When I Was a Swahili Woman: The Possibilities and Perils of “Going Native” in a Culture of Secrecy

Conversion

A large body of scholarship addresses questions related to insider–outsider status of ethnographers. Yet, as Na’amah Razon and Karen Ross (2012) point out, much of this literature suggests “that a researcher’s identity is fixed” and stable (p. 495), even among scholars, like Kirin Narayan (1993), who argue that one can be both insider and outsider, or those who argue that one’s identity may exist somewhere in between “insider” and “outsider,” at some point of hybridity on a continuum. In contrast, Razon and Ross argue for an understanding of identity as “fluid,” with different parts “negotiated, highlighted in some points, or discarded all together in particular moments of research” (p. 495). I find their perspective useful in understanding how my own presentation of self shifted throughout my time in the field as well as how participants’ assumptions about me affected my research. Unlike the participants quoted in Razon and Ross’s work, however, the Zanzibari women who participated in my research never asked me questions about myself. Rather, they seemed to take it for granted that I was, or wanted to be, a “good Muslim wife,” a phrase used synonymously with “Swahili woman” or “Zanzibari woman.”

Research has shown that when people include an ethnographer in their inside secrets, they may have an implicit hope that the ethnographer will “convert,” becoming an insider herself. Thus cultural anthropologist Susan Harding (2000) writes of “saved” fundamentalists who included her among their audience while ostensibly addressing other “unsaved” individuals, assuming that she would be equally transfigured by their narratives. In my case, Zanzibari women who shared their secrets with me had more than just a hope for my conversion. Before I married Hamid, I’d spent a month intensively reading about progressive and feminist interpretations of Islam. At the end of that month, I decided to convert, and by the time I met the women in Hamid’s family who would later participate in my research, I had been a Muslim for about four months.

In April 2011, I wrote the following retrospective account, edited slightly for clarity, of my September 2009 conversion:

After I’d decided to convert I asked Hamid what I needed to do, and he said he wasn’t sure. He suggested that I wait until I came in December for our wedding, and that Ustadhi [my Swahili father] would help me through the process. But after Umi [my Swahili mother] and Ustadhi found out, they called me. They asked when I was going to convert and I told them what Hamid had said. “No,” Umi said, “You should convert now so that by the time you get here in December you will be used to living as a Muslim.” “Okay,” I said, “but what do I need to do? Go to the mosque?” “No,” she said, “you can do it right now over the phone. Here’s Ustadhi.” Ustadhi and I greeted each other. “Uko tayari kusilimu? Are you ready to convert?” he asked me. “Ndiyo. Yes,” I answered. “Okay, repeat after me: Ashada ala ilaha ilallah, wa ashadu anna Muhammada rasulullah.” I repeated his words in Arabic. “Nashuhudia kwamba hakuna Mungu ila Mwenyezi Mungu ushahidi kwamba Muhammad ni Mtume wa Mwenyezi Mungu.” I repeated his words in Swahili. “I bear witness that there is no God but Allah; I bear witness that Muhammad is the Messenger of Allah.” I repeated his words in English. “Allahu akbar! God is great!” Ustadhi said. I could hear my host brother Abdulrahman in the background shouting, “Allahu akbar!” and Umi ululating. Ustadhi told me to seek out a Muslim teacher and promised to send me some links to Islamic websites. We hung up. I wondered what I had gotten myself into.

When I told this story to my archaeologist roommate during fieldwork in Zanzibar, she commented, “Oh, so converting is very connected to Zanzibar?” Her words reminded me of those of my friend Susan back in Los Angeles: when I first told her about Hamid and our plan to marry, she’d remarked, “It’s like you’re marrying Zanzibar!”

Almost as soon as I’d met Hamid in 2009, he’d asked me if I would be willing to convert. I suppose he was already trying to figure out if I was marriage material. Initially, I balked at this question, though I told him I was willing to respect Zanzibari Muslim norms without converting. He suggested that I could just tell people I’d converted even if I hadn’t really done so, but I laughed this off. After we were married, I asked Hamid why he’d suggested that I lie about converting. “I thought that if you acted like a Muslim, you’d come to believe, automatically,” he answered, evidencing a kind of Zanzibari theory of performative language and perlocutionary force (cf. Thompson 2017b). Because an estimated 99% of Zanzibaris are Muslims, most people I met had never known a convert, and had developed their own ideas of how conversion worked, independent of any real “data.” This lack of knowledge is also evident in my retrospective account above, where Hamid said he didn’t know what I needed to do to convert. In another case during my fieldwork, one day I was shopping for a new wedding ring for Hamid to replace the one he’d twisted and broken after we’d had a fight, when a salesman saw my headscarf and modest clothes and asked me if I was converting. I told him I already had two years before and he said, “Good. So you’re covering yourself like this now. Later you will cover more, eventually you’ll get a veil, until you conceal yourself nicely as expected.” He didn’t even say Inshallah (if God is willing), the Arabic expression used by Muslims the world over to express the unknowability of the future. To him it was inevitable that over time I would behave more and more as Zanzibari Muslims expected me to. He saw no room for my own agency.

Similarly, the adoptive family I’d called “my Swahili family” for over a decade and my new in-laws alike seemed to assume that becoming Swahili would follow naturally from my decision to become a Muslim. According to the Zanzibari women who offered me instruction, if I treated my husband the way they taught me to, I’d become one of them: “If you do this to him, he won’t get rid of you or even leave your side. He’ll tell you, ‘Wow, she’s a Zanzibari, this one! Where has she learned all these Zanzibari ways?’ You’ll be a Zanzibari, just like your husband.”

Becoming Swahili

I reread my fieldnotes from July 2011:

Sometime next week they will have an unyago for [Asha] where all the women in the family (including me) will help instruct her and help her get over shyness she may have before her wedding night. “Everyone contributes and debates,” Chanda said. “In other words, to debate the ‘acts,’” Baba [my father-in-law] said, looking at me knowingly and checking to see if I’d understood. “I already understood,” I told him. I reminded them that I received instruction at our wedding, only an abbreviated form because of the short time frame we had. “Yeah, you know the Western ways and the Swahili ways,” Chanda said, laughing.

Both East African and Western scholars have debated for decades what it means to be “Swahili.” Swahili scholar Ibrahim Noor Shariff (1970) writes that, for many mainland East Africans, any Muslim Swahili-speaker may be called Swahili. For him, however, “Swahiliness” (Uswahili) is not an ethnicity, but rather the traditions of coastal people who use Swahili as their “language of inheritance” (Rampton 1990). Shariff goes on to say that anyone who follows and lives by coastal customs and traditions is a Swahili. While “Swahiliness” is defined, at least in part, by the customs of those who use Swahili as a mother tongue, one can be(come) a Swahili person even if the language is not one’s own mother tongue, simply by following coastal norms—becoming “Swahilized” (Mazrui 2000, 75).

Following this expansive notion of Swahili identity, Shariff (1970, 11) argues that a person can be a Swahili and also an Indian, an Arab, a Baluchi (Iranians who settled in East Africa in the 1820s; see Lodhi 2013), a Comorian, and so on. Swahili self-understandings are in keeping with a social constructivist understanding of ethnicity as “an everyday phenomenon accomplished through the interactional routines that come to define . . . the business of everyday life, including the activities of social inquiry itself” (Best 2003, 897). From this perspective, Swahiliness “can be seen as an ongoing interactional achievement and not an attribute or essential property of any individual” (ibid).

Although I spoke Swahili with near-native proficiency, in the sixteen years since I first started learning the language it had never occurred to me I might become Swahili. As a Swahili learner, my “imagined community” (Kanno and Norton 2003, 246) had always been Swahili-teachers rather than a community of native speakers. Despite that Shariff (1970) and Madoshi (1971, 93) include Westerners and “non East Africans” among those who may become Swahili, it had always seemed preposterous to me. I’d never met an American that Zanzibaris considered Swahili, and even Shariff concedes their rarity. But, when combined with my language skills, my new identity as a Muslim convert and the wife of a Swahili man seemed, with little effort on my part, to open the possibility of Swahiliness. More and more, I found that Zanzibaris expected me to become, if not Swahili, at least Swahilized.

Was I falling prey to the ethnographer’s delusion that they share an identity with their interlocutors in the field? “Those we hope to study, our would-be consultants and interlocutors, may deny or reject the strands of common identity or allegiance we proffer, whether of co-residence, nationality, gender, or political sympathy. It remains easy for the more privileged (the educated, the wealthier, the lighter-skinned) to imagine, to project social and emotional ties with the marginalized, who may not want our sympathy and may actively resist our identification with them,” writes anthropologist Maria Lepowsky (2012, 72). Although I was occasionally guilty of this self-delusion—assuming, for example, that my Zanzibari “father” truly saw me as a daughter, but later discovering he did not—overall my experience was the opposite. The Zanzibari women I knew asserted Muslimness and marital status as a common identity we shared and assumed my allegiance to their values, while it was I who resisted such identification with them.

As an ethnographer, a sociocultural linguist, and a professor of Swahili, I had a strong sense of pride in my deep expertise in the language, an achievement I saw as a distinction from other researchers, even some other ethnographers I knew (cf. Gottlieb 2012a). The idea that I could become even more proficient in Kiswahili ndani, “deep Swahili,” through my induction into women’s secret language use appealed to this sense of pride. I began to take pleasure in being taken for Swahili, frequently writing about it in my fieldnotes, and increasingly aware of the access to interesting data I was gaining because of it.

Few researchers ever remain completely outsiders (Mullings 1999). Even if I’d wanted to do so, my Swahili in-laws and others pulled me in, encouraging me to dress, act, and speak more as a Swahili woman every day. Eighteen months into my marriage, I was leaving my in-laws’ home when a repairman passed me on the stairs. “Oh, you have a European!” he said to my father-in-law, laughing. “Asaalaam aleikum,” I greeted him. His face showed surprise that I used this Arabic-cum-Swahili expression. As he stood at the bottom of the landing, I felt his eyes look me up and down, taking in my modest blouse and long skirt sewn from kangas, and the scarf that covered my hair—all wedding gifts I’d received from Hamid’s family. “Or a Swahili?” he asked Baba, a little confused. “Nusu Mswahili,” I said—half Swahili. He and my father-in-law joined me in laughter. Despite my obviously European features, people’s reading of my presentation of self was becoming fluid (cf. Razon and Ross 2012).

When I went to visit Halima, my mother-in-law, she gave me a bunch of kangas, and told my brother-in-law Sudi that he needed to help me have them sewn quickly or I wouldn’t have anything to wear. My sister-in-law Munira said, “You’re such a Swahili! You only wear clothes made from kanga!” But my preference for such clothes seemed old-fashioned to some. One hundred years ago, Zanzibari women wore kanga to mark their Swahiliness and distinguish themselves from slaves (Boswell 2006; Croucher and Wynne-Jones 2006), but today a fashionable urban woman wears dresses sewn from synthetic fabrics and covers them with a black abaya in public. In my attempt to dress in “authentic” Swahili styles, I avoided styles that I perceived as imported from Saudi Arabia along with conservative interpretations of Islam. In a way, I was trying to be even more Swahili than the Zanzibari women I knew, who were themselves becoming Arabized. Affiliation is a moving target.

Going Native; or Becoming a Complete Member Researcher

It used to be that if a researcher “went native,” she was professionally doomed, assumed to have lost some necessary distance and objectivity and criticized for deluding herself that doing so was even possible (Turner 1993). Warning ethnographers not to “go native” is perhaps a holdover from anthropology’s links to earlier colonial projects, an extension of the European fear of colonials’ “losing their cultivated identities” (Burns 2017; cf. Ganter 2017). While the newer term “convert complete member researcher” may put a positive spin on the same phenomenon and attempt to avoid the links to colonialism (Adler and Adler 1987), ethnographers are still advised that to fully join the community in which one does research is to lose sight of one’s research objectives (Behar 1996), even to forget that one is a researcher (Luhrmann 2012). Perhaps because I was trained as a linguist and literary critic rather than an ethnographer, no one ever gave me this advice. But would I have listened if they had?

In sociologists Patricia Adler and Peter Adler’s (1987) terms, I was a mix between an “opportunistic” and a “convert” complete member researcher (CMR). On the one hand, like an opportunistic CMR, I “acquired intimate familiarity” with Zanzibari women through a life event—marriage into a Zanzibari family—and my group membership preceded my decision to conduct research among them (though I was already conducting research on other topics in Zanzibar; e.g., Thompson 2017b). On the other hand, like a convert CMR, I began my work on Zanzibari women with “a purely data-oriented research interest” in my topic but became “converted to complete immersion and membership during the course of the research” (Anderson 2006, 379).

More accurately, I would say that the Zanzibari women I knew took my conversion to Islam as evidence that we shared a belief system. Combined with my knowledge of Swahili and my marriage to a local man, this evidence suggested to them that we belonged to the same community. By showing them only what they wanted to see, and “backgrounding” information about myself that would have contradicted their view (Razon and Ross 2012), I performed a self that allowed them to treat me as one of them: dressing modestly, being quiet, and praying at the right times. But I concealed what I saw as stark differences between our understandings of Islam, gender, and marriage. Because I never talked with them about my own views and doubts, I was able to “pass” as a “good Muslim wife” (cf. Thompson 2011).

My experience speaks to humanist anthropologist Michael D. Jackson’s argument that the advice to an ethnographer “not to lose sight of his or her academic objectives by going native” relies on “culturally and historically determined distinctions that cannot be sustained in reality” (2012, 16–17). I could not be a good Swahili wife (or daughter or daughter-in-law) and sustain a distinction between my “own” culture and that of the Zanzibari women who welcomed me into their families and their instruction ceremonies. I found that taking up the role of complete member researcher through conversion opened up new possibilities, but also created unanticipated perils.

Going Public about Secrets

I am no longer a Swahili woman, if I ever was, and I can now reveal not only my cynicism about the advice Zanzibari women gave me and the extent to which I did not become the “good Muslim wife” they hoped I would, but also the real danger in which trying to do so put me. Ethnographers are meant to keep cynicism a secret: cultural relativism becomes ethical relativism. If it was difficult to admit to Zanzibaris that my marriage was failing and that I did not agree with their advice about how to save it, it has been a different kind of challenge to incorporate that failure and my cynicism into my ethnographic writing about Zanzibari Muslim women’s socialization through weddings and marriages (Thompson 2011, 2013, 2015a).

Just as becoming a Swahili woman requires socialization, so too does joining a community of academics. As Patricia Bizzell (1986) suggests, in order to master academic discourse, one must first “go native” in an academic community (Harris 1989, 16). Before I learned to be a Swahili woman, I learned to be an academic writer, to write in objective, stale prose, and to focus on words and texts rather than the people who used and created them. My membership in these two communities pushed and pulled on my secrets (cf. Jones 2011), forcing me to struggle with the balance between concealment and revelation and between my personal and my ethnographic self (Bruner 1993).

Claiming to have learned secrets in the field is one way that ethnographers position themselves as close to their interlocuters, while they may conceal the secrets themselves. “It is not uncommon for possessors of secrets to flaunt them, revealing the existence of concealment but not its object,” Jones (2011, 79) writes. We see this in a great deal of academic writing about the Swahili, when researchers claim they have been privy to secrets but can’t reveal them to their readers. Middleton (1992) writes:

Most of what I have gathered from conversations with many Swahili men and women has been considered by them as either confidential or as normally private talk, and I have not followed the fashionable habit of giving the names of the sources of every item of information. My “informants” were mostly friends and as such trusted me in their houses, and I learned much that I prefer not to put directly into print: I see no reason to embarrass those who were so generous with their time and knowledge, and without whom I could of course have done nothing. (xi)

In my case, I tended to do the opposite: through extensive quotation from premarital ceremonies that Zanzibari women allowed me to record, I revealed much of the insider knowledge that Zanzibari women shared with me about how a Muslim wife should talk and behave. What I concealed instead was my own secret: the impact of trying to live up to their expectations—staying silent and keeping secrets about my marriage—on my own well-being. In much of my early writing about Swahili marriage, I downplayed my personal experience. I used it as “data,” guiding and controlling the impression readers might form of me, hoping I might be seen as an expert on the anthropology of Swahili marriage rather than simply as the “good Muslim wife” my Zanzibari husband, friends, and family members hoped I would be. I sometimes presented the advice that Zanzibari women offered as typical advice for a bride—which it was—but without specifying that it was advice they offered me. In fact, I was at pains for my colleagues to know that I was not a good Muslim wife, seeing that identity as at odds with my feminism. At the same time, I minimized the extent to which Zanzibari expectations of a Muslim wife were negatively impacting my own mental and physical health. Traditional modes of impersonal academic writing served this project well, since my readers were not privy to actual interactions I had with Zanzibaris that might have contradicted, discredited, or otherwise thrown doubt upon the expert “ethnographic self” I projected (Bruner 1993, 1; cf. Goffman 1990, 23), and the Zanzibaris I knew did not know English well enough to read my work even if they had been interested in it.

Academic norms demand research be “finished, polished, and packaged,” but Goffman reminds us that the end product conceals “the feverish drudgery the author may have endured in order to complete the index on time, or . . . the squabbles he may have had with his publisher in order to increase the size of the first letter of his last name as it appears on the cover of his book” (1990, 53). Like Swahili women who present themselves as good Muslim wives and protect their husbands’ secrets, or like researchers who conceal aspects of themselves that would be considered culturally inappropriate in the field, most academic writers conceal not just the writing process but, in the case of ethnography, also our intimate, unfinished, messy, and painful experiences in the field. While the tide is turning on such personal revelations within anthropology, with the lasting influence of James Clifford and George Marcus’s Writing Culture ([1986] 2010), it is still evident in many neighboring disciplines that use ethnography, such as applied linguistics, the discipline closest to my own interdisciplinary work. Regardless of one’s discipline, ethnographers must grapple with how much of this secret “backstage” material to include in our work, how much of the ethnographic self to reveal. Consider Poulos’s (2008, 62) experience:

As I began my ethnographic journey a few years ago, memories, dreams, and secrets began to slip into my writing, first as hint and innuendo, as thin little threads of storied experience, as little uncontrolled vignettes creeping into my stories of communication in everyday life. At first, I purged these little fragments from my writing during the editing process. After all, I reasoned, even though they might be “real” or honest or important, they were only fragments anyway, and they might be, at best, self-serving and at worst, life damaging.

Similarly, anthropologist Virginia Dominguez (2012, 21) recalls that when she wrote about her fieldwork experience in Jerusalem,

My writing was no doubt too raw, its detailed documentation overwhelming and puzzling to someone in a calm, faraway place; the rawness of the feelings was probably even off-putting. It was neither journalistic writing nor “creative writing” let alone recognizable anthropological writing. But it was “alive.” It was troubled, and it communicated the frenzy and the impatience I saw in myself and in everything around me, and it was what I wanted to communicate to “the outside world.” I had a different presence, an embodied and only partly guarded one. It wasn’t packaged, formal, conventional, or restrained.

Autoethnographic writing foregrounds these “backstage” aspects of the research experience, “where the suppressed facts make an appearance” and “the impression fostered by the performance is knowingly contradicted” (Goffman 1990, 114) through the revelation of secrets, but the decision to foreground these aspects is still a fraught one (cf. Ellis 2008).

As I have moved into more personal modes of writing, mixing ethnography with creative nonfiction (Thompson 2017a) and even fiction (Thompson 2017c), I have tried to be more self-reflexive about my own agency in performing certain aspects of Swahili Muslim womanhood, and rejecting others (even while often concealing that rejection from Zanzibari women I feared would judge me harshly). Such self-reflection has forced me to reveal aspects of myself that I would normally keep hidden not only from Zanzibari friends and family members but also from my students, colleagues, and other academic readers.

In this article I have revealed some painful truths. Why should I disclose these secrets? Anthropologists Annette Leibing and Athena McClean (2007, 13) argue that “the ethnographer’s personal data are relevant only as they relate to, and help illuminate, the ethnographic process” and urge “discretion in sharing only what we must about our personal lives (while at the same time not holding back what needs to be examined) for the purposes of advancing knowledge for ourselves, our interlocutors, and our readers.” While I agree with their view in theory, in practice it is more difficult to know whether one’s “personal data” are helpful or excessive, better shared or held back. What does my personal experience illuminate about, and for, Zanzibari women? To what extent can I use my personal experience to reveal truths about those of others? While cautious to not shift the ethnographic focus too far toward my own experience and away from those of Zanzibari women, I continue to seek ways of revealing, both more effectively and affectively, the dangerous consequences of performing Swahili womanhood.

I am not a Zanzibari woman and my experiences differed from those of the women I researched in many important ways. While I and they were both silenced and expected to conceal ourselves, I want to be careful not to condemn silence or concealment per se; I agree with Blake Poland and Ann Pederson (1998) that a general emphasis on speech and treatment of silence as a problem reflects “Western ways of knowing” (p. 297). We must remember, following Goodall (2005), that “not all secrets are necessarily harmful” (p. 503). Indeed, Zanzibari women’s advice for one another reflects their belief that it is not secret-keeping that is harmful, but rather revelation. Revealing personal secrets bears risks for Zanzibari women: risks of public shame, of divorce and the attendant financial risks, and of accumulating sins. Keeping silent about abuse also protects one from the stigma of expressing grief, allowing one to “pass” as happy, or perhaps even to performatively do happiness. “A story told is a powerful thing that can unleash all sorts of grief; an untold story gives off at least the illusion of control” (Poulos 2008, 51). Silence or self-censorship about marital problems is strategic, offering some protection from these risks and this grief, some illusion of control.

Yet my experience and its impact on my research does reveal something interesting not only about my own ethnographic practice but also about Zanzibari women’s own weddings, marriages, and secrets. After all, the very questions I asked Halima, Chanda, Umi, Asha, and others were influenced by my experience, and no doubt their answers were influenced by their understanding of me as a new Muslim who appeared to be trying to meet their expectations. My experience also reveals that Swahili womanhood is performative: not just performed but rather “produced in the performance” (Pennycook 2007, 58). The advice that Zanzibari women gave me about how to sustain my marriage was of the same sort they gave to their own female relatives and friends, and when they performed the roles of “ideal wives” in my presence they did so for one another as well as for me. If I was performing the good wife that they wanted me to be, so too were they. For both them and me, the “performance of self” was performative in two senses, not only a “performance,” an acting out of what they expected of one another and of me, but also in the sense described by J. L. Austin (1962) for linguistics and later elaborated by Judith Butler (1990) for understanding gender (cf. Pennycook 2007): by performing Swahiliness, Zanzibari women iteratively enacted and (re)created it. Likewise, so did I, creating a Swahili self through interaction with them.

As Goffman says of many enactments, my performance of Swahili womanhood and access to theirs would not have been possible “had not tasks been done which were physically unclean, semi-illegal, cruel, and degrading,” facts that I did not express in more traditionally academic forms of writing but now explore through autoethnography and other forms of literary ethnography. This is the “dirty work” (both mine and that of Zanzibari women) that would remain hidden from a conventional ethnography of Swahili marriage (Goffman 1990, 53).

At the same time, I remain wary of academic reactions to my revelations. These fears are not only tied to the shame I felt at being a feminist in an abusive marriage, but also about writing about violence as tied to culture. When Hamid broke my finger early in my pregnancy, I chose not to call the police because I was afraid he would be deported. My fears of airing Zanzibari “dirty laundry” feels similar. In some ways, doing so risks putting Zanzibaris—or, perhaps even more problematically, Muslims generally—in a negative light. I certainly do not mean to suggest that all Zanzibari women suffer in unhappy marriages or from abusive husbands. Rather, my experience shows that the silence into which Zanzibari women socialize one another creates an environment in which abuse, if present, can remain hidden and do great damage. We know that divorce rates are high on the Swahili coast, but the extent to which domestic violence occurs is difficult to measure precisely because of the high value placed on secrecy.

My decision to talk and write about the “dirty work” of my Zanzibari marriage is an explicit rejection of what I was taught by Swahili women. While they advised me—as they would advise any Muslim bride—to stay silent about my husband’s and my own shameful behaviors, I have refused to do so in my ethnographic fiction (Thompson 2015b, 2017c) and autoethnographic writing (Thompson 2017a). Storying my experience has not only been healing for me but has also felt ethically necessary (cf. Poulos 2008, 53). It felt dishonest to write about Zanzibari women’s experiences of premarital instruction, marriage, and divorce without also writing about the effects of Zanzibari expectations of women on my own experience.

Conclusion

My experience undergoing religious conversion, marrying cross-culturally, and “going native” while conducting research in a culture that highly values secrecy may be a singular case, but what can it tell us about broader issues? Combining “autoethnographic case study” (Duncan 2004, 9) with ethnographic research allows me to contextualize the behaviors and values into which I was socialized with reference to the cultural basis for them (Duff 2008)—both Zanzibari Muslim values and (American) academic ones.

The instructions that Zanzibaris gave me about how to become Swahili, their assumption that I would inevitably do so, and their eventual treatment of me as if I had adds evidence to earlier claims that Swahiliness is not a bounded category; rather, it emerges through not only large-scale cultural interaction but also interpersonal communication. It also throws a wrench into debates about whether a researcher can or should “go native,” demonstrating that sometimes one has little choice in the matter. Finally, it sheds light on the potential ramifications of Zanzibari women’s secret-keeping for their own well-being. Secrecy makes it difficult to conduct research on Zanzibari women’s own use of the teachings they offer one another, and the effects of these teachings on their own marriages, but my experience illustrates at least one possible outcome: While not all Zanzibari marriages are abusive, women’s encouragement of one another to stay silent about their marriages creates a situation in which abuse, if present, can continue.

The parallels between Zanzibari silencing of women about their marriages and academic silencing of scholars about personal issues are many. Both are matters of socialization and both are performative: We learn what to stay silent about, and in doing so we conceive the self in writing (Suleiman 1994, 67). On the one hand, the rarity of autoethnographic and other personal forms of writing in academia contributes to the silencing of negative fieldwork experiences including sexual harassment and rape (Hanson and Richards 2017), in part by socializing younger scholars into silence. What we leave out of our “tales from the field” matters. On the other hand, if we were socialized into leaving personal experiences out of our writing, we know that such socialization does not always result in “culturally predictable outcomes” but can also “occur in ways not predicted or desired”; some of us become “bad subjects” (Kulick and Schieffelin 2004, 354–55). We can resist our socialization, as I did in Zanzibar (and I suspect some Zanzibari women do too) and now try to do in writing about my experience. When graduate students have read work in which I used a literary style and shared some of my own story, they have reacted with surprise and pleasure, asking, “Are we allowed to write like this?” Our writing also socializes others.

To give up on doing research in Zanzibar after my marriage ended was a painful but necessary decision. Writing about this decision publicly is no less painful but perhaps less obviously necessary. My public acknowledgement, in print, of this painful decision contributes to a growing understanding of how personal relationships and emotions impact ethnographic fieldwork and writing. While Gottlieb (2012a, 15) has made the important argument that the choice of a field site is a “human decision . . . limned by structures of experience, . . . our own embeddedness in structures of kinship and marriage, and, lest we forget that most personal but often neglected of all factors, structures of feeling,” my own experience demonstrates that such structures impact far more than that initial choice: the entire fieldwork experience, the decision to abandon a fieldsite altogether, and, finally, the decision to write about it.