights that literature, drama, and soap operas provide for contemporary Westerners (Burke 1973; Katz 1973–4; Press 1991; Radway 1991). The entertainment function of gossip no doubt affects the kinds of events the journalists hear about and thus report. But the relish people take in relaying scandals means that the collective evaluations that the anecdotes provoke are widely available as heuristics when listeners are faced with their own decisions (Kahneman and Tversky 1973; Abelson 1976).

5.2 Seeking Advice, Solving Problems

When Malawian villagers seek or receive advice about AIDS, these are not abstract conversations about some distant threat or repetitions of standard slogans about the dangers of the epidemic (though these do occur). They are instead active, sometimes urgent, attempts to solve problems (Bourdieu 1990). Sometimes the advice villagers receive from friends or even strangers reflects public health campaigns: the woman who fears she has been infected from her husband is advised to get tested before concluding that she is infected, and the young man who is diagnosed by his fellows with gonorrhea is advised to seek treatment lest he become sterile. Even when the message is standard, however, the advice is given with a local twist—if he is not treated, green flies will advertise that he has a sexually transmitted disease—and a local solution, divorcing the partner who, they believe, infected him. We do not know from subsequent journals whether the innocent wife or the infected young man followed this advice, but we do know what sort of advice was being offered spontaneously to those urgently worried about AIDS.

Sometimes people seek help with problems so intractable that conversational partners are unable to offer a solution. In the following excerpt, an acquaintance seeks advice from the journalist. The acquaintance confides that her husband has other sexual partners; he has given her sexually transmitted infections; her newborn baby is ill and a previous one died; and her own body is changing in a way that suggests AIDS. What should she do now?

Mrs. Iweni continued by complaining that she doesn’t know what to do then because she should say that she would like to divorce her husband. That can be possible but still she will be suffering from that disease that she has already been infected. If there is death, she will die and leave her children orphans and if she asks her husband to use condoms with his other sexual partners, she would do nothing since she is already infected and if she asks her husband to use some condoms with her, her husband will not accept that. Now Mrs Iweni was stranded and she asked me what to do on that issue but I told her that I had no say [nothing to say] since there is no any other way that she can do there than keep on staying with her husband as a marriage. Then I asked her to allow me going to my maize garden and discuss again about her problem next time, and she accepted my request. (Alice 021108)

Passages like these expose the uncertainty and confusion of those engulfed by the AIDS epidemic, in a way that would be difficult to elicit in surveys, semi-structured interviews or focus groups.15 The woman has good reason for thinking she is already infected, and there is no good answer to her dilemma, at least at the moment; her conversation with an acquaintance can only help her to worry the frayed ends of a thread that doesn’t lead anywhere, since neither she nor her acquaintance can see an exit from her situation.

5.3 Figuring Things Out

As the excerpts above show, people turn to others to try to solve personal problems, but sometimes they are also trying to solve what we might call cognitive problems: they are using a social process to try to figure out something important about how the world works (Hutchins 1995; Feierman 2005). Cognitive anthropologists have emphasized the “problem solving” role of cognitive frameworks, metaphors, and models with which people think about experience (Lakoff and Johnson 1980; Holland and Quinn1987; D’Andrade and Strauss 1992; D’Andrade 1995; DiMaggio 1997). This is certainly important with respect to illness of all sorts, but perhaps especially with respect to AIDS due to the long period between infection and symptoms (Setel 1999). We see in the journals how rural Malawians combine bits and pieces of information distributed among their network partners to develop a model of what AIDS is, how it is transmitted, and what, if anything, one can do to protect oneself.

In the following excerpt, two men discuss a sort of natural experiment. Much as epidemiologists would do, they draw on comparative evidence to decide for themselves whether condoms really work. The journalist has just been at a tavern, where he overheard two men listing the many disadvantages of condoms—from condoms retarding ejaculation to condoms as a sin against God. The journalist then overhears another conversation that takes the opposite view, based on personal experience:

And some boys were speaking with their friends at N___ Market that they trust using a condom because it protects one from getting sexually transmitted infections. One boy said that he slept with a woman whom people were saying that she had gonorrhea, but since he had sex with her, he has no sign of gonorrhea meaning that the condom he used at that time protected him. And he also said that his friend Fyson Nakoma slept with the very same woman a week after him and after four (4) days, Fyson also had gonorrhea that is when he [Fyson] started trusting a condom. He told his friends that he uses a condom always whenever he is having sex with a sexual partner or any other woman. The boy is still unmarried as of now. Another man told Fyson Nakoma at that time that if he prolongs using a condom he might develop a disease called Cancer. So, he must stop using them. And if he can’t stay without having sex, then he can get married to any woman whom he could love. (Derek 021221)16

Even more common than such natural experiments is the construction of “social autopsies” in which participants pool their local knowledge to formulate a collective account of an individual’s road to death. Such narratives have important cognitive functions: in a context where few are tested for HIV, they help participants decide which deaths are reasonably attributable to AIDS and what behavior might be implicated. As we shall see, they have important moral functions as well. The narratives typically begin with news that so-and-so is visibly ill or has died, or perhaps that his or her spouse has died. The participants then draw on their local knowledge to piece together a medical history and a sexual biography that together create a suspenseful narrative of the growing certainty that the ultimate illness was indeed AIDS. The autopsy often begins with a recounting of a succession of gory symptoms—people vomit, have constant diarrhea, sores, boils, swollen legs; they become “as thin as a two year old child.” But because these are also symptoms of other well-known illnesses, participants then support the physical diagnosis with a parallel sexual biography: genealogies of former sexual partners and their partners’ former partners and whether they had died of AIDS, or children who died in infancy or who are chronically sick (interpreted as a sign that the mother was HIV positive). By the time the funeral occurs, the community has used these autopsies to determine the cause of death.

Through the journals, we see that figuring things out, and the resultant socially composed knowledge (Feierman 2005), is an ongoing—and collective—project. In collectively constructing social autopsies again and again, for the same person as he or she becomes ill and dies, and for successive deaths in the community, people domesticate abstract information about AIDS by developing cognitive models of the relation between sex and death, but models with particular local features that permit participants to identify with the sufferers—to say “this could have been me” (see Holland and Quinn 1987; Quinn 1996).

5.4 Moral Lessons

In rural Malawi ordinary conversations about AIDS are often chaotic, in part because participants drop one thread of conversation and pick up another, in part because participants hold inconsistent views,17 in part because AIDS is particularly grotesque and frightening—it is a disease that has no cure and creates predicaments for which there are no comfortable solutions. But the specifics of physical and sexual histories often end in a moral generalization. The moral is sometimes implicit, but sometimes quite explicit: so-and-so was a womanizer and went to prostitutes but would never use a condom; he deliberately chose death.

The lessons are often partial and probably transitory, to be reworked in subsequent conversations. But at the end of most conversations, the participants reassure themselves that even if they feel overwhelmed, even if cognitive closure has not been achieved, there is a “just world” that offers some direction for action.18 Sometimes the lesson is explicitly what we should all do: we should depend only on our spouses, we should give up extramarital partners, we should use condoms. But sometimes the moral talk is less an overt lesson than a moral interpretation of events, the explication of a meaningful story behind the lurid details of the progression to death. The lesson of the social autopsies is that AIDS is not a random event that can strike anyone but is linked to people’s moral biography.

Sometimes the moral closure occurs in a more civic context such as a court case, where the chief authoritatively expresses a new community standard. In the excerpt that follows, a woman has asked for divorce from a philandering husband. As both sides present their case, some members of the audience shout support for the wife, others their support for the husband. At the end, the chief announces the verdict:

Woman, be free and do what seems good to you and to what you believe, you are a brilliant and courageous woman, I congratulate you, keep it up, such kind of behavior, that by doing that you are trying to teach stupid men a lesson and as well as protecting yourself from this deadly disease AIDS and also protecting the lives of others and children and those who still marry you in the future. (Simon 022602)

The task of reaching a shared opinion or evaluation is sometimes assigned by a researcher to a focus group (Gamson 1992). But conversational journals reveal the extent to which purposeful and collective moral discourse suffuses everyday conversation, as studies of gossip have insisted (Gluckman 1963; Epstein 1969; Sabini and Silver 1982; D’Andrade and Strauss 1992; Dunbar 1992).

5.5 Contexts and Content

We have seen the wide variety of natural contexts in which people actively pursue their own agendas, seeking practical advice, cognitive understanding and moral closure. Other conversations, like the one we describe below, are less urgent, more philosophical. A group of men attending a funeral spend a rainy night in the compound of the deceased’s family, pursuing a leisurely, meandering conversation that brings together an unusual number of discourses taken from an unusual number of authorities.

In a shelter built for the funeral, six men lie in the dark on three palm mats “chatting and snoozing.” The diarist knows only one member of the group by name. Although he “sleeps” (shares a mat) with a stranger, by the end of the conversation the stranger has become a “friend.” They begin with the death that occasioned the funeral (because the deceased was an elderly woman, the death was not diagnosed as AIDS), then move to religious interpretations of the origins of death, from there to AIDS, its epidemiology, the fallibility of modern science, the government’s AIDS prevention programs, and then to potential practical strategies for avoiding infection, which returns them to religion.

[A man said] death is taking anyone regardless of age, being old or young. He continued saying that it is like a joke, for example to us who have gathered here, that indeed the person we saw her yesterday, for example, now is a dead person…. Then he added, saying death started long ago and imfa sizoloweleka <Chichewa, meaning that no one gets used to death–but always fears it>. Then the man whom I slept together with said indeed, no one gets used to death; and it indeed started long time ago and as a punishment for what our forefather Adam did in the Eden and his wife Eva after [she was] bewitched by the Satan.

After a long excursus on Adam and Eva and the forbidden fruit, they return to the question of how death came into the world.

Then the man who slept together with me said that God was clever enough. He knew all about this. He knew that if people could not be dying then the end result will be that the population will be [so] high that no place can be found uncovered, as we see nowadays that there [are] a lot of places uncovered like the national parks and game reserves. But had it been there were no deaths where could all people born everyday be living?

The men then segue from the philosophical to present realities, from death in general to AIDS, and from AIDS as a punishment from God to AIDS as spread by mercenary women:

He went on saying that AIDS is killing a lot of people nowadays. Another one said indeed it’s true, but of AIDS indeed God has really shown himself that he is above all. He is even more above the great scientists who are proudly boasting and claiming that they are wise enough to eradicate any kind of disease, but not in case of AIDS. AIDS came from God and He created it to minimize the population….

The friend who slept with me said that indeed AIDS is claiming a lot of lives, not as some people say that AIDS is only killing those who move around with sexual partners. This is a total lie/total cheating, but this disease is a [world] plague (he said in Chichewa <AIDS ndi mulili>). And he went on saying that since it’s a crisis and an outbreak expect that everyone will contract it because it’s [a] penalty and punishment from God. You may be attempting to refrain from catching AIDS but AIDS is contracted in many different ways including sexual activities [but also] the sharing of razor blades and needles.

Then another one added, saying that indeed AIDS is contracted through the ways you had mentioned but the major way/means is sexual intercourse <he said njira (way) yaikulu (major), Chichewa>. Everyone agreed but my friend who slept with me said that indeed the major way is through sexual intercourse, this means [all will die] because no one refrains from having sex, for a normal human being [is] involved in sexual activities either through [a] love affair <zibwenzi> or through getting married.

Now come admonitions for behavior change, invoking the authority of official views of AIDS in the media:

Another one said but nowadays change of behavior is greatly and urgently needed because if this is not to be done <change of behaviour> then the end result is that we are all to die of it <meaning AIDS>. He went on saying that nowadays normal means of contracting AIDS are widely known to anyone now and only very stupid and very young people, especially children less than 6 years, are unable to have access to knowing more about AIDS, but any child from 6 years above as of nowadays I believe has an access to this message because nowadays, for example, here in Malawi, Malawi is developing in terms of media systems for, say, transmitting/disintermination [dissemination] of messages, of any messages, and as for AIDS these messages don’t miss to be aired out every day through newspapers and those who are lucky that they know how to read are very accessible to reading and be having advices from the readings they read about this disease.

They then turn to a more down to earth discussion of the temptations that make it difficult to avoid infection.

He said a lot of people are dying because they are not satisfied with their own women or husbands they have at home but want every woman to be his if he is a man and a woman wants every man to be hers if she is a woman/girl. Like the prostitutes, they do not want to be loved nor do they love the man, it’s because they love the money the man has which we consider hypocrisy kind of love <he said chikondi cha chinyengo, Chichewa>. Someone added saying that money is the great causing [causative] agent that brings AIDS into one’s life because when a girl or a woman sees that you do always have money or you work and be receiving money, say salary, a girl or a woman doesn’t rebuff you because she wants to be given money and be buying her wants. (Simon 030125)

They continue by discussing strategies of prevention: how to develop self control in order to avoid alcohol and Indian hemp, and the possibility of dedicating oneself to God in order to be faithful to one partner. The conversation ends abruptly when a church elder orders silence so that he can begin preaching.

Such a meandering conversation reveals how in such a setting, participants link one idea or image to others, drawing on a range of authorities. In a way no focus group or interview could match—and in a situation unlikely to be shared by an ethnographer—we see multiple discourses and frames of reference jostling together. More critically, we see here that no “answer” to the problems of life and death–the sort of answer that might appear to be provided by a focus group, for example—reigns unchallenged for long (Billig 1987). Over the long term, we expect to see new common wisdom emerge as a product of ongoing, dynamic collective thought (Hutchins 1995).

6. Substantive Payoffs

Here we summarize briefly some of the insights that the conversational journals added to our understanding of the responses of rural Malawians to the AIDS epidemic. We also suggest the potential application of hearsay ethnography to other areas.19

The journals create a healthy skepticism about some aspects of conventional wisdom on AIDS in sub-Saharan Africa: that there is “silence” about AIDS (e.g. World Bank 1999) and that talk of sex is taboo, that Africans either deny the existence of AIDS or are fatalistic about their chances of avoiding infection. Our method shows instead that AIDS is widely discussed in tones ranging from jocular to anguished, that there appears to be little embarrassment or inhibition in talking about sex among friends, that the majority of rural Malawians are worried about becoming infected and that they are discussing and deploying strategies to prevent infection. Indeed, we learned about strategies of prevention that it had not occurred to us to ask about on the MDICP survey: that careful partner selection is considered an adjunct to partner reduction and that divorce is considered an effective way to avoid exposure to AIDS (Watkins 2004; Reniers 2008).

How much do rural Malawians talk about AIDS? We cannot estimate from the journals how often AIDS comes up in conversation compared to, say, the difficulty of subsisting with limited economic resources.20 But funerals alone provide abundant opportunities: the monthly average of funerals attended based on the 2001 MDICP survey is approximately four for women and five for men. For people who walk long distances to each funeral and may spend the night there, funerals are a major focus of sociability (Dunham and Klaits 2002; Smith 2004). In one five-week period in 2003, a journal-keeper attended five funerals and herself talked with 9 people about these funerals (including relatives of the deceased who diagnosed the death as due to AIDS); she also overheard 16 others talking about these deaths (only two of whom she knew by name) [Alice 031005]. In addition, she had ten other conversations about funerals that she did not attend. There may well be some, even many, people who do not talk about sex or AIDS in circumstances where they can be overheard, or with people who might repeat their confidences to others. Inevitably, these people do not appear in the journals. But even those who do not talk but only listen will overhear, as the journalists do, elaborate stories that people tell about symptoms, sexual histories and death, the morals they draw in recounting these stories, and their advice to treat an STI, divorce a partner, or use condoms to avoid AIDS.

Since the conversations recorded in the journals are driven by participants’ rather than the researchers’ interests, they capture what ideas are salient and what meanings are naturally conjoined in local understandings. The journals can also provide important insights by telling us what is not said. As the journals make abundantly clear, the idea that sex is related to AIDS is widespread. On the other hand, little is said about AIDS orphans or the caregivers of those with AIDS, and there is little call for the government to do anything more or anything different about AIDS (see Whiteside et al. 2004). Either these issues do not come up in public conversation, or the ethnographers—and perhaps others in their communities—do not think these topics are related to AIDS. By calling attention to silences, the journals pose new research questions.

The journals also provide evidence to support skeptics of the validity and reliability of reporting in standard surveys on key issues, such as multiple partnerships. For example, the MDICP surveys found that only 2% of women and 9% of men acknowledge having had more than one partner in the last twelve months. The gossip relayed by conversational participants in our texts suggest that this underestimates multiple partnerships by a considerable, if unknown, amount (see also Helleringer and Kohler 2007). Although we cannot compare reports of sexual partnerships in the journals with objective measures of infection such as the prevalence of sexually transmitted infections or new cases of HIV, an excellent study of the sexual behavior of Tanzanian adolescents did just that (Plummer et al. 2004). They compared biomarkers for sexually transmitted infections with the frequency of premarital sex as reported in a face-to-face survey, an assisted self-completion survey, in-depth interviews, and participant observation.21 None of these reports were consistent with the biological markers. Particularly instructive for our purposes, the participant observers recorded many “third-hand reports”—reports much like those that fill our journals. They note: “These third person reports of scandalous sexual behaviour are interesting in their own right…given that almost all of the individuals for whom there were such reports actually tested positive for two biological markers 2–3 years later.” They go on to suggest that “such reports could be explored as a low cost alternative to a large scale survey in identifying high risk youth in rural areas…” (p. 55). These researchers thus suggest that a method like hearsay ethnography might give superior results even in the effort to get accurate data about sexual behavior.

The unusual texts our method produces allow students of globalization and cultural change as well as policymakers to assess change. We have been struck by the ways some elements of elite or official discourse on HIV have become domesticated in local conversation and gossip, while others, like the recommendation to use condoms, continue to meet articulate—if declining—resistance; still others seem not to have penetrated at all. Although we cannot tell what population is represented in the journals, we may have some confidence that this represents a real change of opinion (or even more important, what local people think it is acceptable or interesting to say to each other), rather than survey respondents’ increasing knowledge of what researchers want to hear. Indeed, because local journalists are local, they can track cultural stability and change as it is occurring. If over time conversations about AIDS become clearer and less chaotic—less likely to circle back on themselves or dissipate in uncertainty—that would indicate a growing collective mastery of the issue.

Although it is difficult to code material as diffuse and meandering as the conversational journals, we have begun coding the diaries for basic topics and themes (topics like condoms, funerals, or AIDS deaths; themes like God’s intervention in human affairs, the dangers of money, or official advice about AIDS). We are also coding the contexts in which conversations occur. We hope eventually to analyze whether the emphasis on particular themes has changed over time (does talk of condoms decline as their use becomes less controversial, for example?) and how the contexts in which issues arise change (do discussions of AIDS occur more frequently in casual conversations among strangers, for example, rather than primarily as villagers discuss the illness and death of neighbors?). We also hope to take advantage of the nearly decade-long series of diaries to ask about change over time in the internal structure of conversations: whether discussions of some topics become more orderly over time, whether fewer—or more—different arguments arise in debates about a topic, or whether discussions of certain issues move more easily toward a resolution.

The methodological advantages of hearsay ethnography extend far beyond the study of AIDS in sub-Saharan Africa. Because journals can be solicited on a multiplicity of topics (e.g. politics, economic distress, gender attitudes) from multiple journalists working in a variety of locations (the shop floor, the neighborhood barbeque, at church, in welfare offices or hospital waiting rooms), they permit systematic analysis of variation in socially composed meaning, such as the association between contexts and content (Eliasoph and Lichterman 2003). Are some topics raised more often or more easily in certain sorts of groups? What contexts encourage more complex arguments about certain issues, versus more casual, unexamined observations (Dillon 1993)? Do some contexts elicit clichés and bromides, while others lead to vociferous debate and contestation? Multiple diarists with access to varied situations also permit addressing new questions about what is systematically missing, what people cannot or need not talk about or which contexts make certain topics difficult (Eliasoph 1998; Noelle-Neumann 1993; Martin 2002; Randall and Kopenhaver 2004). And, if the time period covered by the journals is long enough, they permit tracing changes in collective understandings as these evolve over time.

Such methodological possibilities should be of great interest to those who study political life (who spontaneously discusses politics, the state of the world, or political corruption, and in what contexts?). One can also imagine measuring the salience of an issue by asking conversational journalists to record for shorter periods (a morning, afternoon, or evening) every conversation they overhear or participate in, so that one could ascertain how often people of different sorts in different situations discuss politics, or God’s will, or the price of food.

A particularly intriguing area that systematic analysis of conversational journals can explore is how various streams of discourse come together (for insightful approaches to this using other types of texts, see Spear 2003; Feierman 2005). One can imagine hearsay ethnography used systematically across urban, rural, and village settings—or in the advanced democracies in central metropoles versus more peripheral backwaters—to explore when and how elements of elite discourse are adapted and altered as they are assimilated to common-sense understandings of the world, and how popular, perhaps innovative, ideas filter into elite discourse as well.22

7. Conclusions

In conclusion, we think that conversational journals are a practical methodology that contributes to the project of studying collective meaning making as it unfolds and changes. By fixing episodes of public discourse as texts, conversational journals convey a sense of widespread and ongoing shared agendas. While several theoretical traditions, from the Durkheimian to the symbolic-interactionist, posit such a dynamic public realm, few methods capture its texture. Through many journals, collected from multiple journalists in a wide variety of situations, we witness cultural understandings evolving, tacking back and forth, sometimes folding back on themselves or breaking down in confusion—but over time, even in the course of a single discussion, collective definitions have shifted. At least in our setting, rural Malawi, and about an important issue like AIDS, which is problematic, frightening, salient, and challenging, people are not passive. Collectively and publicly, they dwell on the problem they face, piece together practical knowledge, gossip and authoritative opinion, to try to bring clarity, to construct a conversational and situational universe and to map potential ways forward.


Research supported by grants NICHD RO1HD/MH41713 (P.I.: Watkins, Susan C.) 9/26/01-8/31/06 “Gender, Conversational Networks and Dealing with STDs” and NICHD R01 HD044228 (P.I. Kohler, Hans-Peter) 7/1/03 - 6/30/08 “AIDS/HIV Risk, Marriage and Sexual Relations in Malawi” and by the Canadian Institute for Advanced Research. An earlier version of the paper was presented at the Population Association of America meetings, Philadelphia, March 31-April 2, 2005. We are grateful to to Bill Gentry for assistance and to Adam Ashforth, Rekha Balu, Robert Bellah, Jennifer Browning, Kim Dionne, Steven Feierman, Claude Fischer, Wendy Griswold, Bill Hanks, Jenna Johnson-Hanks, Dick Madsen, Rachel Sullivan Robinson, Abigail Saguy, Peter Stromberg, Bill Sullivan, Iddo Tavory, Steve Tipton, and several anonymous reviewers for valuable comments.



Susan Cotts Watkins is Visiting Research Scientist at the California Center for Population Research at University of California, Los Angeles and Professor of Sociology at the University of Pennsylvania. Her research has focused on large-scale demographic and social change, including fertility transitions in historical Europe and the U.S. and in contemporary Africa; the AIDS epidemic in Africa; and the role of social networks in these changes.


Ann Swidler is Professor of Sociology at the University of California, Berkeley. Her work has focused on culture theory, American society and culture, and more recently on cultural and institutional responses to the AIDS epidemic in sub-Saharan Africa.


1A skilled interviewer with rich knowledge of the local context, such as Farmer (1994), can also capture the evolution of collectively-generated understandings of AIDS, if not the detailed accounts of conversations that our ethnographers provide.

2In a study of a public ritual of king-making, historian and ethnographer Steven Feierman (2005:16) writes: “Oral knowledge that is socially composed is potentially much richer than knowledge that is collectively held and homogeneous. When bits of information are distributed among specialists, or among diverse owners, there is a division of labor in oral transmission…. On the one side of the divide, the academic historian pretends to have an encyclopedic intelligence, capable of assembling all the pieces in a single pattern. On the other side, the men who owned the knowledge assembled it through a social negotiation; the transmission of knowledge and the solution of political problems proceeded in a single seamless process.”

3The historian of Africa Thomas Spear (2003:6) takes a similar approach in a discussion of tradition. “Traditions, then, consist both of fixed precedents and principles and fluid processes of adaptation that enable them to be continually renewed in an autonomous and self-regulating process” (see also Vansina 1990).

4The MDICP, now the Malawi Longitudinal Study of Families and Health, has conducted five surveys in rural Malawi (1998, 2001, 2004, 2006, 2008). The initial sample consisted of approximately 1500 ever-married women and their husbands; in 2004, a sample of approximately 1500 adolescents (ages 15–24) was added. Semi-structured interviews with randomly selected sub-samples of the initial sample were also conducted. More detail is available at

5Much of what the journals record is what V.N. Vološinov (1973) called “reported speech,” with the rich sense of social context such speech conveys.

6In the 2001 MDICP survey, 89% of the AIDS conversational networks of female respondents consisted only of females; 88% of male respondents’ networks consisted only of males (see also Marshall 1970 and Oboler 1986).

7That the conversations are not tape recorded makes it impossible to do rich linguistic analyses of the conversations (e.g. Gibson 1980; Eder and Enke 1991; Stromberg 1993).

8We also miss the words of social isolates who do not talk with friends and neighbors. For some purposes, their absence from our texts is a limitation. But for our theoretical purposes, it is not: those who talk contribute more to the construction and alteration of socially constituted meanings than do social isolates.

9Although meant to stimulate natural discussion and debate, as implemented in research in sub-Saharan Africa at least, focus groups follow the agenda imposed by the research rather than producing more spontaneous conversation. Transcripts show not only participants but moderators following the model of classroom instruction in Africa, with the moderator asking questions and the participants answering one at a time, deferring to rather than joking with the moderator.

10Aaron Cicourel (1974), Charles Briggs (1986), Pierre Bourdieu (1996), and Alex Weinreb (2006) offer penetrating analyses of the difficulties inherent in the researcher-respondent interaction. The solutions they suggest are different from ours, but the spirit of their efforts to capture actors’ own agendas is very similar.

11For a stunning exception see Chernoff’s (2003; 2005) intimate portrait of the long-term career of a West African bar girl. Chernoff manages to capture the conversations in which his informant engages, the lives of her co-workers, friends, and lovers in vivid detail—relying on the reports of one extraordinary bar girl.

12Ethnographers know that their own gender, age, or life-cycle stage gives access to some but not other conversations and views of social reality (Powdermaker 1966; Ortner 1984; Oboler 1986). With multiple journalists, we learn what is said in a diversity of social locations.

13Some patterns, such as market structure, are barely visible in the journals because they are not visible to conversational participants: a woman buys a chitenje in the market, but does not see the political economy that leads from textile factories in Indonesia to the local vendor.

14Ulf Hannerz (1967) underlines the value of gossip for its recipients: “that the individual gets a map of his social environment including details which are inaccessible to him in his own everyday life” (p. 57).

15Jennifer Johnson-Hanks (2005; 2006) has pointed to the role of uncertainty in human action, particularly in insecure settings like those of sub-Saharan Africa. Such uncertainty is much more likely to be captured by conversational journals than by more conventional research methods.

16Tavory and Swidler (2009) use the conversational journals to analyze the meanings of condom use in rural Malawi and the complex sources of local reluctance to use condoms.

17Of course, contradictions and confusions characterize casual conversation everywhere, not just in rural Malawi (see, e.g., Bourdieu 1990; Billig 1992; Swidler 2001).

18In “Religion as a Cultural System,” Clifford Geertz (1973c) argues that religion operates to reassure people that the world is potentially masterable cognitively, emotionally, and spiritually—not to foster optimism, but to ward off “chaos—a tumult of events which lack not just interpretations but interpretability” (p. 100).

19We are in the process of coding the journals using NVivo 7, coding topics and themes as well as the contexts of reported conversations. Here we suggest lines of analysis such systematic coding might allow.

20To get a sense of the distribution of conversational topics, we asked four journalists to keep a record of each conversation they overheard or participated in over a period of seven days, noting the topic, the number of participants and the duration of the conversation. In the record of one journalist, AIDS and money tied for first place; for the other journalists, AIDS was in third, fourth or fifth place, following topics such as work, money, farming, marriages and religion.

21Despite extraordinary care to find interviewers who were native speakers of the local language and to create rapport, the researchers conclude that “If biological markers are used to validate this interview series externally, 32% of respondents provided unreliable responses, while an additional 8% provided reliable but invalid responses (a variable that could only be assessed for those who tested positively for biological markers). Nine (82%) of the 11 respondents with biological markers provided an invalid series of responses; however, if no biological marker data had been available, only three (27%) would have had inconsistent and thus clearly invalid reports, similar to the 32% inconsistent reports for respondents without biological markers” (p. 51).

22Carlo Ginzburg’s classic, The Cheese and the Worms (1980), illustrates the remarkably complex interactions between elite discourse and popular thought, as does Tarver’s (1995) analysis of political themes in talk radio.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Susan Cotts Watkins, Visiting Research Scientist UCLA, Professor of Sociology, University of Pennsylvania, California Center for Population Research, Institute for Social Science Research, 4284 Public Policy Building, University of California, Los Angeles, Los Angeles, CA 90095-1484, ude.alcu.rpcc@sniktaws.

Ann Swidler, Department of Sociology, University of California, Berkeley, CA 94720-1980, ude.yelekreb@reldiws..


  • Abelson RP. Script processing in attitude formation and decision making. In: John S, Carroll JS, John W, Payne JW, editors. Cognition and Social Behavior. Lawrence Erlbaum Associates; Hillsdale, NJ: 1976. pp. 33–45.
  • Alexander JC, Smith P. The discourse of American civil society: a new proposal for cultural studies. Theory and Society. 1993;22:151–207.
  • Bellah RN. Civil religion in America. Daedalus. 1967;96:1–21.
  • Billig M. Arguing and Thinking: A Rhetorical Approach to Social Psychology. Cambridge University Press; Cambridge: 1987.
  • Billig M. Talking of the Royal Family. Routledge; London: 1992.
  • Bourdieu P. The Logic of Practice. Stanford University Press; Stanford: 1990.
  • Bourdieu P. Understanding. Theory, Culture and Society. 1996;13:17–37.
  • Briggs CL. Learning How to Ask: A Sociolinguistic Appraisal of the Role of the Interview in Social Science Research. Cambridge University Press; Cambridge: 1986.
  • Burke K. The Philosophy of Literary Form. University of California Press; Berkeley: Literature as equipment for living; pp. 293–304.
  • Chernoff JM. Hustling Is Not Stealing: Stories of an African Bar Girl. University of Chicago Press; Chicago: 2003.
  • Chernoff JM. Exchange Is Not Robbery: More Stories of an African Bar Girl. University of Chicago Press; Chicago: 2005.
  • Cicourel AV. Theory and Method in a Study of Argentine Fertility. Wiley & Sons; New York: 1974.
  • Clifford J. Introduction: partial truths. In: Clifford J, Marcus GE, editors. Writing Culture: The Poetics and Politics of Ethnography. University of California Press; Berkeley: 1986. pp. 1–26.
  • D’Andrade R. The Development of Cognitive Anthropology. Cambridge University Press; Cambridge: 1995.
  • D’Andrade RG, Strauss C. Human Motives and Cultural Models. Cambridge University Press; Cambridge: 1992.
  • Darnton R. The Great Cat Massacre and Other Episodes in French Cultural History. Basic Books; New York: 1984.
  • Dillon M. Argumentative complexity of abortion discourse. Public Opinion Quarterly. 1993;57:305–314.
  • DiMaggio P. Culture and cognition. Annual Review of Sociology. 1997;23:263–287.
  • Dunbar R. Why gossip is good for you. New Scientist. 1992;136:28–31.
  • Dunham D, Klaits F. Funerals and the public space of sentiment in Botswana. Journal of Southern African Studies. 2002;28:777–95.
  • Eder D, Enke JL. The Structure of gossip: opportunities and constraints on collective expression among adolescents. American Sociological Review. 1991;56:494–508.
  • Eliasoph N. Avoiding Politics: How Americans Produce Apathy in Everyday Life. Cambridge University Press; Cambridge: 1998.
  • Eliasoph N, Lichterman P. Culture in interaction. American Journal of Sociology. 2003;108(4):735–794.
  • Elliott E, Watson AJ, et al. Harnessing expertise: involving peer interviewers in qualitative research with hard-to-reach populations. Health Expectations. 2002;5:172–178. [PubMed]
  • Epstein AL. Gossip, norms and social network. In: Mitchell J Clyde., editor. Social Network in Urban Situations. The University Press; Manchester, England: 1969. pp. 117–127.
  • Farmer P. Aids-talk and the constitution of cultural models. Social Science Medicine. 1994;38:801–809. [PubMed]
  • Feierman S. On socially composed knowledge: reconstructing a Shambaa royal ritual. In: Maddox GH, Giblin JL, editors. In Search of a Nation: Histories of Authority & Dissidence in Tanzania. James Currey; Oxford, UK: 2005. pp. 14–32.
  • Fine GA. With the Boys: Little League Baseball and Preadolescent Culture. University of Chicago Press; Chicago: 1987.
  • Fine GA. Towards a peopled ethnography: developing theory from group life. Ethnography. 2003;4:41–60.
  • Foucault M. History of Sexuality Volume 1: An Introduction. Random House; New York: 1978.
  • Gamson WA. Talking Politics. Cambridge University Press; Cambridge: 1992.
  • Geertz C. The Interpretation of Cultures. Basic Books; New York: 1973a.
  • Geertz C. The Interpretation of Cultures. Basic Books; New York: 1973b. Deep Play: Notes on the Balinese Cockfight; pp. 412–453.
  • Geertz C. The Interpretation of Cultures. Basic Books; New York: 1973c. Religion as a Cultural System; pp. 87–125.
  • Gibson MH. He-said-she-said: formal cultural procedures for the construction of a gossip dispute activity. American Ethnologist. 1980;7:674–695.
  • Ginzburg C. The Cheese and the Worms: The Cosmos of a Sixteenth-Century Miller. Johns Hopkins University Press; Baltimore: 1980.
  • Gluckman M. Gossip and scandal. Current Anthropology. 1963;4:307–316.
  • Griswold W. The fabrication of meaning: literary interpretation in the United States, Great Britain, and the West Indies. American Journal of Sociology. 1987a;92:1077–1117.
  • Griswold W. A methodological framework for the sociology of culture. Sociological Methodology. 1987b;17:1–35.
  • Hammel E. Theory of culture for demography. Population and Development Review. 1990;16:455–486.
  • Hannerz U. Gossip, networks and culture in a black American ghetto. Ethos. 1967;32:35–60.
  • Helleringer S, Kohler HP. Sexual network structure and the spread of HIV in Africa: evidence from Likoma Island, Malawi. AIDS. 2007;21:2323–2332. [PubMed]
  • Holland D, Quinn N, editors. Cultural Models in Language and Thought. Cambridge University Press; Cambridge: 1987.
  • Hutchins E. Cognition in the Wild. MIT Press; Cambridge, MA: 1995.
  • Jacobs RN. Civil society and crisis: culture, discourse, and the Rodney King beating. American Journal of Sociology. 1996;101:1238–1272.
  • Johnson-Hanks J. When the future decides: uncertainty and intentional action in contemporary Cameroon. Current Anthropology. 2005;46:363–385.
  • Johnson-Hanks J. Uncertain Honor: Modern Motherhood in an African Crisis. University of Chicago Press; Chicago: 2006.
  • Kahneman D, Tversky A. On the psychology of prediction. Psychology Review. 1973;80:237–251.
  • Kaler A. ‘My girlfriends could fill a Yanu-Yanu bus’: rural Malawian men’s beliefs about their own serostatus. Demographic Research, Special Collection. 2003. pp. 349–372.
  • Katz E, Blumler JG, Gurevitch M. Uses and gratifications research. Public Opinion Quarterly. 1973;37:509–523.
  • Katz J. How Emotions Work. University of Chicago Press; Chicago: 1999.
  • Keesing RM. Theories of culture. Annual Review of Anthropology. 1974;3:73–97.
  • Lamont M. Meaning-making in cultural sociology: broadening our agenda. Contemporary Sociology. 2000;29:602–607.
  • Lakoff G, Johnson M. Metaphors We Live By. University of Chicago Press; Chicago: 1980.
  • Malawi National Vulnerability Assessment Committee. Malawi: Emergency food security assessment report, Malawi Vulnerability Assessment Committee and SADC FANR Vulnerability Assessment Committee, Lilongwe 2002
  • Marcus GE, Clifford J. The making of ethnographic texts: a preliminary report. Current Anthropology. 1985;26:267–271.
  • Marshall G. In a world of women: field work in a Yoruba community. In: Goode P, editor. Women in the Field. Aldine; Chicago: 1970. pp. 165–191.
  • Martin JL. Power, authority, and the constraint of belief systems. American Journal of Sociology. 2002;107:861–904.
  • McPherson M, Smith-Lovin L, Cook JM. Birds of a feather: homophily in social networks. Annual Review of Sociology. 2001;27:415–44.
  • Mehan H. Learning Lessons: Social Organization in the Classroom. Harvard; Cambridge, MA: 1979.
  • Mohr J, Duquenne V. The duality of culture and practice: poverty relief in New York City, 1888–1917. Theory and Society. 1997;26:305–356.
  • Newman KS. No Shame in My Game: The Working Poor in the Inner City. Random House; New York: 1999.
  • Nguyen V-K. Uses and pleasures: sexual modernity, HIV/AIDS, and confessional technologies in a West African metropolis. In: Adams V, Pigg SL, editors. The Moral Object of Sex: Science, Development and Sexuality in Global Perspective. Duke University Press; Durham: 2004. pp. 245–267.
  • Noelle-Neuman E. The Spiral of Silence: Public Opinion--Our Social Skin. University of Chicago Press; Chicago: 1993.
  • Oboler RS. For better or worse: anthropologists and husbands in the field. In: Whitehead TL, Conaway ME, editors. Self, Sex, and Gender in Cross-Cultural Fieldwork. University of Illinois Press; Urbana, Chicago: 1986. pp. 28–51.
  • Olick JK. Genre memories and memory genres: a dialogical analysis of May 8th, 1945 commemorations in the Federal Republic of Germany. American Sociological Review. 1999;64:381–402.
  • Ortner S. Theory in anthropology since the sixties. Comparative Studies in Society and History. 1984;26:126–166.
  • Parker R, Ehrhardt AA. Through an ethnographic lens: ethnographic methods, comparative analysis, and HIV/AIDS research. AIDS and Behavior. 2000;5:105–114.
  • Plummer ML, Ross DA, Wight D, Changalucha J, Mshana G, Wamoyi J, Todd J, Anemona A, Mosha FF, Obasi AIN, Hayes RJ. “A bit more truthful”: the validity of adolescent sexual behaviour data collected in rural Northern Tanzania using five methods. Sexually Transmitted Infections. 2004;80(Supp lII):ii49–ii56. [PMC free article] [PubMed]
  • Powdermaker H. Stranger and Friend: The Way of an Anthropologist. W.W. Norton; New York, London: 1966.
  • Power R. Some methodological and practical implications of employing drug users as indigenous fieldworkers. In: Boulton M, editor. Challenge and Innovation: Methodological Advances in Social Research on HIV/AIDS. Taylor & Francis; London, Bristol: 1994. pp. 97–109.
  • Press AL. Women Watching Television: Gender, Class, and Generation in the American Television Experience. University of Pennsylvania Press; Philadelphia: 1991.
  • Price N, Hawkins K. Researching sexual and reproductive behaviour: a peer ethnographic approach. Social Science & Medicine. 2002;55:1125–336. [PubMed]
  • Quinn N. Culture and contradiction: the case of Americans reasoning about marriage. Ethos. 1996;24:391–425.
  • Radway J. Reading the Romance: Women, Patriarchy and Popular Literature. University of North Carolina Press; Chapel Hill: 1991.
  • Randall S, Kopenhaver T. Qualitative data in demography: the sound of silence and other problems. Demographic Research. 2004;11:58–93.
  • Reniers G. Marital strategies for regulating exposure to HIV. Demography. 2008;45:417–438. [PMC free article] [PubMed]
  • Sabini J, Silver M. Moralities of Everyday Life. Oxford University Press; Oxford: 1982.
  • Sacks H, Schegloff EA, Jefferson GA. A Simplest Systematics for the Organization of Turn Taking for Conversation. Language. 1974;50:696–735.
  • Salzman P Carl. The Anthropology of Real Life: Events in Human Experience. Waveland Press; Prospect Heights, IL: 1999.
  • Schatz E. PhD. Dissertation, Graduate Groups in Sociology and Demography. University of Pennsylvania; 2002. Numbers and narratives:, making sense of gender and context in rural Malawi.
  • Schegloff EA, Sacks H. Opening up closings. Semiotica. 1973;7:289–327.
  • Setel PW. A Plague of Paradoxes: AIDS, Culture, and Demography in Northern Tanzania. University of Chicago Press; Chicago: 1999.
  • Sewell WH., Jr A theory of structure: duality, agency, and transformation. American Journal of Sociology. 1992;98:1–29.
  • Sewell WH., Jr Historical Event as Transformation of Structures: Inventing Revolution at the Bastille. Theory and Society. 1996;25:841–81.
  • Sewell WH., Jr . Logics of History: Social Theory and Social Transformation. University of Chicago Press; Chicago: 2005.
  • Smith DJ. Burials and belonging in Nigeria: rural-urban relations and social inequality in a aontemporary African ritual. American Anthropologist. 2004;106:569–579.
  • Spear T. Neo-traditionalism and the limits of invention in British colonial Africa. Journal of African History. 2003;44:3–27.
  • Spillman L. Nation and Commemoration: Creating National Identities in the United States and Australia. Cambridge University Press; Cambridge: 1997.
  • Stromberg PG. Language and Self-Transformation: A Study of the Christian Conversion Narrative. Cambridge University Press; Cambridge: 1993.
  • Swidler A. Culture in action: symbols and strategies. American Sociological Review. 1986;51:273–286.
  • Swidler A. Talk of Love: How Culture Matters. University of Chicago Press; Chicago: 2001.
  • Tarver H. PhD Dissertation. Department of Sociology, University of California; Berkeley: 1995. Words of war: the Persian Gulf crisis and American public discourse.
  • Tavory I, Swidler A. Condom semiotics: meaning and condom use in rural Malawi. American Sociological Review. 2009;74:171–189.
  • Tawfik L. PhD Dissertation, Bloomberg School of Public Health. Johns Hopkins University; 2003. Soap, sweetness, and revenge: patterns of sexual onset and partnerships amidst AIDS in rural southern Malawi.
  • Tawfik L, Watkins SC. Sex in Geneva, sex in Lilongwe and sex in Balaka. Social Science & Medicine. 2007;(Special Edition 64):1090–1101. [PubMed]
  • Vansina J. Paths in the Rainforests: Toward a History of Political Tradition in Equatorial Africa. University of Wisconsin Press; Madison: 1990.
  • Varenne H. Talk and real talk: the voices of silence and the voices of power in American family life. Cultural Anthropology. 1987;2:369–394.
  • Vološinov VN. Marxism and the Philosophy of Language. Seminar Press; New York: 1973.
  • Watkins SC. Navigating the HIV/AIDS epidemic in rural Malawi. Population and Development Review. 2004;30:673–705.
  • Weinreb AA. The limitations of stranger-interviewers in rural Kenya. American Sociological Review. 2006;71:1014–1039.
  • Whiteside A, Mattes R, Whillan S, Manning R. What people really believe about HIV/AIDS in southern Africa. In: Poku NK, Whiteside A, editors. The Political Economy of AIDS in Africa. Ashgate, Hants; England: 2004. pp. 127–149.
  • Williams TM, Kornblum W. Growing Up Poor. Lexington Books; Lexington, MA: 1985.
  • World Bank. Africa Lifts Its Silence on AIDS. World Bank; Geneva: 1999.
  • Wuthnow R. Meaning and Moral Order: Explorations in Cultural Analysis. University of California Press; Berkeley: 1987.
  • Zulu EM, Chepngeno G. Spousal communication about the risk of contracting HIV/AIDS in rural Malawi. Demographic Research, Special Collection. 2003. pp. 248–77.