104 + FORMS OF INTERVIEWING In-Depth Interviewing + 105 @ In-Depth Interviewing as a Social Form In-depth interviewing involves a certain style of social and interpersonal interac- tion. As a social form, it differs from the kinds of interactions one usually finds in sales pitches, public lectures, job inter- views, counseling sessions, sexual pickups, board meetings, monologues, or marital conflicts. To be effective and useful, in- depth interviews develop and build on inti- macy; in this respect, they resemble the forms of talking one finds among close friends. They resemble friendship, and they may even lead to long-term friendship. But in-depth interviews are also very different from the kind of talking one finds between friends, mainly because the interviewer seeks to use the information obtained in the interaction for some other purpose. A researcher who uses in-depth inter- viewing commonly seeks "deep" informa- tion and knowledge-usually deeper infor- mation and knowledge than is sought in surveys, informal interviewing, or focus groups, for example. This information usu- ally concerns very personal matters, such as an individual's self, lived experience, values and decisions, occupational ideology, cul- tural knowledge, or perspective. When two close or "best" friends talk, there is no prag- matic purpose that transcends the friend- ship itself. That kind of talk is an end in it- self. But when an in-depth interviewer talks to an informant, the goal is to collect data. Some specific ethical issues arise because of this difference. In-depth interviews rarely constitute the sole source of data in research. More com- monly, they are used in conjunction with data gathered through such avenues as lived experience of the interviewer as a member or participant in what is being studied, naturalistic or direct observation, informal interviewing, documentary re- cords, and team field research. In many cases, researchers use in-depth interview- ing as a way to check out theories they have formulated through naturalistic observa- tion, to verify independently (or triangu- late) knowledge they have gained through participation as members of particular cul- tural settings, or to explore multiple mean- ings of or perspectives on some actions, events, or settings. This was true in the fa- mous case of anthropologist Margaret Mead's Coming of Age in Samoa ([I9281 1960).Mead supplemented her field expe- rience, direct or naturalistic observations, and other interviews with in-depth inter- views with informants (see Atkinson and Coffee, Chapter 38, this volume). Years later, in his reexamination of Mead's research in Margaret Mead and Sa- moa (1983), however, Derek Freeman raised serious questions about these inter- views. Freeman argues that Mead was mis- led by her female adolescent informants, even though she had lived in the Samoan villages for many months, because of the Samoan suspicion of outsiders and other contextual features of their recent contacts with Westerners. When Freeman later in- terviewed some of the same women who had been Mead's adolescent informants, they told him that they had told Mead what they thought she wanted to hear. Another well-known case is represented by the studies of the Mexican village of Tepotzlan done over several decades by two different researchers, Robert Redfield (1930, 1941, 1960) and Oscar Lewis (1951). Redfield and Lewis made different inferences from their observations and in- terviews and drew diverse conclusions from their lengthy experiences of living in this small, remote village. Each justified and legitimated what he reported by refer- ring to his heroic fieldwork and interviews. The two men's findings were different, however, in large part because of certain basic assumptions each made about the na- ture of "conflict" and "shared meanings" (or consensus) in everyday life, assump- tions that predated their research observa- tions and experiences. Another heated debate arose recently concerning what is arguably the most fa- mous sociological ethnography of all time, William Foote Whyte's Street Corner Soci- ety (1943, 1955, 1981, 1993), the classic study of an Italian American community in north Boston. Whyte also utilized in-depth interviewing in his study to supplement and complement other forms of data collection. Years later, W A. Marianne Boelen (1992) returned to north Boston and reinter- viewed virtually all of the people Whyte had interviewed; she then contested the truth of what he had reported (see Whyte 1992; Vidich 1992; Denzin 1992). All of the studies noted above involved multiple research methodologies in addi- tion to in-depth interviews. They illustrate that each research project involves the ob- server or interviewer as an active sense maker and interpreter of what is seen or heard in the research context. Each inevita- bly depends on the researcher's own stand- point and place in the community, aswell as his or her own self-understandings, reflec- tions, sincerity, authenticity, honesty, and integrity. Whether in-depth interviewing should be used in research depends on the nature of the research question. Achieving clarity in the formulation and articulation of the research question commonly enhances the clarity of the methodological goals and ob- jectives. If one is interested in an area of study in which the information sought is relatively limited, such as the marketing choices of individuals, then there is every reason to think that the use of focus groups or fixed-choice questionnaires might be ap- propriate. If one is interested in under- standing forms of urban sociation, then di- rect observation would seem to be a reasonable approach to gathering data. But if one is interested in questions of greater depth, where the knowledge sought is often taken for granted and not readily articu- lated by most members, where the research question involves highly conflicted emo- tions, where different individuals or groups involved in the same line of activity have complicated, multiple perspectives on some phenomenon, then in-depth inter- viewing is likely the best approach, despite its known imperfections. In-depth inter- viewing is often a very appropriate method to use in qualitative research (see Warren, Chapter 4, this volume), life story research (see Atkinson, Chapter 6, this volume), the gathering of personal narratives (see Riessman, Chapter 33, this volume) and oral histories (see Ciindida Smith, Chapter 34, this volume), and the use of grounded theory methodology to analyze the ac- counts of members of some social setting (see Charmaz, Chapter 32, this volume). The important point is this: The nature of the research question determines whether or not the use of in-depth interviewing is advisable. + The Goals and Purposes of ha-Depth Interviewing Many talented researchers have analyzed in-depth interviewing as a method or tech- nique of collecting data (see Atkinson 1998; Cicourel 1964; Denzin 1989a, 1989b; Douglas 1985; Fontana and Frey 1994; Geertz 1988; Holstein and Gubrium 1995; Lofland and Lofland 1984, 1995; Merton, Fiske, and Kendall 1956; Rubin and Rubin 1995; Spradley 1979; Wax 1971).Many authors have taken up the is- sue of "how to do" qualitative or in-depth interviewing, and most additionally affirm the importance of the researcher's goals and purposes, the researcher's moral com- mitment to seek out what istrue, and the re- searcher's ethical imperative to examine his or her own personal ideas, occupational ideologies, assumptions, common sense, and emotions as crucial resources for what he or she "sees7' or "hears" in a particular research interview or project. Many reflective parents have had to learn this important lesson from their chil- 108 + FORMS OF INTERVIEWING In-Depth Interviewing + 109 search," provides many examples of indi- viduals who have conducted such research (see also Higgins and Johnson 1988). Lofland and Lofland (1995) advocate the advantages of "starting where you are," by which they mean potential researchers should seriously consider studying those social phenomena to which they have ready or advantaged access. Some of my own first research experiences and in-depth inter- views fit this pattern. Making use of my knowledge and membership as a former U.S. Navy officer, I conducted in-depth in- terviews to explore others' perceptions and knowledge of routine bureaucratic record- keeping activities and "gundecking" (fudg- ing) of official reports (Johnson 1972, 1980a, 1980b). Later, active participation in the battered women's shelter movement as a founder and worker in a shelter pro- vided me with a foundation from which to conduct in-depth interviews with battered women about their experiences with do- mestic violence (Adhikari, Reinhard, and Johnson 1993; Ferraro and Johnson 1983; Johnson 1981, 1985, 1992; Johnson and Ferraro 1984; Johnson, Luna, and Stein forthcoming). And even today, my years of activist participation as a death penalty ab- olitionist serve to inform my work as I con- duct in-depth interviews with respondents who hold multiple perspectives on these ac- tions. When charting a research project that includes in-depth interviewing, is it better to be an experienced veteran or a relatively ignorant novice? Each status has its strengths and advantages, and each its pit- falls and dangers. Novices are less inclined to possess hardened assumptions about what they are studying, but they often have more difficulty seeing the nuances or lay- ered meanings of participating members. When undertaking a research project through in-depth interviews, they are likely to have a longer learning curve. Veterans with actual lived experience may already possess member knowledge, but they may also take that knowledge for granted. Addi- tionally, their current or former status as members may constitute a barrier when they interview others. It is important that researchers recognize these nuances in ad- vance, so that they can undertake the plan- ning of in-depth interviewing in a manner that will help them to assess these influ- ences on the accounts and reflections col- lected during the interviewing process. Whether the researcher is a neophyte or a returning veteran, in-depth interviewing involves an interactive process in which both interviewer and informant draw upon and use their commonsense knowledge to create some intelligible sense of the ques- tions posed and the ensuing discussions about them. e Locating Informants Planning and preparation are essential for successful in-depth interviews, but few re- searchers do everything they think they should before beginning them. Hardly any- one reads everything he or she feels should be read or achieves the kind of clarity he or she really wants on the protocol of ques- tions. In their recent work on interviewing, Herbert Rubin and Irene Rubin (1995:42) liken the planning for an interviewing ses- sion to planning for a vacation-that is, making plans sufficient to meet practical and emotional expectations while at the same time providing for the possibility of "hanging loose," or altering the course of the interview to go where the informant wants to lead. At some point, the researcher must make a leap of faith and just dive into the process. The research process is a learning pro- cess. Interviewers make mistakes; they make gaffs and alienate informants. They learn that their race, age, gender, social class, appearance, and even achieved sta- tuses make one kind of difference with some informants and another kind of dif- ference with other informants (see in this volume Schwalbe and Wolkomir, Chapter 10; Reinharz and Chase, Chapter 11; Dunbar et al., Chapter 14). The point is that researchers can learn from all this-learn what makes a difference for their specific projects, learn their strengths and how to play to them, and how to cover or compensate for their weaknesses. Individuals have performed the basic forms of asking questions and answering questions countless times before they ever come to their first formal, in-depth inter- views. The role of informant is part of the cultural stock of commonsense knowledge for the vast majority of children and adults. As friends, we talk in an informal manner and engage in cooperative, mutual self- disclosure. Those who elect to conduct re- search in a more formal fashion draw upon and build upon these cultural forms and commonsense practices. When a researcher begins an in-depth interview, he or she be- haves in a friendly and interested manner so as to help build trust and good rapport. An in-depth interviewer begins slowly, with small talk (chitchat),explains the pur- poses of the research, and commonly be- gins with simple planned questions (often referred to as icebreakers) that are intended to "get the ball rolling" but not to move so quickly into the issues of the key interview questions as to jeopardize intimate self- disclosure (or trust). Good rapport is sig- naled by emotions that feel harmonious and cooperative, and trust can commonly be discerned through eye contact, facial ex- pression, and bodily idiom. In-depth interviewingdiffers from other forms because it involves a greater involve- ment of the interviewer's self. To progres- sively and incrementally build a mutual sense of cooperative self-disclosure and trust, the interviewer must offer some form of strict or complementary reciprocity. Strict reciprocity is possible only if the in- terviewer is a former or current member of the group under study, and would take the form of the interviewer's sharing with the informant his or her own views, feelings, or reflections on the topics being discussed. It is more common for an interviewer to bring some form of complementary reci- procity to the informant-not a strict exchange of perceptions, feelings, or reflec- tions, but rather some form of help, assis- tance, or other form of information. When I interviewed the women who came to the battered women's shelter that I helped es- tablish in the late 1970s, for example, I could hardly offer them strict reciprocity for their views on battering, given that Iwas not a battered woman. Rather, I could share with them what many other women had told me they felt and said about their similar circumstances; after a while, I could even offer well-grounded advice on what they might do next (Ferraro and Johnson 1983;Johnson and Ferraro 1984).I did the same in two subsequent interview studies on the effectiveness of domestic violence protection orders (Adhikari et al. 1993; Johnson et al. forthcoming). In my current interviews with male stalkers, I cannot of- fer my informants the solace of strict reci- procity, given that I have never stalked any- one myself. I can, however, share the wisdom I have culled from working in the field of domestic violence for 30 years, in- cluding almost two decades of work and counseling with violent men. In order to conduct in-depth interview- ing, then, researchers must undertake con- siderable self-reflection to get to know themselves; they must also make a self-con- scious effort to observe themselves in inter- action with others. The development and cultivation of trust with informants is slow, incremental, and emotional, in most cases, and the relationship can change quickly (Johnson 1975). The ideal goal is that the informant become a collaborative partner with the researcher in the intellectual ad- venture at hand. Gender is inevitably important in inter- viewing, but it is difficult to generalize about the precise nature of its importance. The nature of the research question is com- monly the main issue. Some research ques- tions may elicit responses or perspectives for which gender has great relevance, whereas others may not. Feminist scholars such as Carol Gilligan (1982) assert that 110 + FORMS OF INTERVIEWING In-Depth Interviewing 1) 111 many researchers interpret women's re- sponses according to male standards (hier- archy, individualization, rationality) while neglecting women's relatively greater uses of relational categories and perspectives. Dorothy Smith (1987, 1990) notes that prevailing institutional priorities and agen- das often devalue women's lived experi- ences in the world, and that the very formu- lation of the questions that animate a research project often implicitly contain hidden gender evaluations or perspectives. She proposes that researchers place the is- sue of women's daily lived experiences at the center of the research process itself. All researchers would be wise to develop a spe- cial sensitivity to the explicit or deeply ob- scured meanings of gender in any particular research topic. The process of locating informants is simplified if members of the group of inter- est are usually or regularly located at the same place or scene; it is more complicated if potential informants do not regularly congregate at one locale. All those persons who are members of some scene or commu- nity, or who participate in some activity, are not equally valuable as informants. Infor- mants differ greatly in their intelligence, knowledge, and ability to reflect. Infor- mants also differ in their motivations to as- sist in or cooperate with an in-depth inter- view or series of interviews. Informants differ widely in their responses to specific individuals, whether because of racial, class, gender, age, or other characteristics, or perhaps just because of timing. It is real- istic for the researcher to anticipate that this will happen. Because those who do in-depth interviews for research purposes have no interest in "counting" them or "adding them up," this reality of noncom- parable interviews poses no problem. Many research projects have been "made" by the researcher's finding that rare, reflective inside informantwho seems to know just about everything that seems to be important and has thought about it and reflected on it forsome considerable period of time before he or she ever meets an eth- nographer or does an in-depth interview. Legendary examples include "Doc" (Dean Pecci), William Foote Whyte's key infor- mant in his research for Street Corner Soci- ety (1943, 1955, 1981, 1993); "Tally," Elliot Liebow's key informant for Tally's Corner (1967);and "Vincent Swaggi," Carl Klockars's key informant for The Profes- sional Fence (1974). The kinds or types of individuals who are likely to become key informants like this can be found in many settings. They are often marginal to the set- ting or scene being studied and are often seen by others in the setting as "lay intellec- tuals," thinkers, eggheads, or know-it-alls. Sometimes they are the politically ambi- tious individuals in the setting, those who have strenuously studied the setting and its personnel for the purposes of occupational or material gain or advancement. Some- times they are the "outsiders" of the setting, stigmatized for some quality that is depreci- ated or deprecated. Ethnographers and interviewers should always develop an awareness of such indi- viduals and be ready to cultivate their trust and friendship for the purposes of gaining member knowledge. Marginal membership status in the setting or activity seems to pro- vide many with an invitation to reflection and usually a certain sense of intellectual detachment from the "official line" among the membership. Finding such individuals and making them collaborators in the re- search process can yield wonderful results. Researchers should take care, however, to check out the observations and reflections of such individuals by getting independent verification through other interviews, if and when possible. Researchers who fail to do such checking can jeopardize the integ- rity of their research findings and possibly their own reputations. Some informants are better than others. Not all members of a setting or community are equally valuable for purposes of in-depth interviews. Not all of those who participate in some activity have a sufficient motive or interest to be interviewed about it. The best informants are those who have been thoroughly enculturated in the setting or community, have recent membership participation, have some provisional inter- est in assistingthe interviewer, and have ad- equate time and resources to take part in the interviews. The best informants are those who can describe a scene or setting or activity, those who can provide "thick de- scription," as Clifford Geertz (1973, 1988) terms it, but not necessarily those who ana- lyze or theorize. In some settings or situa- tions, such individuals may "click" with the interviewer or they may not-this is inevi- table. The issue of "sampling," or how re- searchers decided which informants to in- clude and which to exclude, is one that is rarely addressed in research reports and publications. It is important for researchers to provide accounts or explanations of how this selection was done in specific projects, so that readers may assess the researchers' findings (Altheide and Johnson 1994: 494-95). + ConductingIn-Depth Interviews The act of conducting the first in-depth interviews on a new study is often tinged with anxiety but also great anticipation and excitement. The first interviews usually yield great leaps forward in learning. The learning curve is steep at this point. It isbest for the interviewer to begin with an actual protocol of questions: usually two or three introductory icebreakers to get the ball rolling; several transition questions, which may again explain the purposes of the inter- viewing project or elicit permission from the respondent to use a tape recorder; and then perhaps five to eight main or key ques- tions that address the heart or essence of the research question(s). An in-depth inter- view commonly concludes with the inter- viewer summarizing some of the main points he or she has understood or giving the informant some information about what others have said about the issues dis- cussed.Although interviewers might antici- pate following such a nice, neat, rational plan before they begin interviewing, they inevitably find that the path, tone, and tra- jectory of actual interviews rarely follow this sequence. As an interview progresses, it often takes unexpected turns or digressions that follow the informant's interests or knowledge. Such digressions or diversions are likely to be very productive, so the interviewer should be prepared to depart from his or her prepared plan and "go with the flown- that is, consider following for a while where the informant wants to lead. It is es- sential that the interviewer be assertive enough to return the interview to its antici- pated course when necessary, but not so rigid as to preclude his or her learning un- expected information. Go with the flow, be playful, and be open to an experimental attitude-these are all good pieces of ad- vice for a novice in-depth interviewer in the early stages of a project. USING THE TAPE RECORDER TO LEARN INTERVIEWING SKILLS We now know with some certainty that a human being's individual memory does not remember what the person sees or hears, but rather organizes it into some intelligible coherence based on the individual's past experience. Thus it is essential that inter- viewers tape-record in-depth interviews to obtain verbatim records of those inter- views. Handwritten field notes are impor- tant for any research project, and there ex- ists considerable wisdom about how to make such notes (Emerson, Fretz, and Shaw 1995; Lofland and Lofland 1995; Strauss and Corbin 1990), but field notes are far inferior to tape recording for in-depth interviews. One of the main goals of qualitative re- search has always been to capture the words and perceptions of informants, or, as Bronislaw Malinowski (1922) puts it, "to grasp the native's point of view, his relation 112 s FORMS OF INTERVIEWING In-Depth Interviewing * 113 to life, to realize his visions of his world" (p. 25). So obtaining a verbatim record is the ideal if the subsequent analysis is to be valid and meaningful. Whether or not the researcher tape-records an interview, it is imperative that he or she take process notes regarding the interview itself, to gain an un- derstanding of the interview as a social oc- casion and how the questions and answers mutually constitute the sense of what is said. The questions asked guide and influ- ence the answers given, and so it is impor- tant for the interviewer to grasp why the in- formant proffers one segment of talk as an answer rather than another. Researchers can develop and cultivate the skills needed for in-depth interviewing with practice. Although in-depth inter- viewing is perhaps the form of interviewing closest to the lzind of talking done between friends, the individual who conducts an in-depth interview exercises greater con- trol over the flow and tone of the conversa- tion than does the respondent. The begin- ning of the interview is different from the beginning of a conversation between friends in that the interviewer comlnonly explains the purposes of the research and, these days, perhaps gets the informant's sig- nature on an informed consent statement. The turn talzing is also different from that in a conversation between friends, with the interviewer deferring to the informant. The asking and answering of questions is asymmetrical, with the interviewer having previously prepared a protocol of ques- tions and the will to keep the informant on track, attending to the business at hand. The interviewer is more passive in the role of listener, and, if the interviewer is success- ful, the inforinant is more active as a speaker. During interviews, the rules for pausing are usually different from those in talks between friends, as are the rules for physical proximity. The interviewer's aim is to develop progressively with the infor- mant the kind of mutual and cooperative self-disclosure that is associated with the building of intimacy and trust, but it takes great skill to accomplish this when one is working with asymmetrical communica- tion norms very dissimilar to those one usually associates with building intimacy and trust, as in actual friendship. The inter- viewer's goal is to solicit the informant as a collaborative partner in the sense making and interpretations that flow from the in- terviewing process. USING INTERVIEWS TO EXPLORE VERSUS USING INTERVIEWS TO VERIFY In the early stages of a research project, the in-depth interviewer may feel relatively ignorant about what he or she is studying. After several interviews, however, the in- terviewer begins to build a stock of knowl- edge about the research questions, and in most cases feeds some of this information back to the informants in subsequent inter- views, after those same questions have been covered. This information exchange be- comes part of the complementary reciproc- ity so necessary to the continued building of intimacy, and it also begins the process of verification in the research process. Data collection and verification become inextri- cably intertwined in most in-depth inter- viewing projects. As the research develops, the interviewer should keep and review his or her own jottings and notes (see Emerson et al. 1995; Lofland and Lofland 1995)and should review prior interviews when possi- ble, or when transcripts become available, and should begin progressively to focus the nature of the questioning and probing in later interviews. The later interviews of an in-depth interviewing project are usually more focused on specific probes and verifi- cation of what has been learned in earlier interviews. In more traditional or standardized in- terviewing, interviewers are commonly told to stick to the questions on the re- search protocol, to ask the questions pre- cisely as they are given, to probe for clarifi- cations only in ways that will not influence the respondents' answers, and to record only what the respondents say (see, for ex- ample, Singleton and Straits, Chapter 3, this volume). Further, traditional inter- viewers are trained to be impersonal; that is, they are trained to avoid offering any kind of information or revelations about any of their own values, beliefs, or opinions that might influence respondents in any way (see, for example, Fowler and Mangione 1990). This is not a realistic ideal for in-depth interviewing, because the nature of the research question itself usu- ally entails a deeper process of mutual self-disclosure and trust building. Skilled in-depth interviewers may often deviate from the research protocol, to go where the informant seems to want to go or perhaps to follow what appear to be more interesting leads. The interviewer should record these moves in his or her process notes, so that he or she can see later how one set of interviewing actions influenced and thereby constituted what the infor- mant said. The interviewer can use subse- quent interviews with the same informant or other interviews with additional infor- mants to check the interpretive validity of this strategy. s The Life Cycle of In-Depth Interviewing Excitement runs high when an interviewer is in the springtime of a research project. Genuine students are usually enthusiastic about gaining new knowledge from infor- mants and learning what they have to teach. Eventually, however, the excitement begins to wane. The doldrums of the sum- mer monsoons appear. The animating en- thusiasm begins to lessen, and researchers find themselves using all sorts of excuses, rationalizations, and self-deceptions to al- ter their involvement with the research in- terviews. In some cases, boredoin appears. 0 curveThis happens because the learnin, has ~eaked,and it is less satisfying to do all of the pragmatic work required to set up in- terviews when one learns progressively less from them. Barney Glaser and Anselm Strauss (1967:120-45) refer to this as the "saturation point" of a research project. It is commonly in this context that the re- searcher begins to ask, How many inter- views are needed? How many interviews are enough? Interestingly, the academic literature on interviewing includes various answers to the question of how many interviews are needed. James Spradley (1979:51), an an- thropologist usually interested in using in- terviews to understand cultural forms and members' perspectives, has noted, for ex- ample, that for him, one in-depth interview commonly involves six or seven one-hour sessions, and a given research project might include between 25 and 30 of these. Grant McCraclten (1988:37), a researcher with a business background who uses in-depth in- terviews (which he terms "long inter- views") to gain knowledge about marketing and business questions, says eight such in- terviews are usually enough. The progeni- tors of grounded theory methodology in qualitative research, Glaser and Strauss (1967), do not recommend a specific num- ber of interviews or observations, but say that the researcher should continue until a state of theoretical saturation is achieved; the identification of this point, however, is left ambiguous in their writings on this is- sue. Many others have shared their opin- ions on this question, but as the researchers cited above illustrate, there is no specific, set answer. The number of interviews needed to ex- plore a given research question depends on the nature of that question and the kind or type of knowledge the interviewer seeks. To those students who have asked me how many interviews they need, I have often re- sponded, "Enough." By this I mean that enough interviews must be conducted so that the interviewer feels he or she has learned all there is to be learned from the interviews and has checked out those un- derstandings by reinterviewing the most trusted and most knowledgeable infor- mants. 114 * FORMS OF INTERVIEWING In-Depth Interviewing + 115 It has been a common ideal in in-depth interviewing for the interviewer to check out his or her understandings with one or more key informants since this practice was first articulated and reported by William Foote Whyte (1943:279-358); this is usu- ally called the "member's test of validity." In research that uses interviewing as a basic form of data collection, whether the re- searcher is a neophyte or a returning mem- ber, early interviews will embody much more "grand tour" questioning (Lofland and Lofland 1984:78-86; Spradley 1979: 86-92) than will later interviews, which tend to be more focused on checking out and verifying research observations, analy- ses, and presumptive findings. In a very important sense, all research is "team research" in that it occurs in social, interactional, and community contexts. Even in the case of the heroic "lone ranger," the individual who is for the most part working on his or her own out in the field, there is usually a social support system of family members and friends and a small co- terie of professional colleagues who pro- vide intellectual and social support for the project. Researchers usually acknowledge such ties in the introductions, prefaces, or notes of the reports they publish on their studies. In other cases, interviewers may work in teams on projects with other re- searchers and share the interviewing duties. The interpersonal dynamics among re- search team members can be a source of problems, from the beginning negotiations concerning the "research bargain" (the di- vision of labor and reward) to the eventual analysis and report. Members of an inter- viewing team may feel violated or "ripped off" just as informants may feel violated or "ripped off" if their confidentiality is breached or if promises are not kept (Adler, Adler, and Rochford 1986; Douglas 1976). In one of the extensive team research projects on which I worked, proprietary rights to the interviewing records were specified in a divorce agreement. In addition to the social relationships implicated in and by a particular research project, research reports claim membership in some kind of interpretive community. They do this through the idiom, language, and issues that they embody. Qualitative re- search is a diverse and multifaceted field. The editors of the Handbook of Qualitative Research, Norman Denzin and Yvonna Lin- coln (2000a),identify "seven moments" of qualitative research; in their recent work, Jaber Gubrium and James Holstein (1997) identify four major "idioms" of qualitative method. However one classifies qualitative research communities, each implicates its own standards of acceptable and reportable truth. Researchers would be wise to make their connections to particular research communities explicit and to incorporate these into their research processes and re- porting, so that competent readers may as- sess how standards were created and em- bodied in actual research situations. e Ethical Issues Raised by IE-DepthIaterviewing In-depth interviewing commonly elicits highly personal information about specific individuals, perhaps even about the inter- viewer. This information may include par- ticipants' personal feelings and reflections aswell as their perceptions of others. It may include details about deviant or illegal ac- tivities that, if made known, would have deleterious consequences for lives and rep- utations. It may include expressions of pri- vate knowledge about some setting or oc- cupation that goes against that setting or occupation's public front or public presen- tation. Collecting this kind of informatioil raises some specific ethical issues. HOW DEEP? One ethical issue concerns how far an in- terviewer should go in probing informants' answers. As noted previously, in-depth in- terviewers should be prepared to follow where informants might lead, because this often leads to fruitful territory for those in- formants who wish to use the interviewing situation as an occasion for self-reflection and their own increased understanding. It is sometimes difficult, if not impossible, however, for a researcher to anticipate fully the consequences of such probing. In the case of one in-depth interview conducted by Rubin and Rubin (1995:98), an infor- mant's suicide followed a revealing inter- view by a matter of weeks; the timing of this informant's death led the researchers to wonder if there was any connection be- tween their interview and the suicide. 1985). In that case, Mario Brajuha was a graduate student who was studying a res- taurant that was "torched" (burned down), and when police investigators suspected mob arson, they went to the courts in an ef- fort to obtain Brajuha's research records. Knowing about such potential complexities in advance should stimulate researchers to give prior consideration to their ethical commitments and the lengths to which they will go in order to protect research infor- mants. In another case, a sociology gradu- ate student spent five months in jail in order to protect his subjects in a sociological field project on ecoterrorism; his incarceration produced further reflections on this ethical dilemma (Scarce 1994, 1995, 1999). PROTECTING SUBJECTS PROTECTING COMMUNITIES Professional social science organizations have traditionally addressed potentially difficult issues in their published codes of ethics (see Neuman 1994).One traditional ethical principle has been that the re- searcher must do whatever is necessary "to protect research subjects." There are sev- eral different ways in which such a princi- ple can be interpreted, however, and so there exists some ambiguity about what is required of the researcher. One interpreta- tion of this ethical principle is that the re- searcher should do what is necessary to pro- tect the specific individuals who have assisted him or her in the research, as indi- viduals. This means that a researcher or in- terviewer would feel obligated to take whatever steps are necessary to protect the individuals who have cooperated in the re- search from any misuses of the information they have shared. In one well-known case, a researcher coded all his interview records and kept them in a safe deposit box in a bank located in a state different from the one where the research was conducted (Humphreys 1970).In another famous case, a researcher went to jail rather than yield research and interview materials to court officials (Brajuha and Hallowell 1986; Hallowell Another issue concerning the protection of research informants is whether research- ers should feel any obligation to avoid caus- ing harm to the reputation, social standing, or social prestige of their informants' pro- fessions, occupations, communities, or groups as collectives. Predicting future con- sequences of this kind is highly problem- atic, so it is exceedingly difficult to assess the risk of such harm with any certainty. Another issue concerning the protection of informants is whether a research report will play some role in "deprivatizing" their lived experience (Gubrium and Holstein 1995).The risk of this is also very difficult to assess, and so it is reasonable to antici- pate that different individuals will reach different ethical judgments, even individu- als within the same support community or research team. This seems like one reason- able reading of what occurred when Caro- lyn Ellis (1986), an ethnographer, pub- lished an award-winning book about two fishing villages near the Chesapeake Bay. Ellis studied the villages over a period of 19 years, but when she returned in the early 1990sshe discovered that her published ac- counts had offended some of the commu- nity members, leading her to express some 116 + FORMS OF INTERVIEWING In-Depth Interviewing * 117 reservations about the standards she had used in the research publication (Ellis 1995).It seems clear that Ellis did not use the criteria for privacy that existed in the communities she studied, but instead used a much broader standard familiar to most of the cosmopolitans who live and work in and around universities today. The prob- lematic nature of such ethical judgments does not reduce the need for interviewers to face and address them as best they can. TELLING THE TRUTH The most important ethical imperative is to tell the truth. This issue has become espe- cially important during the current period, which Denzin and Lincoln (2000b:3) call "the postmodern moment." This moment is defined by two crises: the crisis of repre- sentation and the crisis of legitimation (for qualitative research). One response to these crises is the advocacy of "standpoint epis- temologies" (Denzin 199753-89), where the research interviewer not only self- consciously empathizes with the infor- mants as individuals, but self-consciously sympathizes with the political or commu- nity goals of those informants as a category OY collective. John Lofland (1995), a strong advocate of analytic ethnography, heartily disagrees with this position, saying that it amounts to a promotion of "fettered research." Most of the complex settings or situations that the vast majority of social scientists are likely to study are highly variegated, plural- istic, and filled with multiple perspectives and interpretations, so the adoption of a standpoint epistemology does not address certain important ethical questions (Altheide and Johnson 1994). In a situation with multiple perspectives or interpretations, whose standards or cri- teria of truth are to prevail in the final re- port?This is the critical ethical question for in-depth interviewing. In several recent publications, Denzin discusses a short story written by Raymond Carver (1989)about a writer who returns to his home town to find out that everyone there is angry with him because of what he has written about them. Denzin (1997:285-87)interprets the import of this story to be that "a writer is al- ways selling someone out," meaning that, in virtually all complex settings in today's world, all interpretations and voices are subject to conflict and dispute. To resolve this problematic dilemma, Denzin suggests "upping the ante" on the guilt and other professional consequences for not telling a defensible truth in one's writings. Robert Emerson and Melvin Pollner (1992) advocate another way to address this issue: Take the final ethnographic re- port back to the informants and other members of the setting that was studied, not so much to verify the findings inde- pendently (as in Whyte's "member's test of validity") as to gain their impressions and feedback on what has been written about them. The goal is not necessarily to seek a consensus, but to open a dialogue on what is written in the final report. E. Burke Rochford (1992)is one researcher who has actually followed this path. His experi- ences indicate that this practice may be very problematic, however; it call lead to con- flict among members who later dispute what even they will accept as a true inter- pretation, because of subsequent consider- ations about the consequences of publica- tlon. Carl Iclockars (1977) offers the opinion that "the true test of ethics of research with human beings iswhether or not it forces the researcher to suffer with his subjects" (p. 225). This is an ambiguous standard, to be sure. And Jeffrey Reiman (1979:57) would add to this the consideration of whether the publication of the research re- sults enhances the author's career or the in- formant's freedom. Even in a postmodern age characterized by little consensus on the answers to such ethical issues, the questions stay with us to haunt our enterprise. References Adhiltari, R. l?, D. Reinhard, and J. M. Johnson. 1993. "The Myth of Protection Orders." Pp. 294-311 in Studies in Symbolic Interaction: A Research Annual, Vol. 14, edited by N. I<. Denzin. Greenwich, CT: JAI. Adler, l? A., l? Adler, and E. B. Rochford. 1986. "The Politics of Participation in Fie!d Research." Ur- ban Life 14:363-76. Altheide, D. L. and J. M. Johnson. 1994. "Criteria for Assessing Interpretive Validity in Qualitative Research." Pp. 485-99 in Handbook of Qualitative Research, edited by N. I<. Denzin andY. S. Lin- coln. Thousand Oaks, CA: Sage. Atkinson, R. 1998. The Life Story Interview. Thousand Oaks, CA: Sage. Boelen, M. A. 1992. "Street Corner Society: Cornerville Revisited." Journal of Contemporary Eth- nography 21:11-51. Brajuha, M. and L. Hallowell. 1986. "Legal Intrusion and the Politics of Fieldwork: The Impact of the Brajuha Case." Urban Life 14:454-78. Buckholdt, D. R. and J. F. Gubrium. 1985. Caretakers. Beverly Hills, CA: Sage. Carver, R. 1989. "Intimacy." Pp. 444-53 in Where I'm CallingFrom, edited by R. Carver. NewYork: Vintage. Cicourel, A. V 1964. Method and Measurement in Sociology. New York: Free Press. Denzin, N. I<. 1989a. Interpretive Ifzteractionism. Newbury Park, CA: Sage. . 1989b. The Research Act: A Theoretical Introduction to Sociological Methods. 3d ed. Englewood Cliffs, NJ: Prentice Hall. . 1992. "Whose Cornerville Is It, Anyway?" Journal of Contenzporary Ethnography 21:120-32. . 1997. Interpretive Ethu.ography: Ethlzographic Practices for the 21st century. Thousand Oaks, CA: Sage. Denzin, N. I<. and Y. S. Lincoln, eds. 200021. Halzdbook of Qualitative Research. 2d ed. Thousand Oaks, CA: Sage. . 2000b. "Introduction: The Discipline and Practice of Qualitative Research." Pp. 1-28 in Halzdbook of Qualitative Research, 2d ed., edited by N. I<. Denzin and Y. S. Lincoln. Thousand Oaks, CA: Sage. Douglas, J. D. 1976. Investigative Social Research. Beverly Hills, CA: Sage. . 1985. Creative Interviewing. Beverly Hills, CA: Sage. . 1988. Love, Intimacy, and Sex. Newbury Park, CA: Sage. Ellis, C. 1986. Fisher Folk: Two Communities on Chesapeake Bay. Lexington: University Press of ICentuclty. . 1995. "Emotional and Ethical Quagmires in Returning to the Field." Journal of Contempo- rary Ethnography 24:711-13. Ellis, C. and M. G. Flaherty, eds. 1992. Investigating Subjectivity: Research on Lived Experience. Newbury Park, CA: Sage. Emerson, R. M., R. I. Fretz, and L. L. Shaw. 1995.Writing Ethnographic Fieldnotes. Chicago: Univer- sity of Chicago Press. Emerson, R. M. and M. Pollner. 1992. "Difference and Dialogue: Members' Readings of Ethnographic Texts." Pp. 79-98 in Perspectives on SocialProblei~zs,Vol. 3, edited by G. Miller and J. A. Holstein. Greenwich, CT: JAI. Ferraro, K. J. and J. M. Johnson. 1983. "How Women Experience Battering." Social Problems 30:325-39. Fontana, A. and J. H. Frey. 1994. "Interviewing: The Art of Science." Pp. 361-76 in Handbook of Qualitative Research, edited by N. I<. Denzin and Y. S. Lincoln. Thousand Oaks, CA: Sage. Fowler, F. J., Jr. and T. W Mangione. 1990. Standardized Sz~rveyInterviewing: Minimizing Inter- viewer-Related Error. Newbury Park, CA: Sage. Freeman, D. 1983. Margaret Mead and Samoa: The Making and Unmaking of aiz Anthropological Myth. Cambridge, MA: Harvard University Press.