Writing Ethnographic Fieldnotes

There were four people in line with their purchases separated by an approx. 18" rectangular black rubber bar. I put my frozen bags down on the "lazy susan linoleum conveyor belt" and I reached on top of the cash register to retrieve one of the black bars to separate my items. The cashier was in her mid thirties, approx., about 5'2" dark skinned woman with curly dark brown hair. I couldn't hear what she was saying, but recognized some accent to her speech. She was in a white blouse, short sleeved, with a maroon shoulder to mid thigh apron. She had a loose maroon bow tie, not like a man's bow tie, more hangie and fluffy. Her name tag on her left chest side had red writing that said "Candy" on it.

[Describes the first two men at the front of the line.] The woman behind him was dark skinned with straight dark brown hair cut in a page boy. She was wearing a teal blue v-neck knit sweater with black leggings. In her section was juice, a

can of pineapple juice, and a six-pack of V-8 tomato juice. The guy in front of me had a pink polo shirt on and tan shorts. He was about 6'2", slender, tan with blond short hair with a gold 18 gauge hoop in his left ear (I thought he was gay). In his triangle of space he had packaged carrots, a gallon of whole milk, and a package of porkchops.

Candy spent very little time with each person, she gave all a hello, and then told them the amount, money was offered, and change was handed back onto a shelf that was in front of the customer whose turn it was. Before Candy had given the dark-haired woman her change back, I noticed that the man in the pink shirt had moved into her spatial "customer" territory, probably within a foot of her, and in the position that the others had taken when it was their turn, in front of the "check writing" shelf. (I thought it was interesting that the people seemed more concerned about the proper separation of their food from one another's than they did about body location.) . . .

As I walk up to the shelf (where it all seems to happen), I say "Hi," and Candy says "Hi" back as she scans my groceries with the price scanner. . . .

This observer describes the line spatially in terms of individual people (particularly physical appearance and apparel) and their groceries as laid out before being rung up ("in his triangle of space he had . . ."). Indeed, this account notes as an aside the contrast between the care taken to separate grocery items and the seeming disregard of physical space that occurs at the "check writing shelf" as one shopper is about to move on and the next-in-line to move in.

I headed east to the checkout stands with my romaine lettuce, to garnish the rice salad I was bringing to brunch, and my bottle of Gewurtztraminer, my new favorite wine, which I had to chill in the next half hour. As I approached the stands, I realized that the 10-items-or-less-cash-only line would be my best choice. I noticed that Boland was behind the counter at the register—he's always very friendly to me—"Hey, how you doing?"

I got behind the woman who was already there. She had left one of the rubber separator bars behind the things she was going to buy, one of the few personal friendly moves one can make in this highly routinized queue. I appreciated this, and would have thanked her (by smiling, probably), but she was already looking ahead, I suppose in anticipation of checking out. I put my wine and lettuce down. There was already someone behind me. I wanted to show them the courtesy of putting down a rubber separator bar for them too. I waited until the food in front of mine was moved up enough for me to take the bar, which was at the front of the place where the bars are (is there a word for that? bar bin?), so that I wouldn't have to make a large, expansive move across the items that weren't mine, drawing attention to myself. I waited, and then, finally, the bar was in sight. I took it, and then put it behind my items, looking at the woman behind me and smiling at her as I did so. She looked pleased, and a bit surprised, and I was glad to have been

able to do this small favor. She was a pretty blonde woman, and was buying a bottle of champagne (maybe also for Easter brunch?). She was wearing what looked like an Easter dress—it was cotton, and pretty and flowery. She looked youngish. Maybe about my age. She was quite tall for a woman, maybe 5'10" or so.

The woman in front of me didn't take long at all. I've learned quite well how to wait in queues and not be too impatient. Boland, the checker, saw me, and said, "Hi! How's it going?" or something like that. . . .

This observer describes moving through the line as she experienced the process on a moment by moment basis, framing her accounts of others' behaviors as she received, understood, and reacted to them. This style of description gives the reader unique access to the observer's thoughts and emotions; for example, while space is an issue, it is framed in terms not of distance but of its implications for self and feelings (e.g., avoiding "a large expansive move across the items that weren't mine").

In the next excerpt, the writer shifts his focus from self to others:

. . . I picked a long line. Even though the store was quiet, the express line was long. A lot of people had made small purchases today. I was behind a man with just a loaf of bread. There was a cart to the side of him, just sitting there, and I thought someone abandoned it (it had a few items in it). A minute later a man came up and "claimed" it by taking hold of it. He didn't really try to assert that he was back in line—apparently he'd stepped away to get something he'd forgotten—but he wasn't getting behind me either. I felt the need to ask him if he was on line, so I wouldn't cut him off. He said yes, and I tried to move behind him—we were sort of side by side—and he said, "That's okay. I know where you are."

An old woman was behind me now. She had her groceries in one of those carts that old people tend to use to wheel their groceries home. She was thumbing through the National Enquirer, and was clutching a coupon in her hand. She scanned a few pages of the paper, and then put it back in the rack. I looked ahead at the person whose groceries were being checked out—she was staring at the price for each item as it came up on the register.

At this point the guy who I'd spoken to earlier, the guy who was right in front of me, showed a look of surprise and moved past me, over to an abandoned cart at the end of the aisle. He was looking at what was in it, picking up the few items with interest, and then put them back. I thought he'd seen something else he wanted or had forgotten. He came back over to his cart, but then a supermarket employee walked by, and he called out to the man, walking over to the cart and pointing at it, "Do you get many items like this left behind?" The employee hesitated, not seeming to understand the question, and said no. The guy on line said, "See what's here? This is formula [cans of baby formula]. That's poor people's food. And see this [a copper pot scrubber]? They use that to smoke crack." The

employee looked surprised. The guy says, "I was just wondering. That's very indicative of this area." The employee: "I live here and I didn't know that." The guy: "Didn't you watch Channel 28 last night?" Employee: "No." Guy: "They had a report about inner city problems." Employee, walking away as he talks: "I only watch National Geographic, the MacNeil-Lehrer Hour, and NPR." He continues away. . . .

Meanwhile the man with the bread has paid. As he waits momentarily for his change, the "guy" says, "Long wait for a loaf of bread." Man says, "Yeah," and then adds, jokingly (and looking at the cashier as he says it, as if to gauge his reaction), "these cashiers are slow." The cashier does not appear to hear this. Man with bread leaves, guy in front of me is being checked out now. He says to the cashier, "What's the matter, end of your shift? No sense of humor left?" Cashier says, "No. I'm tired." Guy: "I hear you." Guy then says to the bagger: "Can I have paper and plastic please, Jacob" (he emphasizes the use of the bagger's name)? Jacob complies, but shows no other sign that he's heard the man. Guy is waiting for transaction to be completed. He's sitting on the railing, and he is singing the words to the Muzak tune that's playing. Something by Peabo Bryson. Guy's transaction is done. He says thank you to the bagger, and the bagger tells him to have a good day.

Cashier says, "How are you doing?" to me. . . .

In these notes the observer initially writes himself into a prominent role in the line, but then he moves himself offstage by spotlighting another character who says and does a number of flamboyant things as he waits and then gets checked out. This express line becomes a mini-community, first marked by ongoing exchanges between those in line, then drawing in a passing store employee, and culminating in interactions between this character and the checker and bagger.

Writing fieldnote descriptions, then, is not so much a matter of passively copying down "facts" about "what happened." Rather, such writing involves active processes of interpretation and sense-making: noting and writing down some things as "significant," noting but ignoring others as "not significant," and even missing other possibly significant things altogether. As a result, similar (even the "same") events can be described for different purposes, with different sensitivities and concerns.

In this respect, it is important to recognize that fieldnotes involve inscriptions of social life and social discourse. Such inscriptions inevitably reduce the welter and confusion of the social world to written words that can be reviewed, studied, and thought about time and time again. As Geertz (1973:19) has characterized this core ethnographic process: "The ethnographer 'inscribes' social discourse; he writes it down. In so doing, he turns it from a passing event, which exists only in its own moment

of occurrence, into an account, which exists in its inscription and can be reconsulted."

As inscriptions, fieldnotes are products of and reflect conventions for transforming witnessed events, persons, and places into words on paper.⁸ In part, this transformation involves inevitable processes of selection; the ethnographer writes about certain things and thereby necessarily "leaves out" others. But more significantly, descriptive fieldnotes also inevitably present or frame objects in particular ways, "missing" other ways that events might have been presented or framed. And these presentations reflect and incorporate sensitivities, meanings, and understandings the field researcher has gleaned from having been close to and participated in the described events.

There are other ways of reducing social discourse to written form. Survey questionnaires, for example, record "responses" to pre-fixed questions, sometimes reducing these answers to numbers, sometimes preserving something of the respondents' own words. Audio and video recordings, which seemingly catch and preserve almost everything occurring within an interaction, actually capture but a slice of ongoing social life. What is recorded in the first place depends upon when, where, and how the equipment is positioned and activated, what it can pick up mechanically, and how those who are recorded react to its presence. Further reduction occurs with the representation of a recorded slice of embodied discourse as sequential lines of text in a "transcript." For while talk in social settings is a "multichanneled event," writing "is linear in nature, and can handle only one channel at a time, so must pick and choose among the cues available for representation" (Walker 1986:211). A transcript thus selects particular dimensions and contents of discourse for inclusion while ignoring others, for example, nonverbal cues to local meanings such as eye gaze, gesture, and posture. Researchers studying oral performances spend considerable effort in developing a notational system to document the verbal and at least some of the nonverbal communication; the quality of the transcribed "folklore text" is critical as it "represents the performance in another medium" (Fine 1984:3). The transcript is never a "verbatim" rendering of discourse, because it "represents . . . an analytic interpretation and selection" (Psathas and Anderson 1990:75) of speech and action. That is, a transcript is the product of a transcriber's ongoing interpretive and analytic decisions about a variety of problematic matters: how to transform naturally occurring speech into specific words (in the face of natural speech elisions); how to determine when to punctuate to indicate a completed phrase or sentence (given the

common lack of clear-cut endings in ordinary speech); deciding whether or not to try to represent such matters as spaces and silences, overlapped speech and sounds, pace stresses and volume, and inaudible or incomprehensible sounds or words.⁹ In sum, even those means of recording that researchers claim come the closest to realizing an "objective mirroring" necessarily make reductions in the lived complexity of social life similar in principle to those made in writing fieldnotes.¹⁰

Given the reductionism of any method of inscription, choice of method reflects researchers' deeper assumptions about social life and how to understand it. Fieldwork and ultimately the fieldnote are predicated on a view of social life as continuously created through people's efforts to find and confer meaning on their own and others' actions. Within this perspective, the interview and the recording have their uses. To the extent that participants are willing and able to describe these features of social life, an interview may prove a valuable tool. Similarly, a video recording provides a valuable record of words actually uttered and gestures actually made. But the ethos of fieldwork holds that in order to fully understand and appreciate action from the perspective of participants, one must get close to and participate in a wide cross-section of their everyday activities over an extended period of time. Ethnography, as Van Maanen (1988:ix) insists, is "the peculiar practice of representing the social reality of others through the analysis of one's own experience in the world of these others." Fieldnotes are distinctively a method for capturing and preserving the insights and understandings stimulated by these close and long-term experiences. Thus fieldnotes inscribe the sometimes inchoate understandings and insights the fieldworker acquires by intimately immersing herself in another world, by observing in the midst of mundane activities and jarring crises, by directly running up against the contingencies and constraints of the everyday life of another people. Indeed, it is exactly this deep immersion—and the sense of place that such immersion assumes and strengthens—that enables the ethnographer to inscribe the detailed, context-sensitive, and locally informed fieldnotes that Geertz (1973) terms "thick description."

This experiential character of fieldnotes is also reflected in changes in their content and concerns over time. Fieldnotes grow through gradual accretion, adding one day's writing to the next's. The ethnographer writes particular fieldnotes in ways that are not pre-determined or pre-specified; hence fieldnotes are not collections or samples in the way that audio recordings can be, i.e., decided in advance according to set criteria. Choos-

ing what to write down is not a process of sampling according to some fixed-in-advance principle. Rather it is both intuitive, reflecting the ethnographer's changing sense of what might possibly be made interesting or important to future readers, and empathetic, reflecting the ethnographer's sense of what is interesting or important to the people he is observing.

Modes of participating in and finding out about the daily lives of others make up key parts of ethnographic methods. These "methods" determine what the field researcher sees, experiences, and learns. But if substance ("data," "findings," "facts") are products of the methods used, substance cannot be considered independently of method; what the ethnographer finds out is inherently connected with how she finds it out. As a result, these methods should not be ignored. Rather, they should comprise an important part of written fieldnotes. It thus becomes critical for the ethnographer to document her own activities, circumstances, and emotional responses as these factors shape the process of observing and recording others' lives.¹¹

From this point of view, the very distinction between fieldnote "data" and "personal reactions," between "fieldnote records" and "diaries" or "journals" (Sanjek 1990c), is deeply misleading. Of course, the ethnographer can separate what he says and does from what he observes others saying and doing, treating the latter as if it were unaffected by the former.¹² But such a separation distorts processes of inquiry and the meaning of field "data" in several significant ways. First, this separation treats data

as "objective information" that has a fixed meaning independent of how that information was elicited or established and by whom. In this way the ethnographer's own actions, including his "personal" feelings and reactions, are viewed as independent of and unrelated to the events and happenings involving others that constitute "findings" or "observations" when written down in fieldnotes. Second, this separation assumes that "subjective" reactions and perceptions can and should be controlled by being segregated from "objective," impersonal records. And finally, such control is thought to be essential because personal and emotional experiences are devalued, comprising "contaminants" of objective data rather than avenues of insight into significant processes in the setting.

Linking method and substance in fieldnotes has a number of advantages: it encourages recognizing "findings" not as absolute and invariant but as contingent upon the circumstances of their "discovery" by the ethnographer. Moreover, the ethnographer is prevented, or at least discouraged, from too readily taking one person's version of what happened or what is important as the "complete" or "correct" version of these matters. Rather, "what happened" is one account, made by a particular person to a specific other at a particular time and place for particular purposes. In all these ways, linking method and substance builds sensitivity to the multiple, situational realities of those studied into the core of fieldwork practice.

In contrast to styles of field research which focus on others' behavior without systematic regard for what such behavior means to those engaged in it, we see ethnography as committed to uncovering and depicting indigenous meanings. The object of participation is ultimately to get close to those studied as a way of understanding what their experiences and activities mean to them.¹³

Ethnographers should attempt to write fieldnotes in ways that capture and preserve indigenous meanings. To do so, they must learn to recognize and limit reliance upon preconceptions about members' lives and activities. They must become responsive to what others are concerned about, in their own terms. But while fieldnotes are about others, their concerns and doings gleaned through empathetic immersion, they necessarily reflect and convey the ethnographer's understanding of these concerns and

doings. Thus, fieldnotes are written accounts that filter members' experiences and concerns through the person and perspectives of the ethnographer; fieldnotes provide the ethnographer's, not the members', accounts of the latter's experiences, meanings, and concerns.

It might initially appear that forms of ethnography concerned with "polyvocality" (Clifford and Marcus 1986:15), or oral histories and feminist ethnographies (Stacey 1991) which seek to let members "speak in their own voices," can avoid researcher mediation in its entirety. But even in these instances, researchers continue to select what to observe, to pose questions, or to frame the nature and purpose of the interview more generally, in ways which cannot avoid mediating effects (see Mills 1990).

In contrast to views holding that fieldnotes are crutches at best and blinders at worst, we see fieldnotes as providing the primary means for deeper appreciation of how field researchers come to grasp and interpret the actions and concerns of others. In this respect, fieldnotes offer subtle and complex understandings of these others' lives, routines, and meanings.

As argued earlier, the field researcher comes to understand others' ways by becoming part of their lives and by learning to interpret and experience events much as they do. It is critical to document closely these subtle processes of learning and resocialization as they occur; continuing time in the field tends to dilute the insights generated by initial contact with an unknown way of life. Long-term participation dissolves the initial perceptions that arise in adapting to and discovering what is significant to others; it blunts early sensitivities to subtle patterns and underlying tensions. In short, the field researcher does not learn about the concerns and meanings of others all at once, but in a constant, continuing process in which she builds new insight and understanding upon prior insights and understandings. Researchers should document these emergent processes and stages rather than attempt to reconstruct them at a later point in light of some final, ultimate interpretation of their meaning and import. Fieldnotes provide a distinctive resource for preserving experience close to the moment of occurrence and, hence, for deepening reflection upon and understanding of those experiences.

Similar considerations hold when examining the ethnographer's "findings" about those studied and their routine activities. Producing a

record of these activities as close to their occurrence as possible preserves their idiosyncratic, contingent character in the face of the homogenizing tendencies of retrospective recall. In immediately written fieldnotes, distinctive qualities and features are sharply drawn and will elicit vivid memories and images when the ethnographer rereads notes for coding and analysis. Furthermore, the distinctive and unique features of such fieldnotes, brought forward into the final analysis, create texture and variation, avoiding the flatness that comes from generality.

Field researchers seek to get close to others in order to understand their ways of life. To preserve and convey that closeness, they must describe situations and events of interest in detail. Of course, there can never be absolute standards for determining when there is "enough detail." How closely one should look and describe depends upon what is "of interest," and this varies by situation and by the researcher's personality, orientation, and discipline. Nonetheless, most ethnographers attend to observed events in an intimate or "microscopic" manner (Geertz 1973:20–23) and in writing fieldnotes seek to recount "what happened" in fine detail.

Beyond this general "microscopic" commitment, however, our specifically interactionist approach leads us to urge writers to value close, detailed reports of interaction. First, interactional detail helps one become sensitive to, trace, and analyze the interconnections between methods and substance. Since the fieldworker discovers things about others by interacting with them, it is important to observe and minutely record the sequences and conditions marking such interactions. Second, in preserving the details of interaction, the researcher is better able to identify and follow processes in witnessed events and hence to develop and sustain processual interpretations of happenings in the field. Field research, we maintain, is particularly suited to documenting social life as process, as emergent meanings established in and through social interaction (Blumer 1969). Attending to the details of interaction enhances the possibilities for the researcher to see beyond fixed, static entities, to grasp the active "doing" of social life. Writing fieldnotes as soon and as fully as possible after events of interest have occurred encourages detailed descriptions of the processes of interaction through which members of social settings create and sustain specific, local social realities.

In attending to ongoing scenes, events, and interactions, field researchers take mental note of certain details and impressions. For the most part these impressions remain "headnotes" only. In some instances, the field researcher makes a brief written record of these impressions by jotting

down key words and phrases. Jottings translate to-be-remembered observations into writing on paper as quickly rendered scribbles about actions and dialogue. A word or two written at the moment or soon afterwards will jog the memory later in the day and enable the fieldworker to catch significant actions and to construct evocative descriptions of the scene. Or, more extensive jottings may record an ongoing dialogue or a set of responses to questions. Particularly when learning a new language, the ethnographer should jot key expressions and terms.

Through trial and error field researchers evolve distinctive practical styles for writing jottings. An initial choice involves the selection of writing materials. Many fieldworkers use small notepads that fit easily in pocket or purse. Others prefer even less obtrusive materials, using folded sheets of paper, to record jottings about different topics on specific sides. Writers also frequently develop idiosyncratic preferences for particular types of pens or pencils.

Field researchers actually write jottings in different ways. It is time-consuming and cumbersome to write out every word fully, and many fieldworkers develop their own private systems of symbols and abbreviations. Some even learn a formal transcribing system such as shorthand or speedwriting. These procedures not only facilitate getting words on a page more quickly; they also make jotted notes incomprehensible to those onlookers who ask to see them and hence provide a means for protecting the confidentiality of these writings.

Field researchers must also decide when, where, and how to write jottings. Far from simply mundane matters, such decisions can have tremendous import for relations with those in the field. The researcher works hard to establish close ties with participants so that she may be included in activities that are central to their lives. In the midst of such activities, however, she may experience deep ambivalence: on the one hand, she may wish to preserve the immediacy of the moment by jotting down words as they are spoken and details of scenes as they are enacted; on the other hand, she may feel that taking out a note pad and writing jottings will ruin the moment and plant seeds of distrust. Participants may now see her as someone whose primary interest lies in discovering their secrets and turning their most intimate and cherished experiences into objects of scientific inquiry.⁵

Nearly all ethnographers feel torn at times between their research commitments and their desire to engage authentically those people whose worlds they have entered. Attempting to resolve these thorny relational

and moral issues, many researchers hold that conducting any aspect of the research without the full and explicit knowledge and consent of those studied violates ethical standards. In this view, local assistants must be understood as collaborators who actively work with the researcher to tell the outside world about their lives and culture. Such mutual collaboration requires that the researcher ask permission to write about events and also respect people's desire not to reveal aspects of their lives.

Other field researchers feel less strictly bound to seek permission to conduct research or to tell participants about their intention to record events and experiences. Some justify this stance by insisting that the field researcher has no special obligations to disclose his intentions; all social life involves elements of dissembling, as no one ever fully reveals all his deeper purposes and private activities. Other researchers point out that jottings and fieldnotes written for oneself as one's own record will do no direct harm to others. This approach, of course, puts off grappling with the tough moral and personal issues until facing subsequent decisions about whether to publish or otherwise make these writings available to others. Finally, some advocate withholding knowledge of their research purposes from local people on the grounds that the information gained will serve the greater good. Researchers, for example, can only describe and publicize the conditions under which undocumented factory workers or the elderly in nursing homes live if they withhold their intentions from the powerful who control access to such settings.

Many beginning researchers, wanting to avoid open violations of trust and possibly awkward or tense encounters, are tempted to use covert procedures and to try to conceal the fact that they are conducting research or to wait until they leave the field to jot notes. While these decisions involve both the researcher's conscience and pragmatic considerations, we recommend as a general policy that the fieldworker inform people in the setting of the research, especially those with whom he has established some form of personal relationship. In addition to making these relations more direct and honest, openness avoids the risks and likely sense of betrayal that might follow from discovery of what the researcher has actually been up to. Concerns about the consequences—both discovery and ongoing inauthenticity—of even this small secret about research plans may mount and plague the fieldworker as time goes on and relations deepen.

Of course, strained relations and ethical dilemmas are not completely avoided by informing others of one's research purposes. While participants may have consented to the research, they might not know exactly

what the research involves or what the researcher will do to carry it out.⁶ They might realize that the fieldworker is writing fieldnotes at the end of the day, but they become used to his presence and "forget" that this writing is going on. Furthermore, marginal and transient members of the setting may not be aware of his research identity and purposes despite conscientious efforts to inform them.

By carrying out fieldwork in an overt manner, the researcher gains flexibility in when, where, and how to write jottings. In many field situations it may be feasible to jot notes openly. In so doing the fieldworker should act with sensitivity, trying to avoid detracting from or interfering with the ordinary relations and goings-on in the field. If possible, the fieldworker should start open jottings early on in contacts with those studied. If one establishes a "note-taker" role, jotting notes comes to be part of what people expect from the fieldworker. Here it helps to offer initial explanations of the need to take notes; an ethnographer can stress the importance of accuracy, of getting down exactly what was said. People often understand that such activities are required of students and, therefore, tolerate and accommodate the needs of researchers who, they believe, want to faithfully represent what goes on. When learning a new language in another culture, the field researcher can explain that she is writing down local terms in order to remember them. By saying the word as she writes, people might offer new terms and become further interested in teaching her.

Although taking down jottings may at first seem odd or awkward, after a time it often becomes a normal and expected part of what the fieldworker does. In the following excerpt from a Housing and Urban Development [HUD] office, the office manager and a worker jokingly enlist the fieldworker as audience for a self-parody of wanting to "help" clients:

Later I'm in Jean's office and Ramon comes up and waxes melodramatic. Take this down, he says. Jean motions for me to write, so I pull out my notepad. "I only regret that I have but eight hours to devote to saving" . . . He begins to sing "Impossible Dream," in his thick, goofy Brooklyn accent. . . . "Feel free to join in," he says. . . .

Here, the ethnographer and his note-taking provide resources for a spontaneous humorous performance.⁷

Yet even when some people become familiar with open writing in their presence, others may become upset when the researcher pulls out his pad and begins to write down their words and actions. Ethnographers

may try to avoid the likely challenges and facilitate open, extensive note-taking by positioning themselves on the margins of interaction. Even then, they may still encounter questions, as reflected in the following comment by a field researcher observing divorce mediation sessions:

I tried to take notes that were as complete as possible during the session. My sitting behind the client had probably more to do with wanting to get a lot of written notes as unobtrusively as possible as with any more worthy methodological reason. While taking copious amounts of notes (approximately 50 pages per session) did not seem to bother the clients, a few mediators became quite defensive about it. One mediator wanted to know how I "decided what to write down and what not to write down." At staff meetings this same mediator would sit next to me and try to glance over to see what I had written in my notebook.

Given the delicacy of this and similar situations, fieldworkers must constantly rely upon interactional skills and tact to judge whether or not taking jottings in the moment is appropriate.⁸

Furthermore, in becoming accustomed to open jotting, people may develop definite expectations about what events and topics should be recorded. People may question why the fieldworker is or is not taking note of particular events, and they may feel slighted if she fails to make jottings on what they are doing or see as important. Consider the following exchange, again described by the field researcher studying divorce mediation, which occurred as she openly took notes while interviewing a mediator about a session just completed:

On one occasion when finishing up a debriefing. . . . [the mediator] began to apply some eye make-up while I was finishing writing down some observations. She flashed me a mock disgusted look and said, "Are you writing this down too!" indicating the activity with her eye pencil.

Open jotting, then, has to be carefully calibrated to the unfolding context of the ongoing interaction.⁹

Open jottings not only may strain relations with those who notice the writing; jottings can also distract the ethnographer from paying close attention to talk and activities occurring in the setting. A field researcher will inevitably miss fleeting expressions, subtle movements, and even key content in interactions if his nose is in his notepad. Taking open jottings is not always advisable for other reasons as well. In some settings the fieldworker's participation in ongoing interaction may be so involving as to preclude taking breaks to write down jottings; in such instances, he

may have to rely more upon memory, focusing on incidents and key phrases that will later trigger a fuller recollection of the event or scene. For example, in a setting where only a few people write and do so only on rare occasions, an ethnographer who writes instead of participating in an all-night village dance may be perceived as failing to maintain social relationships—a serious offense in a close-knit village.

As a result of these problems, even ethnographers who usually write open jottings may at other times make jottings privately, out of sight of those studied. Waiting until just after a scene, incident, or conversation has occurred, the ethnographer can then go to a private place to jot down a memorable phrase. Here it is often useful for the fieldworker to adopt the ways members of the setting themselves use to take “time out” or “get away.” Fieldworkers have reported retreating to private places such as a bathroom (Cahill 1985), deserted lunchroom, stairwell, or supply closet to record such covert jottings. Depending upon circumstances, the fieldworker can visit such places periodically, as often as every half hour or so, or immediately after a particularly important incident. Other researchers avoid all overt writing in the field setting; but immediately upon leaving the field, they pull out a notebook to jot down reminders of the key incidents, words, or reactions they wish to include in full fieldnotes. This procedure allows the fieldworker to signal items that she does not want to forget without being seen as intrusive.

An ethnographer may write jottings in ways intermediate between open and hidden styles, especially when note-taking becomes a part of her task or role. Those in the field may or may not know explicitly that the fieldworker is writing jottings for research purposes. In one instance, for example, a student in a law office who was asked to take notes during client interviews used the assignment as an opportunity to take down research jottings. This student reported that while she did not make explicit when she wrote jottings, both the attorney and clients knew of her research. Though many activities do not so easily lend themselves to writing jottings, fieldworkers can find other naturally occurring means to incorporate jottings. For example, fieldworkers often learn about settings by becoming members. For the fieldworker who assumes the role of a novice, the notes which as a beginner he is permitted or even expected to write may become the jottings for his first fieldnotes.

Strategies for how, where, and when to jot notes change with time spent in the field and with the different relationships formed between fieldworker and people in the setting. Even after the ethnographer has

established strong personal ties, situations might arise in fieldwork when visibly recording anything will be taken as inappropriate or out of place; in these situations, taking out a notebook would generate deep discomfort to both fieldworker and other people in the setting.¹⁰ One student ethnographer studying a campus bookstore who had grown quite friendly with bookstore workers—with whom she had spoken openly about her study—nonetheless reported the following incident:

One of the younger cashiers came up to me after having seen me during two of my last observation sessions. She approached me tentatively with a question about me being a “spy” from the other campus bookstore or possibly from the administration. Trying to ease the situation with a joke, I told her I was only being a spy for sociology’s sake. But she didn’t understand the joke, and it only made the situation worse.

Sometimes people may be uncomfortable with a jotting researcher because they have had little experience with writing as a part of everyday life. Especially in oral cultures, watching and writing about people may seem like a strange activity indeed. In other instances, people have unpleasant associations with writing and find jottings intrusive and potentially dangerous. On one occasion an elder in a Zambian village became very hesitant to continue speaking after the ethnographer jotted down his name on a scrap of paper, simply to remember it. She later learned that government officials in colonial times used to come by and record names for tax purposes and to enlist people into government work projects.

Finally, even with permission to write openly, the tactful fieldworker will want to remain sensitive to and avoid jotting down matters which participants regard as secret, embarrassing, too revealing, or which puts them in any danger. In other instances, the people themselves may not object and in fact urge the researcher to take notes about sensitive matters. Even though she thinks they may be embarrassing or bring them harm if they were to be made public, the researcher might take jottings but then later decide not to use them in any final writing.

All in all, it is a defining moment in field relations when an ethnographer takes out a pad and begins to write down what people are saying and doing in the presence of those very people. Therefore, fieldworkers take very different approaches to jottings, their strategies both shaping and being shaped by their setting and by their relationships. Hence, decisions about when and how to take jottings must be considered in the

context of the broader set of relations with those in the setting. In some situations and relations, taking open jottings is clearly not advisable. In others, fieldworkers decide to take jottings but must devise their own unique means to avoid or minimize awkward interactions that may arise as a result. When deciding when and where to jot, it is rarely helpful or possible to specify in advance one "best way." Here, as in other aspects of fieldwork, a good rule of thumb is to remain open and flexible, ready to alter an approach if it adversely affects people.

Deciding whether or not to make jottings presupposes some sense for what to observe and write about in the first place. But in the flux of their field settings, beginning students are often hesitant and uncertain about what they should pay attention to as potential issues for writing. We have found a number of procedures to be helpful in advising students how initially to look-in-order-to-write.¹¹

First, ethnographers should take note of their initial impressions. These impressions may include those things available to the senses—the tastes, smells, and sounds of the physical environment, the look and feel of the locale and the people in it. Such impressions may include details about the physical setting, including size, space, noise, colors, equipment, and movement, or about people in the setting, such as number, gender, race, appearance, dress, movement, comportment, and feeling tone. Recording these impressions provides a way to get started in a setting that may seem overwhelming. Entering another culture where both language and customs are incomprehensible may present particular challenges in this regard. Still, the ethnographer can begin to assimilate strange sights and sounds through writing about them.¹²

Furthermore, this record preserves these initial and often insightful impressions, for observers tend to lose sensitivity for unique qualities of a setting as these become commonplace. Researchers who are familiar with the setting they study, perhaps already having a place in that setting as workers or residents, have lost direct access to these first impressions. However, such fieldworkers can indirectly seek to recall their own first impressions by watching any newcomers to the setting, paying special attention to how they learn, adapt, and react.

Second, field researchers can focus on observing key events or incidents. Fieldworkers may at first have to rely on their own experience and intuition to select noteworthy incidents out of the flow of ongoing activity. Here, for example, the fieldworker may look closely at something that surprises or runs counter to her expectations, again paying attention to incidents, feeling tones, impressions, and interactions, both verbal and nonverbal.

Similarly, field researchers may use their own personal experience of events that please, shock, or even anger them to identify matters worth writing about. A fieldworker's strong reaction to a particular event may well signal that others in the setting react similarly. Or a fieldworker may experience deeply contradictory emotions—for example, simultaneously feeling deep sympathy and repulsion for what he observes in the field. These feelings may also reflect contradictory pressures experienced by those in the setting.

To use personal reactions effectively, however, requires care and reflection. Many beginning ethnographers take note of such experiences, but tend to judge the actions of people in the setting, for better or worse, by their own rather than the others' standards and values. Prejudging incidents in outsiders' terms makes it difficult to cultivate empathetic understanding and to discover what import local people give to them (see chapter 5). The field researcher should be alive to the possibility that local people, especially those with very different cultures, may respond to events in sharply contrasting ways. For example, an ethnographer in a Chokwe village may react with alarm to an unconscious man drugged by an herbal drink in a trial-for-sorcery court, only to realize that others are laughing at the spectacle because they know he will soon regain consciousness.

Yet fieldworkers should not go to the other extreme and attempt to manage strong personal reactions by denial or simply by omitting them from fieldnotes. Rather, we recommend that the ethnographer register her feelings, then step back and use this experience to increase sensitivity to the experiences of others in the setting. Are others in the setting similarly surprised, shocked, pleased, or angered by an event? If so, under what conditions do these reactions occur, and how did those affected cope with the incidents and persons involved? Whether an ethnographer is working in a foreign or familiar culture, she needs to avoid assuming that others respond as she does.

Third, field researchers should move beyond their personal reactions to an open sensitivity to what those in the setting experience and react to as "significant" or "important." The sorts of actions, interactions, and events that catch the attention of people habitually in the setting may provide clues to these concerns. The field researcher watches for the sorts of things that are meaningful to those studied. Specifically: What do they stop and watch? What do they talk and gossip about? What produces strong emotional responses for them? "Troubles" or "problems" often generate deep concern and feelings. What kinds occur in the setting? How do people in the setting understand, interpret, and deal with such troubles or problems? Such "incidents" and "troubles" should move the field researcher to jot down "who did what" and "how others reacted." Since a researcher in an unfamiliar setting often pays close attention to others' actions in order to imitate and participate, she can augment her learning by writing down what others do and how they respond. A follow-up strategy that we strongly recommend is to talk to those involved and those witnessing the incident about their impressions.

In this way, the field researcher attends not only to the activities local people engage in but also to the particular meanings they attribute to these activities. She seeks and discerns local knowledge and meanings, not so much by directly asking actors what matters to them, but more indirectly and inferentially by looking for the perspectives and concerns embedded and expressed in naturally occurring interaction. A field researcher, for example, might give close attention to evaluations and distinctions made by members in the course of their daily activities. By way of illustration, those in a work setting may regularly contrast "good" workers and "bad" workers. By noting such distinctions, the researcher learns something about what matters to those in the setting. In addition, by attending closely to how, in conversation, people apply these distinctions to particular workers, the fieldworker may learn how these reputations become resources used to find meaning.

In this sense, the ethnographer is concerned not with members' indigenous meanings simply as static categories but with how members of settings invoke those meanings in specific relations and interactions. This requires, then, not just that the ethnographer describe interactions, but that she consistently attend to "when, where, and according to whom" in shaping all fieldnote descriptions. Those in different institutional positions (e.g., supervisors and workers, staff and clients), for example, may evalu-

ate different workers as "good" (or "bad") and may do so by invoking different evaluative criteria. Indigenous meanings, then, rarely hold across the board but rather reflect particular positions and practical concerns that need to be captured in writing fieldnotes.

When first venturing into a setting, field researchers should "cast their nets" broadly; they should observe with an eye to writing about a range of incidents and interactions. Yet forays into a setting must not be viewed as discrete, isolated occasions that have little or no bearing on what will be noted the next time. Rather, observing and writing about certain kinds of events foreshadow what will be noticed and described next. Identifying one incident as noteworthy should lead to considering what other incidents are similar and hence worth noting. As fieldwork progresses and becomes more focused on a set of issues, fieldworkers often self-consciously collect a series of incidents and interactions of the "same type" and look for regularities or patterns in them.

Even when looking for additional examples of a similar event, the field researcher is open to and indeed searches for different forms of that event, for variations from or exceptions to an emerging pattern. Beginning field researchers are often discouraged by such discoveries, fearing that exceptions to a pattern they have noted will cast doubt upon their understanding of the setting. This need not be the case, although noting differences and variations should prod the field researcher to change, elaborate, or deepen her earlier understanding of the setting. The field researcher, for example, may want to consider and explore possible causes or conditions that would account for difference or variation: Are the different actions the result of the preferences and temperaments of those involved or of their different understandings of the situation because they have different positions in the local context? Or the ethnographer may begin to question how she decided similarity and difference in the first place, perhaps coming to see how an event that initially appeared to be different is actually similar on a deeper level. In these ways, exploring what at least initially seem to be differences and variations will lead to richer, more textured descriptions and encourage more subtle, grounded analyses in a final ethnography (see chapter 7).

In summary, ethnographic attention involves balancing two different orientations. Especially on first entering the field, the researcher identifies significant characteristics gleaned from her first impressions and personal reactions. With greater participation in some local social world, however,

the ethnographer becomes more sensitive to the concerns and perspectives of those in the setting. She increasingly appreciates how people have already predescribed their world in their own terms for their own purposes and projects. A sensitive ethnographer draws upon her own reactions to identify issues of possible importance to people in the setting but privileges their "insider" descriptions and categories over her own "outsider" views.

In order to convey how field researchers actually write and use jottings, we provide two illustrations. Both focus on scenes, observed actions, and dialogue rather than on evaluation or psychological interpretation. The two researchers approach interaction in their settings in very different ways, noting different sensory and interpretive details.

Each of the jottings in the previous illustrations is "a mnemonic word or phrase [written] to fix an observation or to recall what someone has just said" (Clifford 1990:51). As preludes to full written notes, jottings capture bits of talk and action from which the fieldworker can begin to sketch social scenes, recurring incidents, local expressions and terms, members' distinctions and accounts, dialogue among those present, and his own conversations.

Making jottings, however, is not only a writing activity; it is also a mind-set. Learning to jot down details which remain sharp and which

easily transform into vivid descriptions on the page results, in part, from envisioning scenes as written. Writing jottings that evoke memories requires learning what can be written about and how. We have found the following recommendations helpful for making jottings useful for producing vivid, evocatively descriptive fieldnotes.

First, jot down details of what you sense are key components of observed scenes or interactions. Field researchers record immediate fragments of action and talk to serve as focal points for later writing accounts of these events in as much detail as can be remembered. The field researcher studying Spanish rock music, for example, jotted that the promoter she accompanied to a club "now only speaks in Spanish" while he had spoken English in their prior, less public contacts. She also wrote down a key direct quote—"they're not very good"—along with the term "apology" to remind her of the context and meaning of this remark.

Second, avoid making statements characterizing what people do that rely on generalizations. Many novice field researchers initially tend to jot down impressionistic, opinionated words which lend themselves better to writing evaluative summaries than to composing detailed, textured descriptions. For example, it is problematic for a field researcher to characterize the way someone works as "inefficient." Such cryptic, evaluative jottings are likely to evoke only a vague memory when the fieldworker later attempts to write a full description of the social scene. Such jottings also convey nothing of how people in the setting experience and evaluate worker performance. Similarly, jottings that a probation officer "lectures about school" and that a youth is "very compliant—always agrees" during a probation interview are overly generalized; such summary statements are not helpful for writing close descriptions of how probation officer and youth actually talked and acted during a particular encounter.

Third, jot down concrete sensory details about actions and talk. Field researchers note concrete details of everyday life which show rather than tell about people's behavior (see chapter 4). By incorporating such details, jottings may provide records of actual words, phrases, or dialogue that the field researcher wants to preserve in as accurate a form as possible. It is not enough, for example, to characterize an emotional outburst simply as "angry words." Rather the ethnographer should jot the actually spoken words, along with sensual details such as gestures and facial expressions suggesting that the speaker's emotional experience involved "anger."

Jotting these words should evoke recall not only of the details about what happened but also of the specific circumstances or context involved: who was present, what they said or did, what occurred immediately before and after, etc. In this way jottings may be used to reconstruct the actual order or sequence of talk, topics, or actions on some particular occasion.

Beginning ethnographers sometimes attempt to identify motives or internal states when recording observed actions. Having witnessed an angry exchange, for example, one is often tempted to focus on the source or reason for this emotional outburst, typically by imputing motive (e.g., some underlying feeling such as "insecurity") to one or both of the parties involved. Such psychologized explanations, however, highlight only one of a number of possible internal states that may accompany or contribute to the observed actions. Anger could, for example, result from frustration, fatigue, the playing out of some local power struggle, or other hidden factors; the ethnographer who simply witnesses a scene has no way of knowing which factors are involved.¹³

Field researchers do not ignore emotions; they may well note feelings such as anger, sadness, joy, pleasure, disgust, loneliness, but they do so as such emotions are expressed and attended to by those in the setting. For example, in describing the emotional consequences of routinely "having to say no" to clients coming to a HUD office in desperate need of housing, an ethnographer wrote the following:

Laura to me, slouched down on her desk, head in hands: "Sometimes I just don't feel like helping people, you know? You have to say no so often. That's a big part of this program. It gets to you psychologically. (How?) I didn't study psychology, but it affects you" (rolling her eyes).

Here the ethnographer writes not to explain why this HUD worker experienced or reported these emotions (although she herself points to a feature of her work—"You have to say no so often"), but to highlight how she expressed her feelings. He does so not only by direct quotation in her own words but also by providing vivid details of her body posture ("slouched," "head in hands") and by noting her accenting eye movements.

When witnessing social scenes, then, the ethnographer's task is to use his own sensibilities to learn how others understand and evaluate what happened, how they assess internal states and determine psychological motivation. Useful jottings should correspondingly reflect and further this

process of writing textured, detailed descriptions of interactions rather than of individual motivation.

Fourth, jot down sensory details which you could easily forget but which you deem to be key observations about the scene. Jottings are devices intended to encourage the recall of scenes and events in the construction of some broader, fuller fieldnote account. Since jottings must later jog the memory, each field researcher must learn which kinds of details they best remember and make jottings about those features and qualities they easily forget. Thus, fieldworkers come to develop their own jotting styles reflecting their distinctive recall propensities, whether visual, kinetic, or auditory. Some focus on trying to capture evocative pieces of broader scenes, while some jot down almost exclusively dialogue; others record nonverbal expression of voice, gesture, movement; still others note visual details of color and shape. Through trial and error, field researchers learn what most helps them to recall field experiences once they sit down to write up full notes.

Jottings may serve more generally to remind the ethnographer of what was happening at a particular time, in this sense providing a marker around which to collect other remembered incidents. For example, one field researcher teaching in a Headstart Program described a series of incidents that occurred while supervising children playing in a sandbox. Included in her jottings but not in her full fieldnotes was the phrase, "Three new bags of sand were delivered to the sandbox." In discussing this scratch note later she commented: "I don't think it is so important as I would want to include it in my notes because I think it is just—I wrote it down to remind me more what the day was like, what was happening."¹⁴

Fifth, jottings can be used to signal general impressions and feelings, even if the fieldworker is unsure of their significance at the moment. In some cases, the ethnographer may have only a vague intuitive sense about how or why something may be important. Such feelings might signal a key element that in the future could enable the field researcher to see how incidents "fit together" in meaningful patterns. For example, at another point the ethnographer in the Headstart Program made a jotting about a student, "Nicole showing trust in me," which she decided not to write up in her full notes: "It was just an overall feeling I had throughout the day; . . . at that point when I wrote the jottings I couldn't remember an exact incident." But this jotting served as a mental note, subsequently stimulating her to appreciate (and record) the following incident as a revealing example of "children trusting teachers":

At one point, Nicole got on the swings without her shoes on and asked me for a push. I told her that I would push her after she went and put her shoes on. Nicole paused and looked at me. I repeated my statement, telling her that I would save her swing for her while she was gone. Nicole then got off of the swing and put her shoes on. When she came back to the swing, I praised her listening skills and gave her a hug. I then gave her a push. I found this incident to be a significant accomplishment for Nicole, as usually she doesn't listen to the teachers.¹⁵

Having thought about whether or not to write this jotting up as full notes made this student sensitive to the issue of "trust." The jotting later acted as a stimulus to observe and write up a "concrete event" involving such "trust."

In summary, by participating in a setting with an eye to making jottings, an ethnographer experiences events as potential subjects for writing. Like any other writer, an ethnographer learns to recognize potential writing material and to see and hear it in terms of written descriptions. Learning to observe in order to make jottings thus is keyed to both the scene and to the page. Ethnographers learn to experience through the senses in anticipation of writing: to remember dialogue and movement like an actor; to see colors, shapes, textures, and spatial relations as a painter or photographer; and to sense moods, rhythms, and tone of voice like a poet. Details experienced through the senses turn into jottings with active rather than passive verbs, sensory rather than analytic adjectives, and verbatim rather than summarized dialogue.

While a primary goal of ethnography is immersion in the life-worlds and everyday experiences of a group of people, the ethnographer inevitably remains in significant ways an outsider to the worlds of those studied. Immersion is not merging; the field researcher who seeks to "get close to" others usually does not become one of these others but rather continues to be a researcher interested in and pursuing research issues, albeit in close proximity to the ordinary exigencies of life that these others experience and react to (see Bittner 1988; Emerson 1987).¹⁶ The ethnographer, then, stays at least a partial stranger to the worlds of the studied, despite sharing in many aspects of their daily lives. The student-ethnographer working in a bookstore noted that the pull toward involvement as an insider was particularly strong and the researcher's stance difficult to maintain:

There were times when I wanted to be free to listen to other individuals talk or to watch their activities, but friends and acquaintances were so "distracting" coming up and wanting to talk that I wasn't able to. Also, there was this concern on my part that, as I got to know some of the staff people better, their qualities as human beings would become so endearing that I was afraid that I would lose my sociological perspective—I didn't want to feel like in studying them, I was exploiting them.

Field researchers respond to these tensions in a variety of ways. Some try to maintain a detached, observational attitude even toward people whom they like and respect, seeking to keep research commitments somewhat separate from personal attachments.¹⁷ Others find themselves unable to sustain an invariably watching, distancing stance toward people they are drawn to and toward events which compellingly involve them. These ethnographers then take time out, either implicitly or self-consciously, by not observing and/or writing fieldnotes about selected portions of their field experience while continuing to do so about other portions. And finally, some ethnographers may decide that the relationships they have formed in the field are more valuable and enduring than any research product, and eventually they come to abandon entirely the project as research activity.

But the ethnographer remains a stranger as long as, and to the extent that, she retains commitment to the exogenous project of studying or understanding the lives of others—as opposed to the indigenous project of simply living a life in one way or another. When living in a village on a long-term basis, the ethnographer may feel drawn into daily, intimate relations as a neighbor or perhaps even as a part of a family. On these occasions she may participate "naturally"—without a writing orientation or analytic reflection—in ongoing social life. But on other occasions, she participates in local scenes in ways directed toward making observations and collecting data. Here her actions incorporate an underlying commitment to write down and ultimately transform into "data" the stuff and nuances of that life. In this way, efforts to observe in order to write about shared experiences and witnessed events induce a distinctive ethnographic stance. In this sense, we can suggest that the ethnographer's strangeness is created and maintained exactly by writing fieldnotes; such notes reflect and realize this socially close but experientially separate stance.¹⁸

This ethnographic marginality is often manifested interactionally

when the fieldworker ceases simply doing what other people are doing and begins openly writing about these doings. In this sense, overtly writing jottings is a critical, consequential ethnographic activity, publicly proclaiming and reaffirming fieldworkers' research commitments and hence their status as outsiders, as persons in the setting who have clearly delineated tasks and purposes that differ from those of members.¹⁹ Writing down jottings not only reminds ethnographers of their marginal social standing in settings but creates it as well, increasing immediate feelings of isolation and alienation.

It should come as no surprise, then, that many ethnographers, both students and experienced practitioners, feel deeply ambivalent about jottings. Jottings interfere with their interactions with people in the field; they create difficulties in interacting with others while at the same time observing and writing down what is happening. Indeed, students who come back from the field without jottings usually report that taking jottings on the spot would have made others uncomfortable. These students, then, directly experience the distracting, alienating consequences of jotting notes.

Most ethnographers, however, try to balance and juggle these tendencies, sometimes participating without immediate thought about writing down what is occurring, sometimes temporarily withdrawing to some private place to write covert jottings, at other times visibly jotting notes. Several practical writing conflicts arise from these opposing pressures. The inclination to experience daily events either as a "natural" participant or as a researcher shows up in writing as shifts in point of view as well as in varying kinds of details considered significant for inscription. Even where and when to jot notes depends on the person's involvement at a particular moment as a participant or observer. Whether a researcher-as-neighbor in the village and researcher-as-intern on a job, the tension between the present-oriented day-to-day role and the future-oriented ethnographer identity appears in the practical choices in writing both jottings and more complete notes.

In sum, in most social settings writing down what is taking place as it occurs is a strange, marginalizing activity, marking the writer as an observer rather than as a full, ordinary participant. But independently of the reactions of others, participating in order to write leads one to assume the mind-set of an observer, a mind-set in which one constantly steps outside of scenes and events to assess their "write-able" qualities. It may

be for this reason that some ethnographers try to put writing out of mind entirely by opting for the more fully experiential style of fieldwork. But this strategy puts off rather than avoids the marginalizing consequences of writing, for lived experience must eventually be turned into observations and reduced to textual form.

After hours participating in, observing, and perhaps jotting notes about ongoing events in a social setting, most fieldworkers return to their desks and their computers or typewriters to begin to write up their observations into full fieldnotes. At this point, writing becomes the explicit focus and primary activity of ethnography: momentarily out of the field, the researcher sits down to the task of turning recollections and jottings into detailed written accounts that will preserve as much as possible what she noticed and now feels is significant.

In this chapter, we are concerned with the processes of writing up full fieldnotes; we focus on how ethnographers go about the complex tasks of remembering, elaborating, filling in, and commenting upon fieldnotes in order to produce a full written account of witnessed scenes and events.

Writing up fieldnotes requires a block of concentrated time. Incidents that span a few minutes may take the ethnographer several hours to write up; she tries to recall just who did and said what, in what order, and to put all that into words and coherent paragraphs. Indeed, an ethnographic maxim holds that every hour spent observing requires an additional hour to write up. Over time, fieldworkers evolve a rhythm that balances time spent in the field and time writing notes. In some situations, the field researcher may put a cap on time devoted to observing in order to allow

a substantial write-up period on leaving the field. Limiting time in the field in this way lessens the likelihood that the fieldworker will forget what happened or become overwhelmed by the prospect of hours of composing fieldnotes. For beginning ethnographers, we recommend, when possible, leaving the field after three to four hours in order to begin writing fieldnotes.

In other situations the fieldworker might find it more difficult to withdraw for writing. Anthropologists working in other cultures generally spend whole days observing and devote evenings to writing. Field researchers who fill roles as regular workers must put in a full work day before leaving to write notes. In both cases, longer stretches of observation require larger blocks of write-up time and perhaps different strategies for making note-writing more manageable. For example, once having described basic routines and daily rhythms in the first sets of notes, the ethnographer who spends hours in the field may focus subsequent notes on significant incidents that occurred throughout the day. At this stage longer periods spent in the field may in fact prove advantageous, allowing greater opportunities for observing incidents of interest.

Alternatively, the field researcher with regular workday responsibilities may find it useful to designate certain hours for observing and taking jottings, giving priority to these observations in writing up full fieldnotes. Varying these designated observation periods allows exploration of different patterns of activity throughout the day. Of course, while using this strategy, the fieldworker should still write notes on important incidents that occur at other times.

Perhaps more crucial than how long the ethnographer spends in the field is the timing of writing up fieldnotes. Over time, people forget and simplify experience; notes composed several days after observation tend to be summarized and stripped of rich, nuanced detail. Hence we strongly encourage researchers to sit down and write full fieldnotes as soon as possible after the day's (or night's) research is done.

Writing fieldnotes immediately after leaving the setting produces fresher, more detailed recollections that harness the ethnographer's involvement with and excitement about the day's events. Indeed, writing notes immediately on leaving the field offers a way of releasing the weight of what the researcher has just experienced. It is easier to focus one's thoughts and energies on the taxing work of reviewing, remembering, and writing. In contrast, those who put off writing fieldnotes report that

with the passage of time the immediacy of lived experience fades and writing fieldnotes becomes a burdensome, even dreaded experience.

Often, however, it is impossible for an ethnographer to find time to write up notes immediately upon leaving the field. Long or late hours, for example, may leave him too tired to write notes. Under these circumstances, it is best to get a good night's sleep and turn to writing up first thing in the morning. Sometimes even this is impossible: a village event may last through several days and nights, confronting the anthropological researcher with a choice between sleeping outside with the villagers or taking time out periodically to sleep and write notes.

Whether written immediately or soon after returning from the site, the fieldworker should go directly to computer or typewriter, not talking with intimates about what happened until full fieldnotes are completed. Such "what happened today" talk can rob note-writing of its psychological immediacy and emotional release; writing the day's events becomes a stale recounting rather than a cathartic outpouring.¹

Ethnographers use a variety of different means to write up full notes. While the typewriter provided the standard tool for many classic ethnographers, some handwrote their full notes on pads or in notebooks. Contemporary ethnographers strongly prefer a computer with a standard word processing program. Typing notes with a word processing program not only has the advantage of greater speed (slow typists will soon notice substantial gains in speed and accuracy), but also allows for the modification of words, phrases, and sentences in the midst of writing without producing messy, hard-to-read pages. And fieldnotes written on the computer are easily reordered; it is possible, for example, to insert incidents or dialogue subsequently recalled at the appropriate place. Finally, composing with a word processing program facilitates coding and sorting fieldnotes as one later turns to writing finished ethnographic accounts.

The researcher who has been in the field for a long period and has limited time immediately afterward for writing full fieldnotes has several alternatives. First, he may make extensive, handwritten jottings about the day's events, relying on the details of these notes to postpone writing full fieldnotes, often for some time.² Secondly, she may dictate fieldnotes into a tape recorder. One can "talk fieldnotes" relatively quickly and can dictate while driving home from a field setting. But while dictation preserves vivid impressions and observations immediately on leaving the field, dictated notes eventually have to be transcribed, a time-consuming, expen-

sive project. And in the meantime, the field researcher does not have ready access to these dictated notes for review or for planning her next steps in the field.

In summary, beginning ethnographers should not be surprised to experience deep ambivalence in writing fieldnotes. On the one hand, after a long, exciting, or draining stint in the field, writing up notes may seem a humdrum, extra burden; on the other hand, writing fieldnotes may bring expressive release and reflective insight. Having seen and heard intriguing, surprising things all day long, the fieldworker is finally able to sit down, think about, and relive events while transforming them into a permanent record. Writing fieldnotes may bring on an outpouring of thoughts and impressions as the writer reviews and re-experiences the excitement and freshness of the day's events.

Sitting down to write full fieldnotes involves a turning away from the field scene toward "getting it down on the page," toward the worlds of research and writing. In making this turn, the field researcher does not grapple simply with what to write down; she also decides how she is going to present and convey what she has seen and observed. While some of these decisions are relatively straightforward, others are more implicit, arising from the particular stance adopted in writing fieldnotes, that is, from the author's orientation toward and attitudes about the topic or people studied. Whether her stance is primarily influenced by a theoretical position within a discipline or by her personal, moral, and political commitments, the ethnographer expresses this fundamental stance in a distinctive tone which can be heard and sensed in specific writing choices throughout her fieldnotes.

At a fundamental level, a researcher's stance in fieldwork and note-writing originates in his outlook on life. Prior experience, training, and commitments influence the fieldworker's stance in writing notes; these influences predispose him to feel, think, and act toward people in more or less patterned ways. Whether it be from a particular gender, social, cultural, or intellectual position, the fieldworker not only interacts and responds to people in the setting from that orientation but also writes his fieldnotes by seeing and framing events accordingly. The effects of this fundamental stance appear in fieldnote writing in subtle ways: these range from how

he identifies with (or distances himself from) those studied and thus writes about them sympathetically (or not), to the kinds of local activities which draw his attention and result in more detailed descriptions, to the way he prioritizes and frames certain topics and thus writes more fully about any events he sees as relevant.

By self-consciously recognizing his fundamental orientation, the fieldworker may be able to write fieldnotes that highlight and foreground issues and insights made available by that orientation. This recognition may also make him more sensitive to the ways his orientation shapes key interactions with others. For example, in writing up fieldnotes about a school for gays and lesbians, one heterosexual male often wrote about the ways students pressed him to reveal his sexual orientation and watched for his responses to their jokes and teasing. But an openly identified gay male researcher in the same field-site became sensitive to how students "sexualized" stories about their experiences as they constructed gay identities in everyday talk. Indeed, he then began to ask and write about students' talk about sexual activities, as in the following fieldnote:

"Wait," I said, interrupting his story. "Where was this?" "Over by Circus Books," Adam said. "And what was he doing?" I asked as I leaned forward smiling slightly. "He was cruising," Adam said. . . . "What's that?" I asked. . . . "It's a meeting place," answered John. "And this is at a bookstore," I said sounding a bit confused. "Yeah," they both said reassuringly.

The more the field researcher acknowledges those factors influencing his fundamental stance toward people in the setting, the more he can examine and use the insights and appreciations opened up by this stance in fieldnote writing. Furthermore, he can better guard against any overriding, unconscious framing of events—for example, by avoiding evaluative wording or by focusing on members' views of events.

As fieldwork progresses, the researcher's stance toward people and issues may change. As she learns through interactions with individuals in the setting to look at activities, events, and issues in new ways, she may adjust her prior views and reorient herself vis-à-vis others. Having readjusted her stance toward people in the setting, she will more frequently write fieldnotes in ways which not only highlight members' views but which also reveal her ongoing resocialization. Over time, a fieldworker's personal views and theoretical commitments often change; her stance in writing fieldnotes shifts as she more frequently comes to see and respond to events as members do.

Another key component determining the stance expressed in written fieldnotes is intended or likely audience. How a field researcher writes about observed events is linked to often unacknowledged assumptions about whom he is writing for. We first consider anticipated actual readers and then turn to the subtle but significant relevance of more diffusely envisioned audiences.

Under most circumstances, a researcher writes fieldnotes immediately for herself as a future reader. This absence of an actual reader allows the researcher to write in relaxed and shifting styles, moving from stance to stance, from audience to audience without worrying (at that point) about consistency or coherence. In this sense, fieldnotes should be written "loosely" and flowingly. If and when fieldnotes are shown to another reader—usually in a more comprehensive paper or article—the field researcher can at this time take control of this process; she can select, focus, and edit any notes before making them available to others. As future reader of her own fieldnotes, the researcher anticipates a detailed reading in order to code and analyze the notes for a paper or article.

Fieldnotes are also written differently depending upon how far in the future the field researcher projects himself as reader. Student field researchers, for example, may write notes for themselves as readers in the near future—e.g., at the end of the quarter or semester when working on a final paper. Experienced ethnographers envision themselves as readers in a more distant future, recognizing that notes should include details and background sufficient for making sense of them several years hence, when the immediacy of the field experience has faded.

In practice, however, the researcher-writer may have in mind actual readers other than herself. Student researchers in particular must ordinarily submit their fieldnotes to an instructor and write notes for that reader. Similarly, field researchers in team projects (Douglas 1976) will write notes to be read by co-workers and colleagues. Here field researchers may self-consciously write with actual readers in mind, producing accounts explicitly oriented to these others' knowledge and concerns. One common effect of writing with such readers in mind is to include more details of background and context to make fieldnotes more accessible. The ethnographer should nonetheless try to maintain a loose, flowing, and shifting approach, not trying to write with consistency of voice and style.³

The effects of envisioned audiences on how fieldnotes are written are more subtle and complex than those of actual readers.⁴ The ethnogra-

pher's stance in writing fieldnotes involves trying to convey something about the world she has observed to outside audiences made up of those who are unfamiliar with that world. In this sense fieldnotes are ultimately intended for outsiders of one sort or another. Indeed, it is in this respect that fieldnotes differ from a personal diary: fieldnotes are not merely the personal reactions of the writer, intended to heighten self-awareness and self-insight; they are more fundamentally accounts framed and organized to be read by some other, wider audience.

Many ethnographers envision and write for a professional audience, foaming their fieldnotes with eventual publication in mind. These sorts of notes may need some polishing and smoothing, but the writing is intended to be comprehensible to other professionals who are unfamiliar with the people and customs being written about. To the extent that the researcher-writer is self-conscious about writing for an ultimate, broader audience, notes will be richer; they will provide more background, context, and detail.

This is not to say that fieldnotes in "raw" form would be immediately comprehensible to professional or other outside readers. Fieldnotes are an accumulating body of writings in which the sense of later portions will depend upon what has been written earlier. People or events described in earlier notes, for example, need not be described in later ones. And indeed, just who the people are in particular incidents may not be evident to outside readers because of abbreviated names and lack of socially identifying information.⁵ Only with filling in and contextualizing would such a fieldnote actually become comprehensible to someone other than the writer. Thus, accumulating fieldnote entries have an open-endedness which allows for new information and insights and an unfinished in-progress quality which calls for editing later on.

In writing fieldnotes most ethnographers probably shift between self and professional others as envisioned future audiences. When writing in the first person about one's own direct involvement in field events or when reflecting on one's emotional reactions or intuitions about next steps to take in the field, for example, the ethnographer may assume that these accounts will only be read by and hence need only to be comprehensible to oneself. In contrast, when writing up an event that was deeply "important" to those in the setting and that is likely to be excerpted for the final ethnography, the writer may strive for completeness and detail.

In sum, stance and envisioned audiences significantly prefigure the way a researcher composes fieldnotes, even though both take on height-

ened salience when the field researcher self-consciously prepares texts for wider audiences. For writing fieldnotes involves a series of intricate, moment by moment choices in abstracting and processing experience. These choices involve not only what to look at and perhaps jot down, but also for whom and hence how to write up full fieldnotes. For student ethnographers, this audience is usually an instructor who reads and comments on fieldnotes, although more ultimate audiences—e.g., professional readers in one's own discipline—may also have an influence even at this point. These intended and anticipated audiences and the theoretical commitments they reflect linger as an influential presence over every ethnographer's shoulder.

At first glance, writing up may appear to be a straightforward process to the fieldworker. It may seem that with sufficient time and energy, she can simply sit down and record her observations with little attention to her writing process. While having enough time and energy to get her memories on the page is a dominant concern, we suggest that the fieldworker can benefit by considering several kinds of writing choices. In this section we discuss "purposes and styles," "recalling in order to write," "turning jottings into full fieldnotes," "multiple voices and points of view," and "real-time and end-point descriptions" as these influence the process of writing.

To characterize fieldnotes as descriptions initially conveys the prospect of simple, straightforward writing. But once we recognize that description involves more than a one-to-one correspondence between written accounts and what is going on, writing fieldnotes raises complex, perplexing problems. Descriptions are grounded on the observer/writer's participation in the setting, but no two persons participate in and experience a setting in exactly the same way. Moreover, there is always more going on than the ethnographer can notice, and it is impossible to record all that can be noticed. Description inevitably involves different theories, purposes, interests, and points of view. Hence, fieldnotes contain descriptions that are more akin to a series of stories portraying slices of life in vivid detail than to a comprehensive, literal, or objective rendering.¹³

The ethnographer, however, needs to avoid getting drawn into the complexities of fieldnote descriptions while actually writing fieldnotes.

She must initially work in a writing mode, putting into words and on paper what has been seen and heard as quickly and efficiently as possible. In this text-producing mode, the ethnographer tries to "get it down" as accurately and completely as possible, avoiding too much self-consciousness about the writing process itself. She stays close to the events at issue, rekindling her excitement about these events, getting on with the task of inscribing them before memory fades. The writing ethnographer tries to "capture what is out there," or more accurately, to construct detailed accounts of her own observations and experience of what is "out there." At this point, too much reflection distracts or even paralyzes; one tries to write without editing, to produce detailed descriptions without worry about analytic import and connections, to describe what happened without too much self-conscious reflection.

Only subsequently, once a text has actually been produced, can the ethnographer really step back and begin to consider the complexities which permeate fieldnote descriptions; only with fully detailed fieldnotes can the ethnographer adopt a reading mode and begin to reflect on how these accounts are products of his own often implicit decisions about how to participate in and describe events. That is, only with full notes in hand does it make sense to view these writings as texts that are truncated, partial, and perspectival, products of the ethnographer's own styles of participating, orienting, and writing. It is at this point that the ethnographer can begin to treat fieldnotes as constructions, to read them for the ways they create rather than simply record reality.

One key difference between initially working in a writing mode and subsequently in a reflexive reading mode lies in how the ethnographer orients to issues of "accuracy," to "correspondence" between a written account and what it is an account of. In the moment of writing the ethnographer must try to create some close correspondence between the written account and his experiences and observations of "what happened." The immediate task in writing fieldnote descriptions is to create a detailed, accurate, and comprehensive account of what has been experienced. But once notes have been written, this correspondence criterion may lose salience. "What happened" has been filtered through the person and writing of the observer onto the written page. The resulting text "fixes" a social reality in place, but in a way that makes it difficult to determine its relationship with realities outside that text. Readers may attempt to do so by invoking what they know from having "been there" or from experience with a similar reality. But readers are heavily con-

strained by what is on the page; they usually lack any effective means of gaining access to "what actually happened" independently of the written account. In such a reading mode, then, self-conscious, self-critical reflection on how writing choices have helped construct specific texts and textual realities becomes both possible and appropriate.

The ethnographer's central purpose is to describe a social world and its people. But often beginning researchers produce fieldnotes lacking sufficient and lively detail. Through inadvertent summarizing and evaluative wording, a fieldworker fails to adequately describe what she observed and experienced. The following strategies enable a writer to coherently depict an observed moment through vivid details: these strategies are description, presentation of dialogue, and characterization. As is evident in several of the included excerpts, fieldworkers often merge several strategies.

Not only can ethnographers improve their descriptive writing by learning strategies for depicting moments of lived experience, but they also can write more clearly by organizing their writing into coherent units. Clear organizational strategies enable them to conceptualize units and more easily write up an extended and complicated memory as a coherent set of fieldnotes.

Initially, field researchers may find it difficult to sustain a concentrated attention when writing. They may recall disconnected incidents and rush through their writing, thereby producing scattered details and unreadable fragments that will make it difficult to "see" or "hear" a coherent slice of life and therefore hard to subsequently use their fieldnotes. Or, they press on for pages, meandering through endless minutiae which also make their later rereading and analysis of fieldnotes difficult. However, remembering any observed scene as a series of moments and then sustaining that memory as a perceived whole will help one write up a "complete" scene. In addition, by drawing on conventional strategies for organizing their notes, field researchers can more easily sustain a focused attention while writing all the details remembered.

Field researchers frequently attempt to write detailed descriptions which fully document events from beginning to end. They want to be sure, later on, that they have not accidentally skipped writing up an important feature of an observed event or incident. Consequently, fieldworkers write up their fieldnotes in ways which indicate that they have recorded the "entire" scene. They do so not only by providing full details but also by writing in units which create and mark beginnings and endings. Details grouped in a paragraph form one such unit.

By convention, a paragraph coheres because the writer's attention focuses on one idea or insight. That means he perceives some actions as a gestalt and concentrates on them. While writing, he shifts attention from one recalled moment to another: for example, from one person or activity

to another within a classroom. These slight shifts are often indicated by a paragraph break. Thus, in most fieldnotes, a paragraph presents a coherent moment and thereby structures the description.⁸

Ethnographers also organize their writing into units extending beyond a paragraph. They organize memories into larger wholes by sustaining their attention throughout an entire situation or event, so that the separate moments cohere to create a "complete" description of a scene. That larger whole becomes an extended entry, which may include a series of paragraphs. Though the largest unit of writing is that day's entire set of fieldnotes, the ethnographer often finds that she organizes her recollections and jottings to write up several discrete units within that day's entry.

These extended entries within a day's fieldnotes often concern an incident or event which the writer has noted to herself as especially important. Many fieldworkers sit down to write these events first. Especially when working on a computer, one can order the sequence of the day's activities and then scroll down to begin with that key event. In a sense, the ethnographer juggles two competing purposes: to get down all activities as quickly as possible and to write as fully and vividly as possible those scenes she intuits as most vital to her research project. When first in the field, the ethnographer might be more concerned with getting everything down: such writing is often more fragmented and spontaneous. Later on as her research progresses, she may be more invested in writing extended entries about incidents and events she selects as particularly revealing of members' experiences. Such fieldnotes require sustained concentration through longer passages and benefit from the extra effort to write vivid, detailed, and coherent sequences of action.

In writing such extended entries, ethnographers work with a number of different organizing units. In the following section, we consider three such options: sketches which describe a slice of life as an observed scene; and two forms of narrative entry, episodes and more extended fieldnote tales, which recount interactions, incidents, and events as dynamic actions in an unfolding scene.

As the field researcher participates in the field, she inevitably begins to reflect on and interpret what she has experienced and observed. As previously noted, writing fieldnotes heightens and focuses these interpretive and analytic processes; writing up the day's observations generates new appreciation and deeper understanding of witnessed scenes and events. In writing, a field researcher assimilates and thereby starts to understand an experience. She makes sense of the moment by intuitively selecting, highlighting, and ordering details and by beginning to appreciate linkages with or contrasts to previously observed and written-about experiences. Furthermore, she may begin to reflect on how she has presented and ordered events and actions in her notes, rereading selected episodes and tales with an eye to their structuring effects.

To capture these ruminations, reflections, and insights and make them available for further thought and analysis, field researchers pursue several kinds of analytical writing that stand in stark contrast to the descriptive writing we have emphasized to this point. As the result of such writings, the researcher can bring a more probing glance to further observations and descriptive writing and consequently become more selective and in-depth in her descriptions.

These interpretive and analytic forms of writing include asides, commentaries, and in-process memos. These analytic writing strategies move progressively further from the descriptive writing that produces the bulk of the researcher's fieldnotes. Asides and commentaries are analytic writings touched off while composing fieldnotes; the researcher might get an idea as she recalls and puts on paper the details of a specific observation or incident. In-process memos, however, are products of a more concentrated effort to identify and develop analytic themes while still actively in the field and writing fieldnotes. Such memos are not usually stimulated by writing up a specific event. Taking a more consistent analytic stance, the fieldworker produces in-process memos by reading recently completed pieces or sets of fieldnotes to identify and develop particular interpretations, questions, or themes.¹⁵ Clearly, these distinctions are not absolute; the lines between asides, commentaries, and in-process memos are often blurred. We offer them as heuristic devices that may sensitize the fieldworker both to momentary and to more sustained concentration on analytic writing while actively producing fieldnotes.

Asides are brief, reflective bits of analytic writing that succinctly clarify, explain, interpret, or raise questions about some specific happening or process described in a fieldnote. The ethnographer dashes off asides in the midst of descriptive writing, taking a moment to react personally or theoretically to something she has just recounted on paper, and then immediately turns back to the work of description. These remarks may be inserted in the midst of descriptive paragraphs and set off by parentheses, as in the following examples. In the first, the writer pauses in an aside to explain his reaction:

As we're driving along I ask George how long he has been teaching and he tells me, "Well, it's going on my eighth year." I respond with a "Wow! (I didn't think he was as old as that.) He smiles as he looks at me and goes on to explain where he used to teach.

In the next excerpt, the ethnographer notes his uneasy feelings in an aside about someone watching him:

I turn around, away from the office, and face the woman with the blondish hair who is still smiling. (I can't shake the feeling that she's gazing at me.) "I'll see you Friday," I say to her as I walk by her and out the front door.

Fieldworkers often write somewhat more elaborate asides, several phrases in length, again triggered by some immediate piece of writing and closely tied to the events or scenes depicted in that writing. In the fieldnote below, the fieldworker describes a moment during her first day at a crisis drop-in center and then reacts to that experience in a more extended aside:

Walking up the stairs to the agency office, I noticed that almost every step creaked or moaned. At the top stands an old pine coat hanger, piled high with coats. Behind it is a bulletin board containing numerous flyers with information about organizations and services of various kinds. (Thinking about the scene as I climbed those stairs, I think that if I were an upset, distraught client, I would most probably find it difficult to find helpful information in that disorganized mass.)

In providing her own "lived sense" of the agency, the student incorporates in her description the meaning of physical space, while allowing for the possibility that others might perceive it differently. Asides may also be used to explain something that would otherwise not be apparent, or to offer some sort of personal reflection or interpretive remark on a matter

just considered. Ethnographers frequently use asides, for example, to convey their explicit "feel" for or emotional reactions to events; putting these remarks in asides keeps them from intruding into the descriptive account.

A commentary is a more elaborate reflection on some specific event or issue; it is contained in a separate paragraph and set off with parentheses. Commentaries involve a shift of attention from events in the field to outside audiences imagined as having an interest in something the fieldworker has observed and written up. Again, in contrast to descriptive fieldnotes, commentaries may explore problems of access or emotional reactions to events in the field, suggest ongoing probes into likely connections with other events, or offer tentative interpretations. Putting commentary in separate paragraphs helps avoid writing up details as evidence for preconceived categories or interpretations.¹⁶

Commentaries can be used to create a record of the ethnographer's own doings, experiences, and reactions during fieldwork, both in observing-participating and in writing up. A researcher/intern placed in a social service agency wrote the following commentary after describing an incident that she felt marked a turning point in her relationship with staff:

Entering the kitchen, where staff often go to socialize alone, I began to prepare my lunch. Soon, several staff had come in, and they began to talk among one another. I stood around awkwardly, not quite knowing what to do with myself. I exchanged small talk for awhile until D, the director, asked in her typically dramatic tone loud enough for everyone to hear: "Guess where A (a staff member who was also present) is going for her birthday?" There was silence in the room. Turning in her direction, I realized that she was speaking to me. "Where?" I asked, somewhat surprised that she was talking to me. "To Hershey Park!" she exclaimed. "No way!" I said, and feeling embarrassed, I started laughing. "Yeah," D exclaimed, "She's gonna dip her whole body in chocolate so R (lover) can eat her!" The room filled up with laughter, and I, too, could not restrain my giggles.

(With that, the group broke up, and as I walked back to my desk, I began to feel that for the first time, I had been an active participant in one of their kitchen get-togethers. This experience made me believe that I was being viewed as more than just an outsider. I have been trying to figure out what it takes to belong here, and one aspect undoubtedly is to partake in an occasional kitchen get-together and not to appear above such practices.)

In this commentary the researcher not only reports her increased feeling of acceptance in the scene but also reflects on the likely importance of

these informal, sometimes ribald "get-togethers" for creating a general sense of belonging in the organization.

Finally, commentaries can raise issues of what terms and events mean to members, make initial connections between some current observation and prior fieldnotes, and suggest points or places for further observation, as in the following excerpt:

M called over to Richard. He said, "C'm here lil' Homey." Richard came over to sit closer to M. He asked Richard about something Richard said earlier (I couldn't completely hear it) . . . something to do with weight lifting. Richard replied, "Oh, I could talk about it for hours. . . ." M asked Richard if there was a place where he could lift weights on campus. Richard said there was a weightroom, but only "hoops" could use it today. M then asked Richard what "hoops" was. Richard answered that "hoops" was basketball.

(Is the word homey, possibly derived from homeboy, somebody who is down or cool with another person? It seems to me that M, who apparently didn't know Richard, wanted to talk to him. In order to do that, he tried to let Richard know M thought he was a cool person? Homey appears to be applied regardless of ethnicity. . . . Their interaction appeared to be organized around interest in a common activity, weight lifting. Judging by the size of M's muscles, this was something he excelled in.)

This ethnographer has been noticing the ways blacks use the terms cool and down to refer to inclusion of non-blacks in their otherwise black groupings. In this commentary, he reflects on other terms which also seem to be inclusive.

In-process memos are products of more sustained analytic writing and require a more extended time-out from actively composing fieldnotes. Often they are written after completing the day's fieldnotes. Although perhaps touched off by thoughts generated by writing up a day's fieldnote entries, such memos address incidents across several sets of fieldnotes. In writing in-process memos, the fieldworker clearly envisions outside audiences and frames his thoughts and experiences in ways likely to interest them.

In-process memos can profitably address practical, methodological questions that include: Where should I observe next? What questions should I ask to follow up on this event? These questions help direct the ethnographer's attention, focusing and guiding future observations and analysis. Asking such questions helped a researcher in a battered women's shelter identify gaps in her understanding of how staff viewed and accomplished their work:

The goals staff have talked about so far of "conveying unconditional positive regard" for clients and "increasing their self-esteem" seem rather vague. How does the staff know when they have achieved unconditional positive regard? Is it based on their interaction with the client or by their refraining from being judgmental or critical of them during staff meetings? I will attempt to discover how they define and attempt to achieve the goal of "increasing a woman's self-esteem." It has been made clear that this goal is not only seen to be achieved when women leave their abusive relationships. If leaving their abusive partners were the primary indicator of achieving raised self-esteem, the organization would be largely unsuccessful, since most of these women go back to their abusive relationships. Yet, while I have learned what raising self-esteem is not, I have yet to learn what it is.

In this more extended series of comments and questions, the fieldworker identifies two matters that shelter staff members emphasize as goals in their relations with clients: "conveying unconditional positive regard" and increasing client "self-esteem." She then considers ways she might look to understand how these general policies/values are actually implemented and their success or failure practically assessed in interactions within the shelter.

In-process memos may elaborate new interpretations developed from subsequent incidents or understandings, as in the following:

Several weeks ago, I wrote about a client whom staff found to be quite aggravating and "annoying" because she had been continually calling the crisis line at all hours of the morning. At the time I had been under the impression that staff perceived such calls as unnecessary unless they pertained to immediate threats of physical injury. Through a conversation that took place today [included in earlier notes], I realize that this was an accurate but oversimplified notion. Although the staff finds late night crisis calls quite aggravating, they also acknowledge the necessity of maintaining such an option to deal primarily with violence of an immediate and physical nature. But even if the caller's situation does not fit into that category, she wouldn't necessarily be identified as a "nuisance" unless she had called repeatedly and had enough familiarity with the organization to know better. Each caller seems to be viewed as an individual case and is treated accordingly. It is only when their issues become too time consuming or chronic that are they identified as nuisance callers.

Here the student developed a more complex analysis by correcting and extending an earlier analytic claim. Writing helped her clarify her ideas and draw out subtle differences as she systematically examined new information in light of what she had previously understood.

As a general practice in writing commentaries and in-process memos, the field researcher should remain open-minded, avoiding conclusive an-

alytic statements in favor of possibilities and alternatives. These writings are most helpful when offering probing reflections, tentative musings, and open questions. For example, phrases such as "possibly," "apparently," "appeared to be" reflect these qualities in the field researcher's commentaries on the term "homey."

The field researcher may experience uncertainty and strain in deciding when to concentrate on writing thoroughly descriptive fieldnotes and when to record and develop analytic insights. Writing analytic commentary and in-process memos can easily displace time and effort needed for writing core descriptive fieldnotes. Yet we have often found that new ideas, like the descriptive details that make vivid fieldnotes, are fleeting; if not written down immediately, they tend to "get lost" or remain underdeveloped. So the field researcher may constantly have to balance the impulse to write down ideas and insights when they occur against the compulsion to "get it all down" as quickly and completely as possible without interruption.

In sum, ongoing reflection and analysis, even as the fieldworker continues to observe in the field and to actively write fieldnotes, is crucial for ethnographic research. Writing analytic asides, commentary, and in-process memos helps the field researcher carry forward analysis contemporaneously with the collection of field data. Such reflective writing often incites the researcher to pay closer attention to what she sees and, thus, to write more detailed and vivid descriptions. In-process analytic writing in turn increases the possibility of making the kinds of observations needed to develop and support a specific analysis. The more explicitly analytic themes are identified, the better able the fieldworker is to "check out" different alternatives, making and recording observations that can confirm, modify, or reject different interpretations. Theoretical asides and commentary, as well as more elaborate theoretical memos, enhance these prospects.

In writing fieldnotes, ethnographers have as their primary goal description rather than analysis. A researcher writes notes with a specific purpose in mind: to record a slice of life on a page. But these contrasting terms—description and analysis—refer more to recognized kinds of writing than to separate cognitive activities. In that sense, writing fieldnotes is a process

of “analysis-in-description.” Indeed, all descriptions are selective, purposed, angled, voiced, because they are authored. To “write up life” in this way, an ethnographer uses language conventions to create an envisioned scene. Sketches and episodes present a still-life or tell a story; they paint portraits of settings, people, and actions rather than offer causal explanations or build an argument.

All writing, by definition, is an abstracting and ordering process: clear writing always has internal coherence, the product of the writer’s attention to the subject as well as to the potential reader. Ethnographers construct their fieldnotes in a process more accurately captured by the expression “writing up” than “writing down” or “getting down” people’s doings and sayings. Writers do more than inscribe the world. Just as the ethnographer-as-observer participates with members in constructing a social reality, so too the ethnographer-as-writer creates that world through language.

In this chapter, we have seen that even though restricted to actual observed details, an ethnographer always “creates” the described scene. Writing fieldnotes processes experience not only through a researcher’s attention in the field but also through a writer’s memory and compositional choices at the desk. An ethnographer perceives interactions and selects significant details; in writing she groups these details into a coherent whole according to conventional writing strategies.¹⁷

Awareness of writing conventions, however, is not meant to lead a writer to be more craftily inventive through the use of persuasive rhetorical skills. Rather it invites the ethnographer to make more conscious choices when creating fieldnote records that portray social worlds as experienced and perceived by others. Consider the effects of writing: not only does a writer’s theoretical stance influence compositional choices, but the reverse also happens. Even by inadvertently imitating an “objective” social science style, for example, with its measured wording, omniscient viewpoint, and use of the passive voice, descriptions reflect an affinity—though ever so subtle—for that orientation. Certainly, a writing style tends to shape any writer’s vision. How researchers see in the field, in part, results from what they find noteworthy and “writable” as a fieldnote. Consequently, students concerned about research integrity must develop a conscientious respect for how their writing choices influence both fieldwork and note-taking.

Whether carefully or haphazardly written, every fieldnote mirrors an author’s choices: to include these details rather than those, to view actions

either through this character’s or that one’s eyes, to sequence actions in this way or that way (or randomly), or to write with particular readers in mind. These authorial choices, if only subliminal, result in on-the-page descriptions with certain kinds of detail, a particular point of view, a tone and attitude toward the persons depicted, and a distinctive voice. These day-to-day renderings of scenes pile up, and writing choices assume a cumulative effect: the notes portray that world through this particular writer’s lens. In making writing choices, therefore, how ethnographers write fieldnotes becomes as consequential for readers and those depicted as what they write. Whether as privately filed resources or as public excerpts in final documents, fieldnotes persuade.

All too frequently ethnographic fieldnotes fail to attend consistently to members' meanings, instead importing outside or exogenous categories and meanings. Imposition of outside categories produces fieldnote descriptions that fail to appreciate local meanings and concerns (Matza 1969:15–40) and that tend to frame events as what they are not (that is, by reference to categories or standards that differ from those recognized and used by members). In general, field researchers concerned with members' meanings are leery of any classifications which do not refer to the categories that the people recognize and actually use among themselves.

Failures to appreciate members' classifications arise from several sources. First, lapsing into classic ethnocentrism, researchers may take a category, standard, or meaning from one culture or locale and use it to describe events in another context. For example, based on their own expectations, Westerners in an African cinema or theatre might describe as

"disruptive" loud audience remarks to characters and thus fail to appreciate such participation as a locally appropriate way of expressing an evaluation of the performance. Or an observer may employ exogenous criteria to evaluate school classrooms as "noisy" or "chaotic," thereby ignoring teachers' and students' actual understandings of how classroom activities should be conducted. Both procedures caricature rather than describe behavior on its own terms.

Second, in a similar vein, ethnographers may use a term, category, or evaluation that is recognized, used, and honored by one group in a particular social world to describe features or behaviors of another group in that world. For example, psychiatric staff might interpret certain patients' behavior as "acting out" or "denial," even though the patients understand the actions as common, everyday behavior. Often a field researcher who comes across different local understandings of the same event has a tendency to accept one view as "true," thereby marginalizing competing versions. In one situation in Zambia, for example, a diviner-healer determined that an older man who suddenly could not walk had been bewitched and, after treating him for a year with medicines and massage, cured him. However, the medical doctor at the local hospital, on hearing the account and later meeting the man, concluded that he had had a stroke resulting in paralysis. In writing fieldnotes, a Western ethnographer might be tempted to privilege, though ever so subtly, the medical doctor's "scientific" account and then to describe the diviner's interpretation as "belief," thereby prioritizing one practitioner's explanations as more "accurate" and implicitly more efficacious.

Third, field researchers may adopt a dismissive attitude toward members' meanings, treating such meanings as flawed, hypocritical, contradictory, or fallacious. For example, a student ethnographer working in Los Angeles observed in weekly meetings of a study group devoted to the philosophy of Edgar Cayce. In the following fieldnote, she describes an incident recounted by a member to the group:

Dolores lost her purse and did not panic. She threw the white light around it and asked God to protect it. She also asked that no one be tempted to take her identification, credit cards, and money. The next day when she went to work, she asked the guard on duty if the purse had been turned in. Indeed it had and nothing had been displaced.

The student initially interpreted this story as indicating an extremely "passive" approach to the problems of daily life:

The moral of the story was to leave everything in God's hands. . . . To me visualizing the white light and talking to God are very passive ways of dealing with an emergency situation as opposed to going to the police or retracing one's steps.

This interpretation dismisses "casting the white light" and "talking to God" as ineffective nonactions at best and as pathological delusions at worst. It implicitly counters, and hence negates, the group member's claim that the purse had been turned in and nothing taken because of her prayerful actions. The result is to debunk the actual beliefs of this particular group about the efficacy of action in everyday life.¹

Fourth, fieldnote descriptions and memos may be framed in terms of a standard of what is "supposed to be" that derives from official rules or understandings that are held to govern action in some specific setting. For example, noting a discrepancy between an elder's account of the traditional meaning and sequences of a ritual and the actual performance of that ritual, an ethnographer might describe this ritual as "in decline" rather than as subject to adaptation and variation.² Similarly, an ethnographer might describe and analyze police action on the streets in terms of official regulations for the use of force; how actual police officers evaluate specific street situations and decide when and what kinds of force to use is thereby ignored.³ In both instances, ethnographers implicitly determine whether actions should count as conforming to or departing from the "traditional" version or "official" regulations and hence whether these actions are "in fact" "authentic" ritual behavior or a "legitimate" use of force.

Fifth, the researcher may invoke a priori theoretical categories, often those sacred to the core of a particular discipline, to characterize events and settings. For example, in studies of traditional narrating, past researchers relied heavily upon the analytic categories of "myth," "legend," and "folktale" even in explaining non-Western traditions. Since these categories often impose Eurocentric notions and thus misrepresent a people's storytelling traditions and practices, contemporary folklorists increasingly characterize storytelling with the indigenous terms and explanations of the group studied and describe how people use these terms in particular storytelling events.⁴

Indeed, a field researcher may implicitly impose such categories exactly in asking exogenous questions rooted in an a priori research agenda or theoretical framework. Not only might a researcher impose ideas when questioning an "informant," but she might also impose an inappro-

priate form of expression whose constraints distort responses. For example, a field researcher who asks for a list of ingredients in cures or discrete steps in a ritual may get arbitrary lists intended to please the researcher. Or, when asked questions imposing external analysis and itemization, people may offer "non-answers" such as "yes," "no," or "sometimes," especially if they usually describe these healing and ritual events by recounting the story of the experience.⁵ In sound ethnographic research, in contrast, "both questions and answers must be discovered from informants" (Spradley 1979:84).⁶

Finally, describing local settings or actions in terms of dichotomized variables may involve an imposition of exogenous categories. For an ethnographer to describe those present in a bar as "regulars" and "non-regulars," for example, may ignore a range of other, more variegated distinctions bar patrons may draw between one another. In general, the reduction of ongoing social life to variables and to dichotomies tends to produce a radical decontextualizing and destruction of local meanings.

In all these ways, ethnographers tend to produce fieldnotes and in-process memos that ignore, marginalize, and obscure indigenous understandings. In the following sections, we suggest alternative procedures for writing fieldnotes that avoid such impositions and that help develop descriptions and analyses that are sensitive to local concerns, meanings, and categories.

A number of distinct moments in group life highlight how members express, orient to, and create local meaning. Ethnographers begin to construct members' meanings by looking closely at what members say and do during such moments, paying particular attention to the words, phrases, and categories that members use in their everyday interactions.

Members' descriptions, stories, types, and theories, no matter how rich and evocative, provide only a starting point for ethnographic fieldnotes. Deeper, fuller memos and analyses in a final ethnography require examining not simply what terms members use, but when, where, and how they use them and how they actually categorize or classify events and objects in specific situations. ¶

By way of illustration, consider the following fieldnote provided by a student ethnographer with extensive gymnastic experience, in which he

identifies terms for those attending an "open gym night" at a local university campus:

At open gym nights there are different classes and subclasses of people attending. The major classes include the regulars, the visitors, and the walk-ins. Of these there are many subclasses too. In the regulars class, there are the novice, the ex-gymnasts (old-timers), and the advanced amateur. The novices are people that have never taken gymnastics, classes or lessons, and are people that just walked in one day due to interest. The advanced amateurs are people who were never on any gymnastics teams, but have taken classes or lessons, or used to be walk-ins. Finally, the old-timers are those who competed at either the high school or college level. . . . Walk-ins are students who have had a long term interest in gymnastics and would like to learn from the old-timers.

This description provides a typology of those coming to the gym: "Regulars" (subtyped into three further classes—"novices," "advanced amateurs," and "old-timers"), "visitors," and "walk-ins." But as it stands, this typology identifies only categories recognized by the ethnographer; it is not clear that people in this setting actually apply these categories to others (and themselves), and, if they were to do so, when, where, and under what circumstances. Thus, the problem with this typology is two-fold: we do not know whether or not members recognize and use terms such as "regular" and "walk-in"; and more fundamentally, we do not know, if they do use these terms, exactly how, when, and for what purposes.

In insisting on considering members' actual situated use of specific terms or categories, the issue is not the "validity" or "reality" of these categories in a conventional sense. Rather, any object or event can be categorized in multiple ways; and the fact that some objects/events might be classed in one way or another (e.g., on the basis of having this or that trait or attribute in common) is not adequate grounds for recommending a particular classification, since we can always invoke or imagine other traits that would produce very different sorts of classifications. ¶ Gym participants might indeed, at some times for some purposes, recognize "regulars," "visitors," and "walk-ins" as meaningful categories. But we cannot tell from this description, since no effort is made to look at how members actually talk about and identify others on specific occasions; that is, the types are presented without interactional context, as always and everywhere relevant. Since multiple classifications are always possible (Heritage 1984:144-50), the ethnographer should not impose classifications, even when starting with indigenous categories. Rather the ethnographer, alerted to possibly relevant local categories, should look closely at how

members actually classify events on particular occasions for particular purposes.

/ Ethnographic fieldnotes, then, should not simply report any indigenous terms discoverable in a setting. Fieldnotes should more fundamentally detail members' actual, situated uses of such terms. The following pages provide two extended examples of how field researchers' notes and other writings can be made more sensitive to the interactional uses of member-recognized categories.

Because they are committed to members' meanings and experience, ethnographers treat the relevance of gender, race, or class for everyday life in ways that differ significantly from common theoretical approaches which set forth a priori assumptions and definitions. Even though, like such theorists, the ethnographer may assume from the start that these are

significant matters that should always be attended to in understanding social life, she places priority on how people themselves deal with gender, ethnicity, and class within the dynamics of specific instances and situations.

This ethnographic stance toward issues of gender, ethnicity, and class has been criticized on several counts. One line of criticism insists that ethnographic research is uninformed by theories which might enable the fieldworker to transcend the limited view of specific events and allow her to write about more generalizable social forces. Another line of criticism holds that ethnographic treatments of gender, ethnicity, or class are narrowly restricted to empirical observations: that is, that ethnographies describe specific locales and situations as isolated from the broader social structures and forces that critically determine specific events and individual lives.

Certainly, both criticisms highlight areas in which an ethnographic approach to gender, ethnicity, and class differs from more encompassing theoretical approaches. Committed to members' meanings and experience, ethnographers are more attracted by what Geertz (1976) termed "experience-near" as opposed to "experience-far" concepts; thus, they generally give priority to these meanings over a priori, received theories. Valuing the local and the specific, field researchers look in a focused way at daily life rather than in a broad and sweeping manner at general patterns. Ethnographers certainly prefer to see the direct influence of social structures, rather than to assume their relevance and effects at the outset. At first glance, this "experience-near" approach seems to create tensions between ethnography and theories about the effects of broader social structures. However, some of these apparent tensions lessen, perhaps even disappear, by looking closely at several ways that ethnographers can and should bridge their commitment to members' meanings with their concern for gender, ethnicity, and class.

At the most basic level, the ethnographer with strong interests in gender, ethnicity, and/or class should carefully select a site for field research where he expects one or more of these processes to be particularly salient. In choosing a site, the researcher should look for a setting where gender, ethnic, or class diversity not only seems clearly highlighted but also where these issues concern the members. Examples would include police forces with increasing numbers of women or ethnic recruits, or schools with ethnically diverse student populations. In addition, a researcher might

choose to study events through which members directly address these issues. For example, an ethnographer interested in gender issues might research occasions in which elders teach the next generation. In many societies, initiation ceremonies explicitly focus on instructing youth about gender roles and responsibilities. Among the Chokwe in Zambia, such rituals as mwadi for girls after they begin menstruating and mukanda when boys are circumcised are central village events which provide explicit information about gender construction.

Indeed, an ethnographer not only can select a setting and events which focus directly on gender, ethnicity, or class, but she might also design a field research project exactly for its relevance to a theoretical issue derived from these concerns. Those researchers concerned with the reproduction of class distinctions in schooling, for example, might want to follow the progress of a group of working-class youth over time in one particular school (Willis 1977). Or, they might look specifically at the differential decision-making by school counselors in tracking working- and middle-class youth (Cicourel and Kitsuse 1963).

Once in a setting, the ethnographer's first concern should lie in exploring the significance of gender, ethnicity, or class matters for those studied. A first step in this direction requires paying close attention to any occasion upon which people explicitly talk about and/or act toward each other on the basis of race, gender, and/or class. For example, rather than assuming that ethnicity is invariably a causal factor producing a behavior or event, the ethnographer seeks to describe in detail any interaction in which ethnic identification becomes a matter of attention. In the following fieldnote, a student ethnographer describes what happened when an African-American high school teacher opened a discussion of white-black relations in an African-American history class:

Ms. B picked Dapo next. Dapo said that he had just moved to the Valley, Southland Hills. This comment drew a couple of "woo's." Dapo grinned. He said the area he moved to is a "white neighborhood." One time he was walking down the street by his house and passed a white child playing there. The child's parents saw Dapo and grabbed the kid and dragged it inside. Dapo was kind of laughing as he said this. He said he wanted to tell the people, "I'm Black, but I'm not going to kill you." Some classmates burst into laughter and talked among themselves. Dapo continued, "My parents are Creole. . . . They're all (lowers voice to an aside) 'you're not really Black.' My cousins have blue eyes and blonde hair and all that. . . ." He continues, his voice firming up, "I'm Black. I'm a Black person. . . . I'm proud to be Black."

This account conveys a number of dimensions and contradictions of ethnic identity that are meaningful to a high school student. For example, we see the complex tensions that exist between who his parents tell him he is ("You're not really Black") and who he is to those in this neighborhood and for this class ("I'm Black").

However, the significance people attribute to gender, race, or class may be difficult for ethnographers to document because people do not always directly reference them. On some occasions, an ethnographer may feel that people regularly act toward one another in "classed" or "gendered" ways; yet, they may not be able to pinpoint how this is so or to record specific scenes or actions in which these features are explicitly alluded to. It may thus be extremely difficult to identify and tease out these matters in writing fieldnotes. In other situations, a researcher might expect gender, race, or class to be important but find that members fail to acknowledge these factors. In such instances, the ethnographer must push beyond explicit use of relevant terms to make more systematic observations to identify patterns of activities that reflect the relevance of gender, ethnicity, or class.

For example, in her study of storytelling in a Chokwe village, Fretz was consistently told that "anyone may tell yishima." And, indeed, in most villages, men and women, adults and children told stories around the family firesides. But in her own research in the village of the highest chief, after one initial evening in which a woman narrated in the chota (the chief's pavilion), she could not get any woman to tell a story. With continued observation and reflection, she eventually realized that not only did the chief consistently dominate the storytelling, but he requested that all storytelling take place in his pavilion, a locale where men meet to talk and where women, if invited, participate by responding and singing. Thus, questions about women's roles in storytelling did not reveal the status and gender impact of "storytelling rights" in the pavilion because the answers to these questions were not linked to storytelling but to other relational and situational factors. Only repeated observation and comparison between similar situations finally led to an understanding of the complex web of situational, gender, and status influences working in this context.²⁴

On other occasions, specific talk by members will provide a useful starting place for further inquiry to trace out the relevance of ethnicity, gender, or class in wider realms of local life. For example, one group of students set out to study relations between students in an ethnically di-

verse school. One of the group came back with the following talk about different "groups" on campus:

Around the lunch table today, a bunch of guys who hang around together were talking. I thought they could help me understand the different groups at Central. They used a lot of terms I'd heard before to describe the kids. One guy talked about "trendy people," and how I could recognize a trendy person if I saw one. Someone else said there's "ballers," which are people who play basketball, and then there's "footballers," people who play football and then people who "kick" [hang around with] all the groups. And then there's "posses." They said a posse is a group of students who hang around together, kick it together, and they do it because it gives them a sense of belonging. One Black guy goes, "It's just a coincidence that all the people in my posse are Black." We were all laughing so hard. He goes, "No, no, we all come from the same neighborhood. Some of them are interracial." Then there's the "swim team," those are the druggies because they use so many drugs that their eyes are always bloodshot so it's like they were swimming. Then there's the "GCP," the Green Card Posse, they go, "Oh, the Wet-backs." A "cool" person wears nice clothes. I asked, What if you don't have money for nice clothes, does that mean you're not cool? They said, If you have a good personality. But if your personality is the same way you dress, then forget it. "This place is a fashion show."

Here we see that students invoke a range of local categories in distinguishing and categorizing one another. Some of these categories make direct and explicit reference to ethnicity, e.g., "the Green Card Posse." Ethnicity is also directly mentioned with reference to "posses," but specifically how it is relevant appears more open: one speaker identified his posse as all black; another claimed that some posses are "interracial." In contrast, another speaker minimizes ethnicity as the basis for group formation. Finally, most of the categories are not explicitly identified with any particular ethnicity, e.g., "ballers," "druggies," "cool" people. An ethnographer in this setting would want to follow up and seek to establish the ethnicities of students identified as belonging to each of these various categories. This inquiry would be primarily a matter of observing the ethnic status of students identified with each category, then perhaps talking to students about observed ethnic patterns.

An ethnographer could also use this incident as a starting place for tracing out connections between these student categories and gender or class. It appears that this talk about groupings occurs among and is about boys; but the field researcher would want to find out specifically if any of these categories include girls and to ask further questions about similar or different groupings among girl students at the school. Here in particu-

lar the ethnographer should trace out gendered patterns of segregation and difference, as well as of integration and overlap, among students and their activities.

Handling issues of social class is more complex. Compared to gender and ethnicity, class is an "experience-far" rather than an "experience-near" concept (Geertz 1976). As a result, ethnographers rarely encounter members explicitly talking about "class" per se. But people employ a number of terms that refer to elements or components of the concept of social class. This instance, for example, directly involves a kind of naturally occurring "ranking" of persons that mirrors one concern of the social class concept. Furthermore, these students discuss "money," "nice clothes," and the school as a "fashion show," suggesting that parental income and conspicuous consumption might bear on how one is categorized within the school. Thus the field researcher might further question and observe these matters in order to describe what students consider "nice clothes," the care they take to display them, and where these clothes and the money needed to purchase them come from.

Finally, field researchers can employ a number of different strategies to try to link local events and specific outcomes to more distant social settings and more remote consequences. Initially, the ethnographer can look for specific connections within the setting to outside social influences. The ethnographer should attend foremost to how the people involved talk about and understand their connections with these outside entities and forces, but he would not be limited to these member-recognized understandings. Field research on the homeless, for example, might well begin by examining how people living on the street understand and cope with the conditions of their daily existence on a day-to-day basis, including how they see their relation to the wider society (e.g., Snow and Anderson 1993). But the researcher would also observe the various persons, agents, and institutions that have recurrent contact with the homeless: for example, missions, hotels, and other places that provide occasional residence; regular feeding lines and informal arrangements with restaurants as sources of food; relations with police patrol officers and jailers; caretaker agencies and welfare/relief workers.²⁵ And then a researcher (or

other researchers) might move out to examine these institutions and agents and their conditions of existence.

Ethnography can also explore links to broader social processes by observing people and settings as they change over time. Long-term, continuous field research is necessary, for example, in order to understand how working-class youth react to and are affected by their contacts with schools. Introducing a longitudinal component to field research, while practically difficult, allows the researcher to describe different life chances and to understand how these chances are shaped and determined. Field researchers, for example, often examine particular "institutional careers" (Goffman 1961) and the factors that shape them, whether these involve moving through schools to different final outcomes or through processing by the police or courts to different fates. The limited "breadth" of many ethnographies can be attributed directly to a lack of concern with change over longer time spans, to an assumption that one can look atemporally at one piece of life without considering changes over time.

In this chapter we have proposed strategies for writing ethnographic fieldnotes that collect and represent members' meanings in a rigorous, grounded manner. These strategies require the ethnographer to bracket preconceptions about what is important in order to attend to people's indigenous ways of ordering and interpreting their worlds. In so doing ethnographers assume that members' meanings are consequential, that how people act is based on their understandings of their local social worlds.

In pursuing members' meanings, ethnographers begin by looking at how members describe and categorize people and events; they try to discern their terms, phrasings, classifications, and theories. But indigenous categories provide only a starting point; the ethnographer's task is not simply to identify member-recognized terms and categories but also to specify the conditions under which people actually invoke and apply such terms in interaction with others. For no term or category is self-applying to actual situations; its relevance to specific circumstances cannot be determined in advance. Hence the ethnographer should not describe social scenes by applying member-recognized terms and categories to situations indepen-

dently of members' actual applications. Indigenous terms and categories must be indigenously invoked and fitted to specific situations. In writing about those in the open gymnastics session discussed previously, the ethnographer who identifies someone as a "regular" in his fieldnotes independently of actual (interactional) treatment by members of that setting has appropriated the power of fitting categories to specific situations.

Several implications flow from recognizing that the ethnographer who writes fieldnotes about indigenous meanings should specify the conditions under which members' meanings are invoked and applied. First, such fieldnotes must incorporate not words and phrases abstracted out of context but the actual interactional occasions in which these members' terms are used. Fieldnotes useful to appreciating members' meanings, then, will be interactionally rather than cognitively focused; they will document not how members talk about various social objects, in general and out of context, but how members construct meaning through interactions with other members of the group, how they actually interpret and organize their own and others' actions.)

Some methodological implications follow. Many ethnographers seem to assume that the pursuit of members' meanings is equivalent to interviewing people about what is important to them. But ethnographers collect materials relevant to members' meanings by focusing not on members' decontextualized talk but on naturally occurring, situated interaction in which local meanings are created and sustained. Writing ethnographic fieldnotes that are sensitive to members' meanings is primarily a matter not of asking but of inferring what people are concerned with from the specific ways in which they talk and act in a variety of natural settings. Thus, interviewing, especially asking members directly what terms mean to them or what is important or significant to them, is not the primary tool for getting at members' meanings. Rather, the distinctive procedure is to observe and record naturally occurring talk and interaction. It may indeed be useful or essential to interview members about the use and meaning of specific local terms and phrases, but the researcher's deeper concern lies in the actual, situated use of those terms in ordinary interaction.²⁶

Finally, focusing on interactionally situated uses of indigenous terms heightens the ethnographer's sensitivity to the intricate processes of situated judgment and skilled interpretation that characterize members' use of local categories. Members' categorizations are not invariant and transcendent but are tied to specific situations for varying purposes. Extensive

local knowledge and judgment-making skill are necessarily involved in their use. In the gym, for example, those about to undertake a particular gymnastic routine requiring a "spotter" may have a practical interest in recognizing and distinguishing between the experience and skill level of others present. Indeed, experienced gymnasts can see at a glance how much training another has had, on the basis of her performance and actions. In general, a deeper appreciation of indigenous meanings requires learning when and how members actually make such assessments and what knowledge they rely on in so doing.

The ethnographer begins concentrated analysis and writing by reading her fieldnotes in a new manner, looking closely and systematically at what has been observed and recorded. In so doing, she treats the fieldnotes as a data set, reviewing, reexperiencing, and reexamining everything that has been written down, while self-consciously seeking to identify themes, patterns, and variations within this record.

We strongly recommend reading line-by-line through as many pages of fieldnotes as possible, at least until coding seems to generate no new ideas, themes, or issues. Reading notes as a whole and in the order they were written confers a number of benefits. First, the fieldworker can perceive changes in her relations with those in the field over time. The grad-

ual movement from distance to rapport, for example, may only become apparent when reading in a matter of hours a record of events which took place over weeks and months. Second, the ethnographer gains fresh insights as she changes in her own understanding and interpretation of people and events by reviewing the completed set of notes. Based upon what has subsequently been learned, initial interpretations and commentaries now reencountered may seem naive or erroneous. This contrast between initial and later understanding is often striking when working in a totally unfamiliar culture and language. The fieldworker begins to reinterpret the import and significance of events and actions in novel ways. She may feel that foreign concepts and terms have no equivalent in English. And patterns and tendencies recognized when reading all of the notes may suggest alternative interpretations of actions or talk previously understood in another way. Finally, working with the complete set of fieldnotes allows the ethnographer to take in for the first time, in a relatively concentrated time stretch, everything that she has been able to observe and record. Reading notes as a whole encourages recognizing patterns and making comparisons. She begins to notice how an incident is like others in previously reviewed notes. Conversely, she also begins to note important differences between incidents previously seen as similar.

To undertake an analytically motivated reading of one's fieldnotes requires the ethnographer to approach her notes as if they had been written by a stranger. Indeed, many fieldworkers find it difficult to achieve the sort of emotional distance required to subject to analysis those with whom she has been deeply immersed. Some fieldworkers report discomfort at "examining under a microscope" the lives of people with whom they have become deeply involved and in many cases care about. For some, analysis comes close to an act of betrayal; many fieldworkers report having taken several weeks or months after they stopped writing fieldnotes before they could begin their analyses. Indeed, a number of ethnographers find that relations with those in the setting have become primary. In that case the ethnographer may set the project aside for years or even abandon it altogether. Some researchers resolve this internal conflict by working collaboratively with people in the setting, even occasionally co-authoring their writing with a local assistant.

Although the deliberate and self-conscious analysis ethnography entails may contribute to feelings of estrangement, it may be helpful to remember that making sense of "what's going on" is an activity that members of the setting engage in and that it is one of the usual and expected

activities of social life. And it is also sometimes helpful to remember that while our analysis of patterns of social life in the field site is ordinarily for audiences and purposes outside of it, we seek to convey an appreciative understanding of the world and lives of persons under study.

Coding begins with the ethnographer mentally asking questions of specific pieces of fieldnote data. In asking such questions, the ethnographer draws on a wide variety of resources, including direct experience of life and events in the setting; sensitivity toward the concerns and orientations of members; memory of other specific incidents described elsewhere in one's notes; one's own prior experience and insights gained in other settings; and the concepts and orientation provided by one's own profession or discipline. Nothing is out of bounds!

But the secret of coding lies in turning the answers to these questions into a distinctive kind of writing—a word or short phrase that captures and signals what is going on in a piece of data in a way that links it to some more general analytic issue. Such writing is integrally linked to the processes of thinking and interpreting whereby the ethnographer “comes up with” a code to write down. In turn, writing down codes—putting an idea or intuition into a concrete, relatively concise word or phrase—helps stimulate, shape, and constrain the fieldworker's thinking and reflection. This mutually necessary relationship between reflection and writing is expressed in John Forester's (n.d.) apt phrase, “thinking with your fingers.”

We have found the following sorts of questions useful in beginning to examine specific fieldnotes:

Such questions reflect and advance several specific concerns linked to our approach to ethnography and writing fieldnotes. First, these questions give priority to processes rather than to “causes” or internal psychological

“motives.” Specifically, this priority means asking questions that identify what is occurring and in what order, rather than “why” questions that ask what is causing or producing some outcomes. In this sense, we view open coding as a means for developing interpretations or analytic themes rather than causal explanations.

Second, these questions reflect a sensitivity to the practical concerns, conditions, and constraints that actors confront and deal with in their everyday lives and actions. This concern with the practical or the pragmatic requires paying attention to the mundane, ordinary, and taken-for-granted rather than looking only or primarily at the dramatic or exceptional action or event.

Finally, these questions can help specify the meanings and points of view of those under study. We try to frame questions that get at how members see and experience events, at what they view as important and significant, at how they describe, classify, analyze, and evaluate their own and others' situations and activities. Yet, to get at these matters, it is initially crucial to clarify what the ethnographer felt was significant about what occurred by asking: “Why did I include this item in my fieldnotes?” It is then important to ask whether or not and on what basis members seem to attribute this same significance to events or incidents. These procedures keep the ethnographer aware of the complexities involved in pursuing members' meanings; in other words, they remind the ethnographer that there is no “pure” way to capture what is important to members, their meanings or points of view. Rather, the ethnographer always writes her interpretation of what she feels is meaningful and important to them.

Such questions will lead to codes which the ethnographer writes in the margins of his fieldnotes. The following example, from a student whose ethnography examined her work as an usher, illustrates this process:

customer types: late arrivals

Dance audiences do tend to come right at curtain, so we have to hold many out. Tonight was no different. I'd say we had about 50 people waiting in the lobby through the first number. . . . One man we held out was irate. He had already been in but had come out for some reason. When we closed the door, he began yelling at the door attendant. He said he was already in—not like these other people who were “LATE.” He was not late and shouldn't be treated like them! The house manager came over and smiled and told him in a quiet voice why he was being held out—that it was requested by the

holding out audience members

waiters: irate

latecomer claims exception

mgr intervenes

passing the buck

calming latecomers
keeping occupied
distracting

smiling
minimizing the wait

dancers. He calmed down but was still angry. He waited without another word, except when I came around. I went around giving out programs so they could read something and so the ushers wouldn't have to waste time doing that when these people charged their doors. I also asked people if I could tell them which aisle to go to so as to alleviate confusion for the door attendant. When I got to this man and asked him if he wanted me to tell him which door to enter through, he said in a huff that he had already been in and knew where to go. Other people were just as irritated. I just smiled and told them it would just be a few minutes. I think that calmed them a little, because the exasperated look left their faces.

This student ethnographer focused on the practical situation of ushers, implicitly asking how ushers understood and made sense of behavior and events and how they interacted with one another and with customers. Specifically, in the code "late arrivals" she identified the practical work consequences of the inevitable tendency for some ticket holders to come late. Furthermore, the codes "holding out audience members" and "calming latecomers" identify specific processes for dealing with and managing latecomers as practical work problems. The ethnographer then asked herself how these activities were actually done by ushers, which led to a series of more specific codes for "calming," e.g., "keeping occupied," "distracting," "smiling," "minimizing the wait."

These codes begin to identify and elaborate a variety of analytic distinctions. For example, the code "late arrivals" names a particular "type of customer"; in framing "late arrivals" as a "type," she asserts that coming late is a normal, routine event in this setting and that "late arrivals" are one among a range of customer types. In identifying one customer type, this code raises the possibility that other customer types exist and hence opens the question of just what these other "customer types" might be. That is, the process here is a dialectical one that consists of asking, "Of what more general category is this an instance?" In answering this question, the field researcher may draw upon a wide variety of experiences and different sorts of knowledge: her own experience as an usher, her awareness that dealing with people who come late is a practical matter that ushers must routinely confront, her experiences as someone who has

come late to a performance, and her familiarity with sociological thinking about waiting as a key to power differences (e.g., Schwartz 1975).

But while latecomers are expected at dance performances, the code "irate waiters" distinguishes an audience type, a latecomer who is a source of trouble and special concern. The code "latecomer claims exception" identifies both the responses with which ushers have to deal and the categories and distinctions advanced by this particular latecomer. The next codes—"mgr intervenes," "passing the buck," "keeping occupied," and "distracting"—identify additional forms of "backup" responses. These responses include the manager's efforts to placate the disgruntled patron, and the writer's attempts to take waiting audience members' minds off the delay.

Codes, then, take a specific event, incident, or feature and relate it to other events, incidents, or features, implicitly distinguishing this one from others. By comparing this event with "like" others, one can begin to identify more general analytic dimensions or categories. One can do this by asking what more general category this event belongs to, or by thinking about specific contrasts to the current event. For example, the response of "holding out" customers would stimulate a concern with the reverse situation (e.g., "taking latecomers in during a performance") and, hence, would suggest looking for observations describing how this would have to be managed.

While many of the codes used here involve members' concerns and specific terms, we also see attention to members' meanings in the code "latecomer claims exemption." This code tries to capture the actual distinction that this audience member advances in trying to get back in to see the performance—that some people arrived after the show had begun, but he had arrived before, was now trying to reenter and, therefore, was "not late" and should be treated differently than those in the first category. In the staff response, we see the practical irrelevance of this distinction; to the staff, what presumably matters are not considerations of justice and fairness (such that "real latecomers" should be treated differently from those who had to leave momentarily and hence were returning) but the disruption that would be caused by anyone entering at this time.

Through an initial line-by-line reading of her fieldnotes, this student began to clarify the socially ordered work activities of an usher for dance audiences. As she continues through her notes, asking the question, "What are the processes by which the ushers accomplish their work?" she will generate more codes; some will be further instances or elabora-

tions of earlier codes, while others will suggest entirely different themes and lines of analysis. Having identified, for example, "latecomers" as one group of dance audience members, she will proceed through her notes looking for other member types or categories. Similarly, having a code, "waiters: irate," implying that getting irate is only one response in the general category of audience responses, she could go on to look for others. She could also wonder: this goes on here, but does it always to on? What are the conditions under which it occurs?

Similarly, the student may identify an order or natural sequence of events or stages that make up the larger activity. She can further develop themes along these lines by continuing to look for expected or routine events that are problematic at each stage and the kinds of skills and practices used to respond to them. For example, the strategies noted in the codes—"keeping [customers] occupied," "distracting," and "smiling"—suggest that she look for further instances to illustrate the general issue of ways that ushers manage, respond, control, or cope with different types of audience members.

In creating codes, the fieldworker is engaged in an analytic process; she seeks to move beyond the particular event or situation in the fieldnotes to capture some more general theoretical dimension or issue. While it is often useful to begin coding by focusing on a term in the notes—whether the fieldworker's or a member's—the fieldworker must transform that term so that it references a more general category. Yet, at the other extreme, it is not useful to use very general categories as codes. For example, it would not be helpful to code as "social control" staff procedures for searching residents' rooms for "buzzes" and other contraband in a reform school. This category is too general and without specific connection to the events described in the notes. But, a code like "staff control—room searches" would categorize these staff activities as a specific kind of control and perhaps stimulate the field researcher to think about and identify other forms of "staff control."

While subjecting fieldnotes to this careful, minute reading, the ethnographer begins to sift through and categorize small segments of the fieldnote record by writing words and phrases that identify and name specific analytic dimensions and categories, as these are suggested by the recorded

observations. Such codings can be written in the margin next to the pertinent fieldnote, on a separate sheet of paper (with some marking of the location of the relevant fieldnote), or in a "comment" field in some word processing programs or a keyword field in a text database. In such line-by-line coding, the ethnographer entertains all analytic possibilities; he attempts to capture as many ideas and themes as time allows but always stays close to what has been written down in the fieldnote. He does so without regard for how or whether ideas and categories will ultimately be used, whether other relevant observations have been made, or how they will fit together.

Coding fieldnotes in this way differs fundamentally from coding in quantitative research. In quantitative coding, the researcher proceeds deductively by constructing questionnaires with categories derived from theory. He fits peoples' responses to the questionnaire into the already established categories in order to determine the frequencies of events within those categories. Qualitative research proceeds inductively by writing fieldnotes that reflect the significance of events and experiences to those in the setting.² Qualitative coding is a way of opening up avenues of inquiry: the researcher identifies and develops concepts and analytic insights through close examination of and reflection on fieldnote data. Such coding is not fundamentally directed at putting labels on bits and pieces of data so that what "goes together" can be collected in a single category; the ethnographer is indeed interested in categories, but less as a way to sort data than as a way to name, distinguish, and identify the conceptual import and significance of particular observations. In contrast to quantitative coding, then, in qualitative coding we identify, elaborate, and refine analytic insights from and for the interpretation of data.³

This view of qualitative coding implies that the same set of fieldnotes could be coded differently by ethnographers with different theoretical sensitivities and commitments. Disciplinary background and interests in particular will exert a deep influence on analytic coding: anthropologists working with the concept of culture, for example, might formulate different analytic categories than folklorists interested in performance and the dynamics of performer-audience interaction. Theoretical differences within a discipline may produce almost as marked variations in coding. For example, two sociological field researchers studying households might well write and code their fieldnotes quite differently (even, we would argue, were they to carry out their studies in the same setting); one might focus her coding on the consequences of economic policy on

household relations and the division of labor, the other on women's invisible work in families. In sum, there is no single, correct way to code fieldnotes inasmuch as ethnographers ultimately decide which among a number of possible patterns and ideas, including member concerns and interests, to choose as a focus.

In open coding, the ethnographer should not use preestablished categories to read fieldnotes; rather he should read with an eye toward identifying events described in the notes that could themselves become the basis of categorization. Nor should code categories be avoided because they do not fit with the fieldworker's initial "focus"; this focus will change as the researcher moves through the notes. Rather, the ethnographer should seek to generate as many codes as possible, at least initially, without considering possible relevance either to established concepts in one's discipline or to a primary theoretical focus for analyzing and organizing this ethnography. All ideas and concepts that can be linked to or generated from specific fieldnotes should be treated as of possible interest and should be framed and expressed as clearly and explicitly as possible. Hence, any particular code category need not necessarily connect with other codings or with other field data; integrating categories can come later, and one should not ignore or disregard codings because they suggest no obvious prospects for integration within a major focus or with other emerging categories.

By way of illustration, consider the following open coding of an incident from a support group for those taking care of family members afflicted with Alzheimer's disease:

trouble: memory loss;

bad driving

dr does not "help"

asks advice

fam pressures dr

med test → no results

don't rely on dr

cger to DMV

Lucie says her husband is in good health, but his symptoms include memory loss and poor and dangerous driving. The doctor does nothing to stop him from driving. She asks, "What does everyone else think?" Some other members say, "Change doctors." Lucie explains the doctor is a friend of the family. Her son has stressed to the doctor that his father's driving is dangerous and they could be legally involved. The doctor has done a catscan but there is no direction from that.

Pat, the group leader, recommends, "Take it into your own hands." She suggests that Lucie go to the DMV. Lou says she thinks there is a new law that states anyone with a mental deficiency, including Alzheimer's disease, is not supposed to drive. Lucie

no med dx

advice: coalition w/dr

practical remedy: deception

proposed remedy will not work

"talking to"

says, "But I don't have a name on it—that's what hinders action. I am so frustrated."

Vie says, "Isn't it important for the doctor to tell him not to drive?" Lucie says, "Why won't he do that? Maybe he's too close and he doesn't want to get involved." Lou: "What about Nicholson? He's a geriatric psychiatrist." Others suggest that she hide the car keys. Joey says, "You need to lie to him." Lucie says, "I must say I have been doing that." Joey says, "We all have." . . . Lucie says in terms of the car keys, he knows there is a second set. Another woman says she talked with her husband and he doesn't drive anymore. "I've done this. It is not working." Someone says, "You need a good diagnosis from a medical doctor." Lucie: "That's what I think." Others in the group agree.

Through these marginal codes the fieldworker has identified a variety of loosely related (or even unrelated) issues:

Some of these codes reflect issues that the field researcher was interested in from the start: practical "troubles" and how people respond to or "remedy" such troubles (see Emerson and Messinger 1977). But many of these codes elaborate or specify a prior concept in original and unanticipated ways, e.g., "hiding the keys" as a practical remedy for dangerous driving. Other codes identify issues that are entirely unexpected; for example, doctors as both barriers and possible allies in handling unfit rivers.

By the time the ethnographer finishes reading the complete set of fieldnotes, her categories and themes will have fundamentally changed. And many of those categories will be dropped, in turn, as the researcher becomes more focused and aware of other, more interesting and recurrent

issues. Furthermore, the process of generating codes may help clarify the meaning or import of previous as well as upcoming notes, for coding shapes and may alter the fieldworker's sense of what the notes "contained" in the first place. As one student commented: "You feel you know your notes because you wrote them, but the thing is, you wrote them so long ago that it doesn't click."

Many students report that the evolving, seemingly unending character of coding initially proved discouraging and upsetting:

The coding process, it happened once and then it happened again. I ended up coding again and again and again. . . I had to get over the fact that I would do it the wrong way, or I wouldn't really find any good categories or things wouldn't relate to one another. I had to get over the fear of thinking that there was nothing there.

Coding is indeed uncertain, since it is a matter not simply of "discovering" what is in the data but more creatively of linking up specific events and observations to more general analytic categories and issues. Though researchers inevitably draw on concepts from their particular disciplines to develop linkages, coding keeps them focused on and anchored in their data. Often the researcher is already familiar with the key concepts and interests of her discipline and quickly sees how a given piece of data is relevant to them; but at other times the researcher may have to turn to specific writings that she has not previously read to find pertinent concepts. With time, practice, and wider exposure within a discipline, the researcher gains confidence that she can make analytic connections, and coding becomes less threatening and uncertain.

This open-ended approach can lead to anxiety on several different levels: some students fear they may never come up with a specific focus for a paper. Others, finding line-by-line coding time consuming and tedious, want to focus on a smaller number of themes in order to move ahead quickly, without a lot of "wasted" effort. Still others express concern over a procedure that, in seeking to generate so many different codes, contradicts what they have been taught about "logical" (i.e., carefully planned in advance) thinking and writing. Consider the comments of two students:

I didn't have any categories before I began. I just was looking at the notes and jotting down codes, but it didn't seem that I was going about it in a very logical way.

I went through two or three sets of notes and there were so many random, recurring themes and not anything that was organized.

But the fact that fieldnotes seem unwieldy, with codings leading in many different directions, is actually a good thing at this stage; such codings will suggest a myriad of possible issues and directions. Especially early on in the process of open coding, we recommend resisting these inclinations to focus on specific themes and topics while continuing to go through the fieldnote record and generating additional codes.

Yet, we have also found that continuous open coding can generate a great deal of frustration as ideas begin to coalesce; continuous open coding may actually discourage developing a specific focus when it would be possible and useful to do so. A strategy of selective open coding, in which the fieldworker uses these procedures at different times and with discrete sets of fieldnotes, may therefore be advisable. For example, one may begin with systematic open coding, but then after going through a significant portion, code remaining notes and recode previously coded notes selectively, focusing on "key," "rich," or "revealing" incidents.

Inspired by reading and coding fieldnotes, the fieldworker begins to entertain a wide variety of ideas and insights about what is going on in the data. He can preserve and elaborate these ideas by writing initial theoretical memos. We encourage writing memos about as many ideas, issues, and leads as possible. While some of these ideas reflect concerns and insights the fieldworker brings to the reading, others grow out of reengaging the scenes and events described in the fieldnotes.

Memos written while coding fieldnotes are generally more analytic than in-process commentaries. At times, it is helpful to take a specific, "rich" fieldnote and explore its theoretical implications. The ethnographer studying family members caring for persons with Alzheimer's disease composed the following memo as a series of "observations" on a single, brief but "suggestive," fieldnote excerpt:

Fieldnote: During the support group Fumiko comments on her husband's behavior: "Once in a while he is a pussycat" (laughter), "but he was a raging bull when the VNA came to give him a bath." She adds that recently he has fought her shaving him, but "this morning he let me do it."

Memo: Note how this description suggests that caregivers recognize that cooperation can vary independently of ability or condition for the person with Alzheimer's. Thus, it is one issue whether or not the person with Alzheimer's can feed or bathe him/herself, shave himself, etc.; the stance the person with Alzheimer's takes toward these helping/caring for activities is another matter.

Note also how unpredictable these matters may be for the caregiver; bathing and shaving go smoothly on some occasions, but produce major hassles on others. And the caregiver does not seem able to find a reason or explanation for when and why one outcome rather than another occurs.

Furthermore, it may well be uncooperativeness or resistance in caregiving matters, rather than the amount or kind of help per se, that generates critical problems and burdens for caregivers. In this respect, the core of a caregiving management regime may rest in those devices and practices that inhibit, overcome, or sidestep resistance. With someone with Alzheimer's who is cooperative (or nonresistant)—in most matters—the caregiver can say: "I can still guide him." Similarly, a person with Alzheimer's who is cooperative is one who can be "talked to," i.e. convinced to make changes in his/her daily life, more or less "voluntarily."

In this memo the fieldworker identifies two initial, somewhat unrelated issues in the fieldnote: some caregivers report that patient cooperation can vary independently of physical condition and that cooperation may wax or wane unpredictably. In the final paragraph, she speculates on the possible relevance of one of these issues—cooperation (and its counterpart, resistance)—in shaping the broader pattern and course of family caregiving for persons with Alzheimer's disease.

At other times, the ethnographer can use an initial memo to try to name and specify a particular analytic issue that cuts across a number of particular incidents. Here, for example, one might try to identify and explore a general pattern or theme, drawing upon and attempting to link a number of disparate incidents or events. Along these lines, consider the following memo from a study of support and interaction among courtroom personnel (clerks, recorder, bailiff), which explores patterns of "sustaining community and insideness" in courtroom proceedings:

Examples of "sustaining community and insideness" tend to occur during dead time (recess), on easy days with little business and also after session ends for the day. . . . For example, after today's session, all of the participants except the judge, who always leaves, were actively looking for interactions. Their methods included making eye contact with each other, walking toward each other, making jokes, and interrupting conversations. In this way, information could be shared and opinions could be aired.

This category can be distinguished from idle chatter during recess by the involvement of the participants in the events. High involvement equals commu-

nity and insideness; low involvement, which is evidenced by briefness of interaction and lack of emotion and eye contact, equals idle chatter.

Here the field researcher identifies a regular pattern of more intense, animated talk and action between courtroom workers which she contrasts with other occasions of less engaging interaction ("idle chatter"). In writing she offers some initial observations on when this pattern of relating occurs (during recesses, on slow court days, etc.), as well as on what it involves (actively seeking out others, joking, etc.).

In sum, initial coding and memoing require the ethnographer to step back from the field setting to identify, develop, and modify broader analytic themes and arguments. Early on, these efforts should remain flexible and open, as the ethnographer reads, codes, and analyzes fieldnotes to foster a wide range of new ideas, linkages, and connections. Eventually, however, the ethnographer will move beyond these open, inclusive procedures to pursue more focused, analytic themes more intensively. Initially, this narrowing and focusing process involves selecting a small number of core themes which the researcher will subsequently pursue through focused coding and integrative memoing.

Through initial coding and memoing the ethnographer identifies many more ideas and themes than she will actually be able to pursue in one paper or monograph. Hence she must decide which ideas to explore further and which to put on the back burner, at least for the moment.

Field researchers have different ways of selecting core themes. One consideration is to give priority to topics on which a substantial amount of data has been collected and which reflect recurrent or underlying patterns of activities in the setting under study. Fieldworkers may also give priority to what seems significant to members, whether it is what they think is key, what looks to be practically important, or what engages a lot of their time and energy. For example, one student who wrote fieldnotes while an intern at a county probation office described the following process:

I was going through [the notes] and kept thinking of things like we have all this paper work to do, and people have to sign this and that and I started to get the sense of this larger issue—how is the department dealing with so much paper

work? And as I went through it, I found, "Oh, well, a lot of times we help each other out." One probation officer will say I saw your client yesterday on the Commons; that will count as a collateral contact [a kind of contact that must be noted in the paperwork] for you because I saw him. There are shortcuts that way. There are summary reports called "quarterlies" that summarize basically three or four months' worth of work into one sheet. So three or four things like that are subtopics of this larger issue.

This student, in going through her notes, began to notice the different tasks that probation officers must accomplish with a sensitivity to the conditions and constraints that accompany the work. Looking at what probation officers actually did amid the practical constraints and opportunities offered by other agencies—police, clinics, etc.—provided a frame for drawing together what had initially seemed like discrete tasks. Discovering additional themes of this sort provided a guide to reading and coding the rest of her notes.

The fieldworker must also consider how a selected theme can be related to other apparent themes. A theme that allows the researcher to make linkages to other issues noted in the data is particularly promising. Finding new ways of linking themes together allows for the possibility that some of the themes that might have been seen as unrelated and possibly dropped can in fact be reincorporated as "subthemes."

In the process of identifying promising themes and trying to work out possible linkages, the fieldworker may lose for the moment a sense of focus and have to rework ideas until she can reclarify matters. A student who studied the band at a public high school started coding with a good sense of what her paper would be about only to find her direction changed. She reflected on these processes in an interview:

I first thought I would explain how, in the face of budget cuts, somebody could keep a program, an extracurricular program like this going. And then in listing the ways that the teacher does that, I came across the idea that he has to do things to get all of these kids to be friends together. And then I thought, wait a minute, that could be a whole topic of its own. There's so many things going on. How do I explain in my paper the different social cliques with 110 kids; there's so many social cliques? And then I just started looking at the relationships that students have with each other inside band and outside. It was just the weirdest thing—I lost my paper! The more I coded, the more I lost my paper.

Eventually, this student shifted her focus from the many differences between social cliques to how the teacher kept the program going both in the face of budgetary cuts and the divisive tendencies of these different

cliques. What was reported negatively as "having lost her paper" really indicates an openness to new meanings and ways of putting things together.

Students engaged in this process often talk about a particular theme "jumping out at them" or, alternatively, of the "focus" for the ethnography "disappearing." This experience is so strong and pervasive that it is important to recall two closely related issues that were touched on previously. First, while the ethnographer's experience is often one of "something going on in the notes," neither the fieldnotes nor their meanings are something "out there" to be engaged after they are written. Rather, as creator of the notes in the first place, the ethnographer has been creating and discovering the meaning of and in the notes all along. Particular sensitivities led to writing about some topics rather than others; these sensitivities may derive from personal commitments and feelings as well as from insights gained from one's discipline and its literature and/or the course instructor. Second, when an ethnographer thinks he has "a substantial amount of data" on a topic, it is not so much because of something inherent in the data; rather it is because the ethnographer has interpreted, organized, and brought the data to bear on the topic.

Once the ethnographer has identified a set of core themes for further analysis, he may find it useful to sort fieldnotes on the basis of these themes. Here, the fieldworker breaks down the corpus of fieldnotes into smaller, more manageable sets, collecting together, in one place, all those pieces that bear on each core issue. This sorting or retrieving procedure involves physically grouping segments of the data on a theme in order to more easily explore their meanings. Sorting into one place or pile facilitates analysis by concentrating fieldnotes relevant to an emerging issue.⁴

In sorting fieldnotes, it is advisable to use themes that are inclusive, allowing for notes that may have been identified with different but related codes to be grouped together. In the study of family caregiving for persons with Alzheimer's disease, for example, after extensive open coding the researcher decided upon "management practices" as one core theme. Management practices were to include any actions caregivers took to manage and control the patient's circumstances and behavior. This category was intentionally inclusive, and it allowed the researcher to incorporate fieldnotes given widely varying codes; thus, under the rubric "management practices," she collected incidents including: incessant monitoring of the patient; warning or "talking to" the patient; and delib-

erately deceiving the patient in order to manage troublesome behavior. The analysis at this stage is still preliminary, and the meaning and significance of any fieldnote is open to further specification and even fundamental reinterpretation. For this reason, the ethnographer should feel free to include any particular fieldnote excerpt in multiple categories.⁴

Sorting requires physical movement of the data in ways that alter the narrative sequence of the fieldnotes. In the past, fieldworkers often cut up a copy of their fieldnotes and sorted the pieces into piles which would then be repeatedly rearranged as analysis proceeded. A variety of computer programs can now perform the sort function very quickly and efficiently, although some fieldworkers still prefer the flexibility that an overview of fieldnotes spread out on a table or the floor affords.⁵ We strongly recommend that those who use this later method keep either a computer copy (with a backup) or a hard copy of the original notes intact for later reference.

Having decided on core themes and perhaps having sorted the fieldnotes accordingly, the ethnographer next turns to a fine-grained, line-by-line analysis of the notes in focused coding. This involves building up and elaborating analytically interesting themes, both by connecting data that initially may not have appeared to go together and by delineating subthemes and subtopics that distinguish differences and variations within the broader topic.

As an example, the fieldworker whose research focused on caregivers looking after family members with Alzheimer's disease became aware of the stigma frequently attached to the latter's condition and behavior. Sorting all fieldnotes on stigma (broadly conceived) into one long document, she then reread and recoded all these materials, in the process developing a series of subthemes of stigma. For example, she came to distinguish "passing"—efforts to prevent the stigma from becoming publicly visible—from "covering"—efforts to cover up, normalize, or distract attention from visible stigmatizing behavior. She also recognized and coded for situations in which the caregiver cooperated with the person with Alzheimer's to manage stigma and for situations in which the caregiver entered into some kind of "collusion" with others to apologize for or manage the stigmatizing incident and its social effects.⁶ In focused coding,

the researcher constantly makes comparisons between incidents, identifying examples that are comparable on one dimension or that differ on some dimension and hence constitute contrasting cases or variations. When the ethnographer identifies such variation, he asks how the instance differs and attempts to identify the conditions under which these variations occur.

By breaking down fieldnotes even more finely into subcodes, the ethnographer discovers new themes and topics and new relationships between them. The same openness to new ways to understand and fit pieces of data together that we encouraged earlier applies to focused coding as well. In some cases, this process generates new issues or opens up new topics that carry the analysis in an entirely different direction and may even require a rethinking and regrouping of the fieldnotes. One student ethnographer reported:

You're both discovering and creating the pattern as you create the pieces—the initial codes—and these begin to structure and frame what the other pieces are going to be and how they will fit together. You have one note and you say to yourself, "Oh, this note seems to fit and be similar to the first note, but it's slightly different and that's what I mean by variation. But somehow they seem to follow one another." Then you continue and read and maybe 15 pages later there's something that seems like it follows or fits. You begin to find pieces that fit together in some kind of way. Don't worry how they all fit in the total paper, just keep fitting them together even if you don't have the connections between them. The aim is to identify what is going on irrespective of whether you will use it later on.

Another student, initially overwhelmed by the number of preliminary codes, said: "I felt that there were so many codes that it wasn't very logical." But she persevered until she could begin to see that there was more to discover in the notes: "I did see that within the more general codes I could see how that once I cut them up I could separate them out into smaller subgroups. What I need to do is do them again." Through the process of focused coding, the ethnographer begins to recognize a pattern in what initially looks like a mass of confusing data. With focused coding, the ethnographer may also begin to envision possible ways of making an argument or telling a story.

Students often express concern when they have only one example of a particular kind of incident or issue. They are concerned that writing about just one instance may distort their analysis if it reflects the response of only a few of those in the setting. Finding only one example would be a problem if the field researcher's purpose were to make claims about

frequency or representativeness. But frequency is only one dimension for analysis. While the researcher delights in numerous examples of a theme or topic, the goal in ethnographic analysis is not representativeness. Rather, the ethnographer seeks to identify patterns and variations in relationships and in the ways that members understand and respond to conditions and contingencies in the social setting. That there is "only one case" often does not matter.⁷ When the ethnographer is fortunate enough to find more than one instance, it is important to note how they are the same and how they vary. Useful questions to keep in mind at this point include: How is this example the same and different? What were the conditions under which differences and variations occurred?

As the ethnographer turns increasingly from data-gathering to the analysis of fieldnotes, writing integrative memos which elaborate ideas and begin to link or tie codes and bits of data together becomes absolutely critical. In writing integrative memos, the ethnographer seeks to explore relationships between coded fieldnotes and to provide a more sustained examination of a theme or issue by linking together a variety of discrete observations. At this point, many ethnographers continue to write primarily for themselves, focusing on putting the flow of their thoughts on paper and maintaining the loose, "note this" and "observe that" style characteristic of several of the memos we have considered to this point. Others, however, find it useful to begin to write with future audiences explicitly in mind. For these researchers, integrative memos provide a first occasion to begin to explicate contextual and background information that a reader unfamiliar with the setting would need to know in order to follow the key ideas and claims. Imagining this future readership within a particular discipline spurs the ethnographer to write in a more public voice, that is, to word ideas in concepts and language that approximate the analytic writing in a final text. It is a first attempt to formulate a cohesive idea in ways that would organize a section of the final ethnography (see the discussion in chapter 7). Thus, such memos sound more polished.

Substantively, integrative memos may move through a series of field-note incidents, linking these incidents by connecting sentences. We ex-

amine the following extended memo on "remedial covering" by family members caring for persons with Alzheimer's disease to illustrate these processes:

Remedial covering involves attempts to correct the troublesome behavior once it has occurred. Caregivers take it upon themselves to watch over the family member and attempt to "smooth over things" in a variety of public places. For example, Laura explains what she does in the presence of friends:

He may take the cup off the saucer and just put it somewhere else on the table. And I'll say, "I think you'd probably get that cup back over here because it'll get tipped over, and it's easier if you have it close to you like that." . . . I try to smooth over these things.

In a similar case, Carol recounts how Ned embarrasses her by removing his dentures in a restaurant and how she handles this:

I got up real quick and stood in front of him and said, "Get your teeth in your mouth." Then she explains to me, "I felt I had to protect him. What if the waitress came?"

In this first segment, the ethnographer links two separate incidents occurring in restaurants through the themes of "watching over" and "smoothing over things." In doing so, differences between the incidents—for example, in the first instance, that something untoward is prevented from happening, while in the second the untoward action has occurred but is literally "covered" and then corrected—are subordinated to these commonalities.

The researcher then takes up a further dimension of remedial covering, specifying the contrast between covering that relies upon the cooperation of the person with Alzheimer's and covering that is carried out directly by the caregiver:

Remedial covering involves having to negotiate the individual's cooperation when he or she is capable of doing so. For example, Laura describes her husband in a local restaurant, how she instructs and physically maneuvers him through various eating tasks ("puppeteering," Polner and McDonald-Wikler 1985) and how he responds. Her description of their interaction gives a real flavor of the minute detail to which the caregiver must attend:

I'll say, "Now turn around some more so that your legs are under the table and then move over so that you're in front of the placemat." . . . Then he would set the beer out very perilously near the edge, and I'd move it back. . . . And then I'd have to arrange things . . . he picked up the tortilla, and it wasn't appropriate. And if anybody were watching, they'd say, "Tsk tsk."

While Laura suggests remedial practices to William in the above example, Tess in her situation takes over and attempts to remedy the situation on her own. She describes going to a buffet restaurant with some of her co-workers, where she tries to cover her father's mistakes so the co-workers are less likely to notice:

Him and I go to buffets all the time . . . and I watch him. I make him go ahead of me so I can fix everything he screws up. He like takes the spoon, puts some cheese on his salad, puts the spoon on his plate. . . . And I grab the spoon and put it back . . . all the employees that I work with are behind me. . . .

Here the ethnographer sets up a contrast between two different responses to the problematic acts of a person with Alzheimer's. First, she notes Laura's handling of her husband by means of orders; in so doing she sees and marks a parallel with the concept of "puppeteering" developed in an article she is familiar with. Second, she examines Tess's ways of managing her father by directly "taking over." She then continues by considering the conditions under which one or the other form of remedial covering is likely:

As the person with Alzheimer's is less and less able to cooperate with the caregiver in these covering practices, the caregiver is forced to take more control of the situation. For example, Carol states, "I'm more ready to be the ultimate authority. . . . This is the way it's going to be done. In other words, take total control."

In composing this memo, then, the writer outlines a progression from milder to more active and restrictive forms of remedial covering likely to occur as the disease progresses. She ends by arguing that this progression fundamentally involves increasing control over the behavior of the person with Alzheimer's disease; she quotes a caregiver who talks openly of her need to now "take total control."

In writing analytic, integrative memos of this sort, the central task is to develop theoretical connections between fieldnote excerpts. In so doing, the ethnographer confronts difficult analytic choices. One major issue is deciding which theme to make the primary focus, which to include as subthemes, and which to exclude entirely. To return to the dilemma of the student who "lost her paper" while focusing and sorting her notes: one strategy was to divide the paper up into different sections, such that the issues of the teacher's strategies for managing the band and of the students' grouping themselves into cliques would be analyzed as a topics unto themselves. A second possibility was to see these strategies as different aspects of the more general theme. Here, the paper would focus on

how the teacher managed to keep an extracurricular program going in the face of overwhelming odds—declining resources and a large and heterogeneous group of students. Specific subtopics would include how he tried to motivate kids to spend extra time on weekends or extra time during the week, and how he managed the tensions and different interests between the various student cliques.

Deciding how to frame an analysis often requires taking a step back from the particulars of the analysis in order to answer the question: What is the larger, more encompassing question(s) I am responding to? One student who studied an alternative school, for example, was able, once she clarified the story she wanted to tell, to incorporate themes from the following incident involving negotiations over the use of a chair at an all-school meeting:

The chair was just sitting there, and I was sitting behind a group of guys who were saving chairs, and this girl took this chair and started to put her feet on it, and the guy says, "Hey, someone's sitting there." She said, "Well, can't I just use it until he comes back?" Then a student teacher comes along, and you can see him eyeing the chair, and he says, "Can I use your foot rest?" She said, "Someone's sitting there." He said, "Well, I'll just use it until he comes back," and then he sits down. But the first guy says, "Excuse me, someone is sitting there." He says, "Well, I'll give it back when he gets back." The student [whose chair it is] comes back and the teacher just got up and left.

The student ethnographer saw in this fieldnote ways the students at the school negotiated with one another and with a student teacher over seating. But while she found the incident and several like it of interest with regard to relations between students and between students and teachers, she struggled with how to link such incidents to a variety of other themes. She decided at this point to step back and attempt to relate the incident more broadly to what she knew and found interesting about the school. She thought, for example, about the pride that both students and teachers at the alternative public school took in the ethic of "democratic decision making" and "shared power." She contrasted this with many more traditional schools where teachers readily exert authority. With the more general issue of this contrast in mind, the student saw that, on some occasions, teachers in the alternative school may not hold or choose to exercise authority but rather negotiate or defer to student claims to space. This led the student to see that she could tie negotiating for space to a range of other incidents that were decided in nonauthoritarian ways. She also began to look for contrasts in this theme and specifically for examples of

matters that were closed to negotiation. By pursuing this line of analysis, the student saw that what initially might have seemed an isolated, mundane incident was related to larger questions of power and authority. More fundamentally, finding a frame for this incident helped her not to take teacher and student claims to "democracy" and "power sharing" at face value or as givens but rather as achievements that were variously honored in the setting.

Again, there is no single, correct way to organize themes and sub-themes. Part of the decision about which course to take depends on the kind of data that has been recorded. In the study of the high school band, very rich and detailed notes on types of students in the school would allow focusing on student cliques. But if such observations are lacking, cliques must move from the center of the picture and become part of the context or background with something else in the foreground. It is usual for ethnographers to try on, modify, discard, and reconsider several possibilities before deciding which tells the best story. As was the case when writing fieldnotes in the first place, organizational decisions will be influenced by factors that range from how inclusive an organizational scheme is to how well it highlights particular theoretical and substantive interests and preferences.

Analysis of ethnographic data begins with concepts that are grounded in and reflect intimate familiarity with the setting or events under study. From close, systematic attention to the fieldnotes as data, the ethnographer seeks to generate as many ideas, issues, topics and themes as possible. This is an inductive process.

Ethnographers differ from other social scientists who proceed deductively with a theory that explains phenomena and attempt to find instances in the data that illustrate or disprove it. While fieldnotes may privilege certain kinds of events—those of significance to members or that illustrate social processes, for example—the ethnographer proceeds in a more open-ended way, seeking to identify issues and ideas by a careful sifting through and piecing together of fieldnotes. The ethnographer remains open to other possibilities and gives serious consideration to processes and issues that become apparent as she reviews the data.

This should not be understood to mean that the fieldworker com-

pletely ignores existing theory and has no theoretical commitments prior to reading through the notes. It does suggest, however, that for the ethnographer, theory does not simply await refinement as analysts test concepts one by one against events in the social world; nor do data stand apart as independent measures of theoretical adequacy. Rather, the ethnographer's assumptions, interests, and theoretical commitments enter into every phase of writing an ethnography and influence decisions that range from selecting which events to write about to those that entail emphasizing one member's perspective on an event over those of others. The process is thus one of reflexive or dialectical interplay between theory and data whereby theory enters in at every point, shaping not only analysis but how social events come to be perceived and written up as data in the first place.

The goal in fieldwork, then, is to generate theory that grows out of or is directly relevant to activities occurring in the setting under study. This contrasts not only with those who practice deduction from received theory but with proponents of grounded theory (Glaser and Strauss 1967). As suggested earlier, grounded theorists focus on the "discovery" and modification of theory through the close examination of qualitative data. But such an approach dichotomizes data and theory as two separate and distinct entities; it avoids seeing theory as inherent in the notion of data in the first place. But as we have emphasized, data are never pure; they are ripe with meanings and always products of prior interpretive and conceptual decisions. Grounded theory slights the processes whereby data are assembled, processes that build concepts into the data from the start in the very process of writing fieldnotes.

In this respect, it is something of a distortion to talk about "discovering theory," as we often are tempted to do. For while reading our fieldnotes carefully and reflectively, we frequently feel that we "discover theory" in the data contained in those notes. But theory only seems to jump out of the data and hit the researcher in the face; this flash of insight occurs only because of the researcher's prior analytic commitments built into the notes, the theoretical concerns and commitments she brings to the reading, and the connections made with other "similar events" observed and written about. Thus, it is more accurate to say that the ethnographer creates rather than discovers theory. She does so not simply in the culminating moment of reading and reflecting on what she has seen and written about previously, but also throughout that prior process of seeing as she writes fieldnotes.

In fieldwork, then, events and actions become meaningful in light of an emerging meaningful whole. The analysis of fieldnotes is not just a matter of finding what the data contain. Rather, the ethnographer selects out some incidents and events, gives them priority, and comes to understand them in relationship to others. Grasping the continuously analytic character of fieldwork often entails a shift in the ways we often think of the ethnographer's relationship both to the fieldnotes and to analysis of them. One student describes the process of finding her "ethnographic voice":

At first, I wanted the paper to emerge through the notes in the sense that it had its own story, and I was supposed to tell its story. But I had to make the shift from just wanting to talk about what was in the notes to making something solid out of them—my ideas, instead of thinking that it's hidden somewhere in the notes.

Rather than simply tracing out what the data tell, the fieldworker renders the data meaningful. Analysis is less a matter of something emerging from the data, of simply finding what is there; it is more fundamentally a process of creating what is there by constantly thinking about the import of previously recorded events and meanings.

In coding and memo-writing, the ethnographer has started to create and elaborate analytic themes. In writing an ethnographic text, the writer organizes some of these themes into a coherent "story" about life and events in the setting studied. Such a narrative requires selecting only some small portion of the total set of fieldnotes and then linking them into a coherent text representing some aspect or slice of the world studied.

Writing a thematic narrative differs fundamentally from writing an analytic argument, both in the process of putting that text together and in the structure of the final text. Structurally, in a text which presents a

logical argument, the author sets forth a formal thesis or proposition in the introduction as a stance to be argued, then develops each analytic point with evidence logically following from and clearly supporting the propositional thesis.⁴ In contrast, an ethnographic story proceeds through an intellectual examination of evidence to eventually reach its contributing central idea. While a thematic narrative begins by stating a main idea or thesis, it progresses toward fuller elaboration of this idea throughout the paper. Indeed, the more precise, fuller statement of the thesis is often most effectively presented at the end of the story, in a conclusion to the paper.

In addition, the structure of an ethnographic story results from an ordered progression of fieldnote excerpts. The details in the fieldnotes stand as the essential kernels of the story. That is, thematic narratives use fieldnotes not as illustrations and examples of points that have already been made, but as building blocks for constructing and telling the story in the first place. In this sense, the main idea grows out of the process of coding and selecting excerpts, rather than prefiguring the choice of fieldnotes to include. The excerpts in an ethnographic story are not so much evidence for analytic points as they are the core of the story.

In terms of writing processes, developing a thematic narrative requires constant movement back and forth between specific fieldnote incidents and progressively more focused and precise analysis. To facilitate this process, we do not recommend beginning with a tentative thesis or working hypothesis. Instead, we urge the writer to hold off formulating an explicit thesis until the paper is finished, so that even in the process of writing, she will make discoveries about data and continue to balance her analytic insights with the demands of sticking close to indigenous views. We suggest that the ethnographer begin developing a thematic narrative by writing out a statement of a general topic or question. A topic ties a broad analytic concern or sensitivity to the events that occurred in the setting. For example, "ethnicity as social construction in a high school" and "parental involvement in juvenile court hearings" provide such topic statements.⁵ At this early stage, topic statements point to a concern or phenomenon, but they do not pose a specific problem or question, nor propose a formal thesis or explanation. Rather, a topic or question identifies a more general focus and helps the author to begin tying fieldnotes together into a coherent whole.

In general, the topic of the ethnographic story will incorporate several more specific analytic themes, i.e., claims about key patterns, processes,

or regularities within the setting. Hence one way to develop a topic is to review the earlier codings and memos, identifying a number of the more interesting or relevant themes in one's fieldnotes. At this point we advise that one write out phrases stating possible themes clearly and explicitly. Initially, the researcher need not be concerned with deciding how these themes relate to one another or how they might be tied together; the writing is intended simply to clarify and specify themes of possible interest. But once several promising themes have been identified, the ethnographer looks for ways of relating some of these themes to one topic and then decides to drop those themes which cannot be tied to this topic.

Alternatively, the ethnographer may come away from his coding and memo-writing with a clear sense of an interesting and unifying general topic. He should then write out this topic as explicitly as possible and then attempt to specify more particular themes that might develop that topic by reviewing his codings, memos, and perhaps even original fieldnotes. For example, having written the phrase, "I will show that parents become involved in court decisions," the student ethnographer studying juvenile court then asked in what different ways parents might become involved in these hearings. On reviewing his codings, he found two distinct patterns, one in which the judge used parents as a source of information about youth, the other in which the judge sought to help parents control their children. He then wrote out these two more specific themes: "The judge sometimes uses parental information against the minor in order to sentence him" and "the judge also might support the parents in disciplining the minor and therefore threaten punishment."

In either case, the ethnographer will move back and forth between topic and themes, writing an overview statement that relates themes to a topic and to one another, and/or developing explicit phrasings for each identified topic. The relations between themes need not be tight and closely reasoned; in "thematic narratives" the themes can be loosely integrated. Relating and ordering themes will usually require changes in wording and conceptualization. Clearly, some themes may not "fit" with others, even on these terms, and may have to be dropped. In fact, even after developing an overall plan for a first draft, it is quite common to revise both the specific thematic statements and their interconnections a number of times as work progresses and the ethnographic story begins to take shape.

Consider how one student began to develop a thematic narrative around the general topic "ethnicity as social construction" in a public

high school. First, he wrote out an elaboration of his topic: "Through people's interactions 'ethnicity' is constantly being recreated and modified within a situation." Then, he wrote out a number of specific themes or issues that he wanted to deal with. Finally, to present these themes, he worked out the following order for five specific sections of the text—each centered on one theme:

In developing these themes, the ethnographer does more than name different situations; more fundamentally, he points out distinctions and interconnections between related phenomena. For example, the theme of how students talk about and identify "different social and ethnic groups" not only considers a range of ethnic (and social) groups but also deals with the ethnic identities assigned to others; in contrast, the theme addressing how "black social groups maintain ethnic boundaries" will involve examining how group members establish their own ethnic identity. Yet he also suggests important linkages between these phenomena; for example, exploring "whites' uses of black styles" suggests a concern with the blurring and crossing of ethnic boundaries that will elaborate and extend his interest in the maintenance of black ethnic boundaries.

To pick a topic and specific themes, the ethnographer must make choices. Fieldworkers regularly find that they have many more themes than they are able to include in any particular manuscript. The process of developing a story is essentially one of selecting some themes that resonate with personal or disciplinary concerns and that recur in a number of specific fieldnotes. In selecting these themes and the data they make relevant, the ethnographer inevitably ignores other themes and data, at least for this particular manuscript.

In developing a topic and then assembling themes into a story, the ethnographer should make every effort to incorporate multiple voices and perspectives. To do so often requires giving special attention to selecting and framing the topic and subsequent, interrelated themes, for how a topic or theme is named and developed can implicitly privilege some

voices and perspectives and exclude others. For example, one student studying the relations between domestic workers and their employers initially identified "hiring" as one topic in her ethnography. But "hiring" frames events from the point of view of the employer, highlighting and privileging her concerns with finding a worker who is "reliable" and "trustworthy." "Hiring" implicitly neglects the domestic worker and her practices for "getting hired" or "finding work." A more relational framing—e.g., "the hiring situation"—would incorporate the perspectives of both employer and domestic worker.

In the following sections, we present ways of turning fieldnotes into ethnographic texts. While recognizing that the initial commitment to a general topic and several initial themes informs this process, we emphasize how the ethnographer elaborates, specifies, and excerpts fieldnotes—which may be only loosely associated with a common theme—in order to develop a finished ethnographic story.

Atkinson (1990:103) argues that the "persuasive force" of an ethnographic text derives from the "interplay of concrete exemplification and discursive commentary." We are explicitly concerned with producing such fieldnote-centered texts—stories that stay close to and are highly saturated with bits and pieces of fieldnotes. To create such a text, the ethnographer must conceptualize the relevance of local happenings so that they relate to analytic issues; but simultaneously, the ethnographer must remain sensitive to how these reframings might distort the meaning of member categories.

To begin this process, the fieldworker must return to the fieldnotes that inspired the story to look for potential excerpts that could develop a story line. The ethnographer first identifies pieces of fieldnote data and then writes interpretive commentary about these excerpts; she also edits each excerpt-and-commentary unit so that the analysis elaborates and highlights the fieldnotes which are the kernels of the story. Finally, the researcher must organize these excerpt-commentary segments into coherent sections of the ethnography; that is, she orders them in a sequence which creates a compelling story line which leads readers to an ever fuller understanding of the people and issues addressed.

Depending upon the time available, the ethnographer might rework units and sections a number of times, replacing initially selected excerpts with new ones, refining analytic commentary and transitions, reordering units within a section and/or rearranging sections within the overall ethnography. Although she sees still further possible changes and refinements, at some point she must stop revising and take up a series of final writing tasks required to turn the now substantial body of text into a completed ethnographic document. These projects include titling the ethnography, writing an introduction linking the topic and major theme to other research, describing the setting and methods, and providing an overall conclusion to the ethnography.

Finished ethnographic texts usually end with a section which reflects on and elaborates the thesis addressed in the introduction to the paper. Hence, while naturally among the last pieces of writing the author does, conclusions are intricately tied to introductions.¹⁸ Often the conclusion explores the implications of the theoretical and/or substantive issues raised in the paper's introduction. In an ethnographic paper, the wording of the introductory thesis focuses the reader on the central idea, but often this idea may not be as sharply delineated as the concluding presentation of the thesis. Whereas the introduction prepares the reader to understand the upcoming analytic points and excerpts, the conclusion more precisely interconnects the ideas because, by the paper's end, the reader has read

the whole ethnographic story and absorbed the details of its fieldnote excerpts. In other words, the ethnography tells a story which can be understood fully only by reading the progression of analytic ideas and fieldnote excerpts. Each section with its theme, points, and discussion of excerpts moved the reader further along toward the conclusion, with its more finely tuned thesis.

To write a conclusion, the ethnographer should review the now completed tale, paying particular attention to the framing of that story in the introduction. In most cases it is useful to write a summary of the major findings and themes of the paper. This summary should generally restate the thesis of the paper and then in short, concise sentences suggest how each section advances or contributes to this thesis. In some cases the ethnographer may choose to use the summary to begin the conclusion. In others, she may move directly to other issues without a summary. Yet even when one does not plan to make the summary part of the conclusion, writing a summary is beneficial to the ethnographer; it forces the author to turn from the minute problems of writing up specific ideas and segments to a review of the overall structure and flow of her paper. The result is a gestalt view of the ethnography's initial promises compared to where it has actually gone: a view which gets the writer thinking about some of its wider implications.

Whether the author summarizes or not, conclusions take up the paper's thesis. The ethnographer may do so in at least three ways: (1) by extending or modifying the thesis in light of the materials examined; (2) by relating the thesis to some more general theory or current issue in the relevant literature; and/or (3) by offering a meta-commentary on the thesis or on the methods or assumptions associated with it. An author might employ only one of these options, or might weave together two or even all three options in one longer, more elaborate conclusion.

As an example of the first option, consider some of the concluding portions of the ethnographic study on how family caregivers of persons with Alzheimer's disease manage the stigma associated with this condition. The introduction to this study had highlighted Goffman's (1971) concept of the "family information rule," i.e., the preference for family members to keep knowledge of the stigma (discrediting information) within the family to prevent outsiders from learning about the problem. In the conclusion, the author returns to this issue, suggesting that, as the disease worsens, there is a radical change in the family's ability to honor this information rule:

The Alzheimer's caregiver will try for as long as possible to collude with the family member, continuing to abide by the family information rule to the extent that she or he is able, and limiting initial disclosures to intimates and medical personnel. However, there may well come a point where the caregiver realizes that she or he cannot count on the person with Alzheimer's to be cognizant of what is discrediting, let alone motivated or able to collude in trying to cover it up or minimize its embarrassment. Thus, the information control within the family tends to give way to more direct caregiver interpersonal and interactional control.

The caregiver increasingly relies on a variety of management practices to control the individual, both within the private family domain as well as outside it. And, as the person with Alzheimer's can no longer play the collusive game, caregivers gradually come to align with outsiders, disclosing discrediting information.

Here the author argues that while the family caregiver initially seeks to honor the family information rule, to do so requires cooperation from the person with Alzheimer's disease. When such cooperation can no longer be counted on, the caregiver increasingly violates the rule by disclosing discrediting information to outsiders in order to enlist their help in managing the patient. In this way the author highlights in the conclusion how her findings have modified Goffman's notion to point out previously unnoted conditions underlying the operation of the family information rule and to appreciate the kinds of circumstances which may lead family members to violate it.

Another way authors might extend a thesis statement is to develop theoretical linkages between separate components of the thesis. For example, in the introduction to the study of residents in a home for experimental patients, the introductory thesis pointed to two conflicting tendencies within the home: resident dependence on staff members and the residents' ability to actively influence staff views of them. In the conclusion, the author uses her more specific analyses of these relations to explicitly connect these contradictory tendencies as parts of an ongoing vicious circle. Residents feel vulnerable to the power of staff and may respond by trying to build credit and good-will with them. In order to do this, they participate in therapy sessions and other staff-initiated activities. As a result they gain the staff's support and protection but thereby become more directly dependent upon the staff members who "sponsor" them. This analysis thus links two patterns which initially appear separate and indeed contradictory, pointing to an ironic outcome whereby residents' actions intended to lessen vulnerability and dependence on staff end up tightening that dependence. In this option, the writer tells an

ethnographic story which progresses from an initial thesis which highlighted conflicting tendencies, through an in-depth discussion of analytic points with appropriate excerpts, to finally come to a conclusion which intertwines these conflicting strands.

Second, a conclusion may attempt to connect the ethnography's thesis to issues raised in a relevant disciplinary literature. In the study of Alzheimer's family caregiving, after the paragraphs quoted above, the author relates the contrast between colluding with the person with Alzheimer's and colluding with outsiders to a more general issue in the sociology of deviance: When do family members accept, tolerate, and continue to look out for another family member with some kind of stigmatizing condition or behavior, and when do they turn against, exclude, and implicitly reject this family member? This issue had been recently addressed in a journal article entitled "Toward a Sociology of Acceptance: The Other Side of the Study of Deviance" (Bogdan and Taylor 1987) which the author cites in developing her argument:

Recognition of these two phases of caregiver stigma management, collusion with the person with Alzheimer's, and a realignment and collusion with outsiders, allows for an integration of a sociology of acceptance with a sociology of rejection (Bogdan and Taylor 1987).

Here the author suggests the possibility of unifying sociological theories about why and how people tolerate deviants with theories about why and how people exclude and reject deviants. These reactions need not be opposed, alternative courses of action; some forms of exclusion develop exactly because of a deep and abiding commitment to caring for another, under conditions where the afflicted family member can no longer be "counted on" to aid the caregiver in tolerantly managing the situation. This unity of acceptance and rejection is frequently highlighted, the author argues, in caregivers' deeply ambivalent feelings about having to take overtly rejecting actions toward the person with Alzheimer's disease:

Many caregivers were disturbed about having to take more and more control over their family members. In monitoring the person as well as using physical coercion, they made such comments as, "I hate my nagging voice." Or as one caregiver said with regard to taking control over his wife: "I have no right."

In this way the concept of acceptance is also extended to encompass rejecting actions that are performed reluctantly and are combined with deep regret.

A third option in writing a conclusion is to pause, step back, and reflect on the ethnography in offering some meta-commentary on its methods, assumptions, tone, or conclusions. In the study of resident life in the home for ex-mental patients, the author not only addressed staff-resident relations but also considered how residents related to and developed important social and supportive ties with one another. One section of the ethnography explored the ways in which residents regularly exchanged certain items with one another—cigarettes, food, small amounts of money. The author suggested that these exchanges and the continuing relations they created and sustained helped residents deal with the chronic deprivation that they faced. In one portion of her conclusion, however, she reflected on how this earlier consideration of resident exchange “strategies” presented an “overly rationalized,” game-playing view of these exchanges. This view, she argued, needed to be complemented by appreciation of the caring and emotional qualities also characteristic of these exchanges as well as the role of these exchanges in fostering a sense of sharing and community among a number of residents. The prior strategies-and-tactics analysis tended to obscure and distort these critical processes.

In all these approaches to writing a conclusion, the ethnographer takes up once again the problem of identifying and writing out in explicit, elaborated form the relevance of some of her experiences, observations, and insights into others’ way of life for an outside audience. But by proceeding in a way that keeps fieldnotes at the center of the analytic process, the ethnographer is often able to reach understandings and make connections that do not neatly fit existing explanations and theories in the discipline. The refined, more precise thesis to be presented in the conclusion will more likely privilege members’ views and show what is interesting (and has theoretical import) about this local life in ways which convince one’s scholarly readers. Thus, the more explicit thesis in the conclusion not only represents what the ethnographer saw and heard of members’ experience, but also will further clarify known issues, or propose an entirely original perspective.

In producing an ethnography for wider audiences, fieldworkers are constantly pulled by conflicts between representing some indigenous world

and its meanings and making their own experiences with that world speak to the very different concerns of scholarly readers. In creating a finished ethnographic story, the ethnographer self-consciously orients toward the latter; in regularly returning to his fieldnote record and to the memories bound up with and evoked by this record, he is again and again reminded of the former.

While the give-and-take of relations in the field continue to shape the ethnographer’s understanding, the finished ethnography is the ethnographer’s version of those happenings and events. Most ethnographic conventions allow the writer to represent others (and her experience with them) as she sees best. In this sense the ethnographer openly assumes and exercises authorial privilege. Even in those instances when ethnographers ask members to read portions or to comment on certain analyses, the author has the final say about both the text itself and the extent of members’ evaluations of it (see Bloor 1988; Emerson and Pollner 1988; Rochford 1992). Despite the efforts of intensive participation, the attempts to learn members’ meanings, and the self-reflection in representing others’ realities, the final document turns into a rather linear narrative, defined and controlled primarily by the author.¹⁹

Over time—beginning in the fieldwork and extending beyond writing a paper—an ethnographer develops a complex view of field experiences. Nevertheless, the ethnographic text offers one overriding perspective—his own at the time of writing. Only when the reader’s interpretation differs from that of the fieldworker do the many ways to interpret a set of notes become explicit. But to keep these various possibilities in mind while writing may paralyze the writer, preventing any story from being told. Hence, an ethnography remains one author’s vision of field experiences. And thus, because the author controls the text, she takes on an authoritative voice in writing.

Nonetheless, the ethnographer sometimes provides unintended glimpses into others’ everyday lives. Readers may discern things which the ethnographer did not intend to reveal. In fact, reader participation in text-making can be a double-edged advantage in ethnographies built around fieldnote excerpts. On the one hand, readers more directly engage in the described social scenes and thereby convincingly follow the story line. On the other hand, they can also more readily assess the proposed analysis, at least the version presented by the author, and derive different insights from the fieldnotes.

Readers ultimately make their own sense of these fragments, even

though the note-taking ethnographer created, selected, and arranged them in the text. By choosing fieldnotes for their rhetorical effects as well as for their signifying and conceptual functions, an ethnographer tries to prefigure a reader's likely range of interpretations. Still, the original fieldnotes stand there embedded within the analysis, allowing any reader to listen closely to members' voices, to vicariously experience their actions, and to imagine other interpretations. In the end, it seems, the reader has her say.

In the preceding chapters we initially examined the processes whereby field researchers transform direct experience and observation into fieldnotes. We then considered ways of using fieldnotes to develop and tell an ethnographic story, exploring a variety of procedures that can facilitate the construction of fieldnote-centered texts. In this final chapter we want to offer some further reflections on learning to write and use fieldnotes and on some broader implications of these writing processes for ethnographic research.

As we have seen, in writing fieldnotes the ethnographer makes a number of specific writing choices; through these choices she transforms experience and observations into text and data. Obviously many of these choices involve decisions about what to write—to note and describe the practical efforts of Alzheimer family caregivers, patterns of racial and ethnic distribution in a school playground, or audience participation in storytelling in a Zaïre village. But these choices also involve intricate decisions about how to write about what has been observed and experienced. As we have emphasized, writing fieldnotes is not simply a matter of putting observed details on paper. Rather, the ethnographer draws on a variety of writing conventions in order to actively create characters and scenes on a page, to dramatically depict action and speech, and to effectively convey the meanings of events as perceived by those involved in them.

Ethnographers, of course, may not always make these choices consciously; because the immediate task is to get descriptions and accounts

on the page, experienced writers may use skills and make choices without a conscious thought. But increased awareness of the options which make such choices possible will, we believe, improve the overall quality of ethnographic research. In the first place, heightened consciousness about writing should help ethnographers produce richer, more varied, and useful fieldnotes. In becoming aware of and adept at using effective writing conventions, the ethnographer is more likely to capture significant detail, create vivid imagery, and provide nuanced depictions of talk and events. The field researcher will have greater flexibility in making writing choices. She will know and employ to her advantage the different effects of writing in the third as opposed to the first person, of describing a scene or event from particular or from varying points of view, and of writing up others' talk as direct or indirect speech.

But in addition, increased awareness of writing choices can also inspire the ethnographer to be more attentive to details while in the field. Envisioning scenes as written can make the researcher a better observer. With knowledge of writing options, he will be attuned to features of action and talk that might be captured on paper. Furthermore, a researcher who makes choices about different points of view in his writing is less likely, when observing, to confuse his own perspective with the views of others; hence, he will be able to recognize and represent those members' voices more fully.

Furthermore, sensitivity to writing options in constructing a final fieldnote-centered text also allows field researchers to produce more compelling and detailed ethnographic stories. The writer, for example, becomes explicitly attuned to responding both to voices from the field and to the voices of envisioned scholarly readers. She realizes that she must translate and interpret members' voices into the analytic language of intended readers in order to address issues, theories, and concerns that might interest them. Thus, in creating a fieldnote-centered final text, the ethnographer includes excerpts which report members' voices but with an awareness that she controls and orchestrates their presence; she re-frames and reorders members' words and doings into her ethnographic story. Sensitive to members' concerns and meanings, she can directly confront the task of re-presenting those meanings—for example, making them "interesting" or "relevant" to the concerns of anticipated readers. As a result, in a good ethnography the reader can hear these two sets of voices speaking in harmony or at least not creating dissonance. The ethnography should provide a vehicle through which the voices of the

field can, in their own distinctive ways, speak; and at the same time, the ethnographer should also speak the language of the readers, addressing their issues, theories, and concerns.¹

In all these ways, increased awareness of writing choices allows for a deepening appreciation of the power and implications of writing. The ethnographer cannot help but realize that he is not simply recording witnessed events; rather, through his writing he is actively creating realities and meanings. In writing fieldnotes, he is not simply preserving those moments in textual form, for he is shaping observed moments as scenes, characters, dialogue, and recounted actions in the first place. Subsequently, in reworking fieldnotes and transposing them into a final ethnographic story, he does not simply recount the tale of something that happened; instead, he reconstructs "what happened" so as to illustrate a pattern or a make a point. Inevitably, in interpreting his fieldnotes for readers unfamiliar with that world, he constructs a version of events. Thus, while writing and analyzing fieldnotes, the ethnographer-as-author grows increasingly aware of his role and responsibility in telling the story of the people he studied; for in writing he re-presents their everyday world.² In so doing, he is continually reminded how the act of writing constructs meaning and knowledge.

In this sense, awareness of writing choices generates an appreciation of the reflexivity of ethnographic research. Reflexivity involves the recognition that an account of reality does not simply mirror reality but rather creates or constitutes as real in the first place whatever it describes. Thus "the notion of reflexivity recognizes that texts do not simply and transparently report an independent order of reality. Rather, the texts themselves are implicated in the work of reality-construction" (Atkinson 1990:7).

The growing number of critical analyses of ethnographies that focus on reflexivity (e.g., Clifford and Marcus 1986; Atkinson 1990; Van Maanen 1988) tend to address the rhetorical structure or unstated political and cultural presuppositions of completed ethnographies, examining how the ethnographer represents another culture, develops a particular line of analysis, or constructs a persuasive argument or engaging tale in a published account. However, these analyses reveal significant limitations in themselves, for they implicitly depict final ethnographies as original, unconstrained constructions produced wholly from the ethnographer's struggles to come to terms with experiences in the field. While polished ethnographies are in part culled from memories of and reflections on field

experiences, they also draw heavily on the already created fieldnote record of that experienced reality. Final ethnographies are rarely new edifices built up entirely by original writings but are more commonly jerry-built projects incorporating and constrained by prior fieldnote writings. The representational processes through which fieldnote segments are selected (or ignored), linked to one another, reworked into a consistent voice, and integrated to produce a clearly recognized rhetorical style have received little or no attention. In this way most reflexive analyses neglect or marginalize fieldnotes in the construction of finished ethnographic accounts and, thereby, ignore the role of fieldnotes in the in-process work of actively constructing a polished ethnographic account.

Such critical analyses have overlooked not only the use of fieldnotes in writing final ethnographic accounts but also the prior processes whereby ethnographers actually created a fieldnote record in the first place. Yet this initial transformation of field experiences and observations into written texts involves equally profound and consequential reconstructions of social reality as does the production of polished, full ethnographies. Growing consciousness of the reflexive qualities of ethnographic texts, then, has advanced by and large without attention to day-by-day writing practices for producing what comes to be treated as ethnographic data.

To fill this gap between reflexive analysis and practice, one must look closely at exactly how ethnographers go about writing fieldnotes: how they produce, process, and finally assemble fieldnotes into texts intended for wider audiences. For, in significant ways, describing people, events, and scenes in fieldnotes gives definite shape and substance to these matters for the writer. The writer, after all, does not simply sit down and put directly on paper something already worked out in his head. Rather, he constructs his descriptions: he must decide where to start, what to put first and what later, what to include and what to ignore. While writing, he determines whose points of view to present, what is significant about a person or event, and what is incidental and can be left out. These decisions are even more salient for subsequent readers, who have no independent access to the reality often presumed to lie behind and to have shaped the written account. From a reader's perspective, then, the text about a people's way of life creates that world as a phenomenon.

In the preceding chapters we have emphasized several specific writing choices that can highlight awareness of the reflexive character of ethnographic research. First, we have advocated writing fieldnotes so that the ethnographer can be seen and heard in them, since the ethnographer's

interactions in the field shape her writing. In this way the processes whereby ethnographic texts come to be produced can be preserved and made available to readers: "Being reflexive is structuring communicative products so that the audience assumes the producer, process, and product are a coherent whole" (Myerhoff and Ruby 1982:6). But in addition, in so writing the ethnographer reminds herself that what she learned and writes about occurred on a specific occasion and was shaped by her own methods and mode of participation.

Second, we have urged writing fieldnotes in ways that effectively capture and represent members' meanings—the perspectives, understandings, concerns, and voices of those studied. In order to do this effectively, the writer must clearly understand that she is in fact re-presenting member's meanings, creating, to paraphrase Geertz (1973), "meanings of meanings," or "interpretations of interpretations."

But a seeming problem arises when we recognize that members' meanings are not things in themselves, but representations of something: Why should members' meanings have priority over any other representation an ethnographer might make? Here again, we hold that fieldnotes and finished ethnographies are inevitably and unavoidably mediated by the ethnographer's person, experiences, point of view, and theoretical priorities. But the researcher's point of view and theoretical priorities are not simply pre-given; they are shaped and influenced by the relationships he forms with the people whose social worlds he is trying to understand. As a participant who has a place in the local setting and who has some degree of involvement with the people in it, the researcher is part of the world being studied and not a neutral, detached observer. The process of forming relationships with specific people subjects the ethnographer to their meaning systems, ones that must be learned and understood, if only in order to get by. The more the ethnographer involves himself in others' social worlds, the more he subjects his own presuppositions, his own ways of doing and giving meaning to events and behavior, to the challenges of members' everyday life. The ethnographer's fieldnotes, then, consist of descriptions of and reflections on the meanings acquired and jointly constructed over the course of participation in relationships with those studied. Hence, fieldnotes reflect understandings gained through subjecting oneself to the logic of others' social worlds, a logic that comes to partially constitute the lens through which the ethnographer views and understands those worlds. In the end, what he inevitably writes is his version (informed by theoretical and other concerns and priorities) of their ver-

sion. But the versions an ethnographer constructs are negotiated and mediated by members' points of view, logics, and constructions of the world as well as by the researcher's. Hence, through relationships with others, the possibility exists for appreciation and understanding of the interactions the researcher observes in their, not simply his own, terms.

Reflexivity is central both to how we understand the worlds of others as well as to how we understand the research enterprise. Reflexivity, when applied to the understanding of members' worlds, helps us to see those worlds as shaped not by variables or structures that stand above or apart from people but rather as meaning systems negotiated and constructed in and through relationships. Hence, when self-consciously applied to ourselves as researchers, the reflexive lens helps us see and appreciate how our own renderings of others' worlds are not and can never be descriptions from outside those worlds. Rather, they are informed by and constructed in and through relationships with those under study. Hence, in training the reflexive lens on ourselves, we understand our own enterprise in much the same terms that we understand those we study.