WRITING AN ETHNOGRAPHY
1. To understand the nature of ethnographic writing as part of the translation process
2. To identify different levels
of ethnographic writing
3. To identify the steps in writing an ethnography.
4. To write an ethnography.
Every ethnographer probably begins the task of writing a cultural description with the feeling it is too early to start. Doing ethnography always leads to a profound awareness that a particular cultural meaning system is almost inexhaustibly rich. You know a great deal about a cultural scene but you now realize how much more there is to know. It is well to recognize that what you write is true of every ethnographic description: it is partial, incomplete, and will always stand in need of revision. Most ethnographers would do well to set aside the feelings that writing is premature and begin the task sooner rather than later. In the process of writing one discovers a hidden store of knowledge gained during the research process.
As most professional writers will affirm, the only way to learn to write is to write. In the same way that learning to swim cannot occur during classroom lectures on swimming, discussion of principles and strategies to follow in writing do not take one very far in learning to write. It is best to observe other swimmers, get in the water yourself and paddle around, and then have an experienced swimmer point out ways to improve your breathing and stroke.
One of the best ways to learn to write an ethnography is to read other ethnographies. Select those that communicate to you the meaning of another culture. Those written in a way that brings that culture to life, making you feel you understand the people and their way of life. If you read well-written ethnographies during the process of writing, your own writing will improve spontaneously.
Every ethnographer can identify books and articles that are well-written cultural descriptions. In the past eight years, my colleague David McCurdy and I have scoured the professional literature in search of brief examples of ethnographic writing of the highest caliber. Our standard has been to identify writing that translates the meanings of an alien culture so well that someone unfamiliar with ethnography grasps those meanings. These selections of ethnographic writing have been collected in three successive editions of Conformity and Conflict: Readings in Cultural Anthropology (1971, 1974. 1977). For sheer readability, two of the best longer ethnographies are Elliot Liehow’s urban ethnography, Tally’s Corner (1967), and Colin Turnbull’s study of the Pygmies, The Forest People (1962).
In this step I want to examine briefly the nature of ethnographic writing as part of the translation process. Then I want to discuss the principles of the D.R.S. Method as applied to writing an ethnography. In the process I will give some specific suggestions on writing, but always keep in mind that the way to learn to write an ethnography is to write an ethnography.
THE TRANSLATION PROCESS
A translation discovers the meanings in one culture and communicates them in such a way that people with another cultural tradition can understand them. The ethnographer as translator has a dual task. For one, you must make sense out of the cultural patterns you observe, decoding the messages in cultural behavior, artifacts, and knowledge. The more fully you apprehend and digest the cultural meaning system operating in the social situation you study, the more effective your final translation.
Your second task is to communicate the cultural meanings you have discovered to readers who are unfamiliar with that culture or cultural scene. This means that every ethnographer must develop the skills of communicating in written form. It requires that you take into consideration our audience as well as our informants. In a real sense, a truly effective translation requires an intimate knowledge of two cultures: the one described and the one tacitly held by the audience who will read the description.
Many highly skilled ethnographers fail to finish the work of ethnographic translation. They give months of time to the intensive study of another culture, analyzing in great detail the meanings encoded in that culture. Then, without taking time to learn the skills of written communication, without understanding their audience, without even feeling the importance of communicating in a way that brings the culture to life, they write an ethnography. Their audience becomes a very small group of other ethnographers, who, by virtue of their interest in the culture, are willing to wade through the vague and general discussions, examine the taxonomies, paradigms, and other tables or charts, and glean an understanding of the people and their way of life. The ethnographic literature is plagued by half-translations that cannot be used as guides to another way of life.
In discussing the steps in writing an ethnographic description I will make numerous suggestions for creating a full translation, one that communicates the cultural meanings you have discovered. However, one fundamental cause of inadequate cultural translations lies in the failure to understand and use different levels of writing. During the writing of any ethnographic description, the ethnographer must keep these various levels in mind and consciously use them to increase the communicative power of the translation.
Levels of Ethnographic Writing
Every ethnographer deals with the most specific, concrete, human events as well as the most general. In our fieldnotes we identify an infant with a specific name, held by a specific mother, nursing at that mother’s breast, at a specific time and in a specific place. In those same fieldnotes we will make observations about human love, nurturance, and the universal relationship of mothers and children. In the final written ethnography, the range of levels is enormous. More than anything else, the way these levels are used will determine the communicative value of an ethnographic translation.
Kenneth Read, in his beautifully written ethnography of the Gahuku peoples of Highland New Guinea, The High Valley (1965), suggests the underlying cause of partial translations in ethnography:
Why, then, is so much anthropological writing so antiseptic, so devoid of anything that brings a people to life? There they are, pinned like butterflies in a glass case, with the difference, however, that we often cannot tell what color these specimens are, and we are never shown them in flight, never see them soar or die except in generalities. The reason for this lies in the aims of anthropology, whose concern with the particular is incidental to an understanding of the general (1965:ix).
In anthropology, as in all social sciences, the concern with the particular is incidental to an understanding of the general. But when this principle is transported wholesale into doing ethnography, it creates a travesty of the translation process. When an ethnographer studies another culture, the only place to begin is with the particular, the concrete, specific events of everyday life. Then, through the research process described in this book, the ethnographer moves to more and more general statements about the culture. With the discovery of more general categories and cultural themes, the ethnographer begins to make comparisons with other cultures and make even more general statements about the culture studied. And all too frequently, it is primarily this kind of analysis and understanding that finds its way into the ethnographic description.
In writing an ethnography, as a translation in the full sense, the concern with the general is incidental to an understanding of the particular. In order for a reader to see the lives of the people we study, we must show them through particulars, not merely talk about them in generalities.
There are at least six different levels that can be identified in ethnographic writing as we move from the general to the particular. Let’s examine each of these different kinds of translation statements.
Level One: Universal Statements
These include all statements about human beings, their behavior, culture. or environmental situation. They are all-encompassing statements. Although the beginning ethnographer may often feel incompetent to make any universal statements, all of us know things that occur universally and can include them in our ethnographies. Most cultural descriptions include universal statements. A study of air traffic controllers, for example. might assert, ‘In all societies, people manage the movement of their bodies through space in such a way that they do not constantly collide with other human beings.” Such a statement is relevant to controlling the movement of vehicles in which humans move about as well. A study of clerks who record burglaries in the police department might assert the following universal statement: “in all human societies, some people keep records of one sort or another about their affairs.”
For each of the six levels of abstraction that appear in ethnographic writing, I want to give an example from The Cocktail Waitress: Woman’s Work in a Man’s World (Spradley and Mann 1975). This will clarify the nature of the various levels by showing their expression in a single work. The following universal statement is one among several: “Every society takes the biological differences between female and male to create a special kind of reality: feminine and masculine identities” (1975:145). In the context of a specific bar in a specific city, we made an assertion about a universal feature of human experience.
Level Two: Cross-Cultural Descriptive Statements
The second level of abstraction includes statements about two or more societies, assertions that are true for some societies but not necessarily for all societies. Consider the following statement from The Cocktail Waitress: “When anthropologists began studying small, non-Western societies they found that people participated in a single web of life. . . . [W]hen we turn to complex societies such as our own, the number of cultural perspectives for any situation increases radically” (1975: 8, 9). This statement says something about two very large classes of human societies—the small, non-Western ones and the complex ones. Such a descriptive statement helps to convey an understanding of even the most specific place such as Brady’s Bar. Cross-cultural descriptive statements help place a cultural scene in the broader picture of human cultures, something every ethnographer is concerned about doing. These kinds of statements say to the reader, “This cultural scene is not merely one little interesting group of people; it is a part of the human species in a particular way. It is like many other cultural scenes, but it is also different from many others.” By means of contrast you have conveyed an important dimension of the culture.
Level Three: General Statements about a Society or Cultural Group
This kind of statement appears to be specific, but in fact remains quite general. ‘The Kwakiutl live in villages along coastal bays” is a general statement about a cultural group. ‘The Pygmies live in the forest and play musical instruments” is another general statement. We can make such statements about complex societies also: “American culture is based on the value of materialism.” Or we can make such statements about recurrent cultural scenes, or groups of people who have learned similar cultural scenes: “Air traffic controllers work under great stress”; “Police departments must gather, classify, and record a great deal of important information.”
In our study of Brady’s Bar, we included statements at this level; they did not refer exclusively to Brady’s Bar, but to all the institutions of which Brady’s was one example: “Bars, in general, are places of employment for hundreds of thousands of women, almost always as cocktail waitresses. Their role in bars tends to be an extension of their role at home—serving the needs of men. . . . Like most institutions of American society, men hold sway at the center of social importance” (1975: 145).
Level Four: General Statements about a Specific Cultural Scene
When we move down one level of abstraction, we can note many statements about a particular culture or cultural scene, Most ethnographies are filled with statements at this level. “The Fort Rupert Kwakiutl engage in seine fishing.” “The air traffic controllers at the Minneapolis International Airport work one of three shifts.”
Participant observation provides many statements at this level. We can say things like, “The waitresses at Brady’s get hassled by customers” or “Tramps aren’t really tramps unless they make the bucket,” These are descriptive statements about a particular scene or group, but even so, they are still general in nature. Moreover, even when expressed by an informant and used in an ethnography as a quotation from an informant, they represent an abstraction. Every culture is filled with these low-level abstractions that must find their way into any ethnographic description. Here is an example from Brady’s Bar: “At one level, Brady’s Bar is primarily a place of business. At another level, Brady’s Bar is a place where men can come to play out exaggerated masculine roles, acting out their fantasies of sexual prowess, and reaffirming their own male identities, Brady’s Bar is a men’s ceremonial center” (1975: 130-31).
This level of ethnographic writing contains many of the themes the ethnographer wants to present to the reader. Thus, the theme of males expressing their identities in many different ways—in the way space is organized, the way drinks are ordered, and the like—is described in statements at this level. Sometimes one can encapsulate general statements at Level Four in a quotation from an informant: they still remain statements of a very general nature. Making use of an informant quotation helps provide a sense of immediacy and gives the reader a closer acquaintance with the culture, but we must move to even more specific levels.
Level Five: Specific Statements about a Cultural Domain
At this level, the ethnographer begins to make use of all the different terms in one or more cultural domains. We are now dealing with a class of events, objects, or activities as you have discovered them in the cultural scene. For example, here is an ethnographic statement at this level from my research on the factory that makes tannery equipment: “One of the most important jobs that men do is to make drums. A drum can be small, such as a barrel, or more than thirty feet across. There are many stages in making a drum, including making heads, making pins, making cross pieces, making staves, making doors, and making door frames. The entire process of making a drum can take as long as a week and involve the work of several men.”
Descriptive statements at this level can make reference to taxonomies and paradigms that encapsulate a great deal of information. However, these representations in themselves seldom communicate more than a skeleton of relationships to the reader. In order to translate these into a description that will be understood, a great deal of narrative description at this level and the next more specific level is required.
Here is a brief example of a specific statement about the domain, “asking for a drink,” which makes up part of the culture of cocktail waitresses. ‘One frequent way that men ask for a drink is not to ask for a drink at all. In the situation where it is appropriate to ask for a drink, they ask instead for the waitress. This may be done in the form of teasing, hustling, hassling, or some other speech act” (1975:132).
Level Six: Specific Incident Statements
In one sense, Levels One through Five all contrast sharply with Level Six, which takes the reader immediately to the actual level of behavior and objects, to the level of perceiving these things. Consider an example from Brady’s Bar at this level, closely related to all the examples given at the other five levels of abstraction: “Sandy is working the upper section on Friday night. She walks up to the corner table where there is a group of five she has never seen before: four guys and a girl who are loud and boisterous. She steps up to the table and asks, “Are you ready to order now’?” One of the males grabs her by the waist and jerks her towards him. ‘I already know what I want! I’ll take you,” he says as he smiles innocently up at her” (1975:132). As a reader, you immediately begin to see things happening, perhaps feel things the actors in this situation feel. Instead of merely being told what people know, how they generate behavior from this knowledge, and how they interpret things, you have been shown this cultural knowledge in action. A good ethnographic translation shows; a poor one only tells.
Perhaps another example of the six levels in ethnographic writing will clarify the effect on the reader. Drawing from my research among tramps, the following statements all describe a single aspect of their experience: begging, borrowing, panhandling, lending, and otherwise exchanging things.
LEVEL ONE: Reciprocity among human beings is balanced where two people give to each other over time, each giving and each receiving. Such reciprocity occurs in all societies.
LEVEL TWO: Tramps, like those who live in tribal villages, depend on one another in time of need. They expect others to reciprocate. A Kwakiutl Indian will give in a potlatch and later receive gifts at someone else’s potlatch. A tramp will give to another tramp and also beg from another tramp.
LEVEL THREE: Tramps engage in much more reciprocal exchange than do other members of the larger society. This kind of exchange takes many forms.
LEVEL FOUR: A tramp in the Seattle City Jail will exchange goods and services with other tramps. If he is a trusty in the jail, he might exchange a service for money with someone in lockup.
LEVEL FIVE: (Informant’s statement) “Yes, a tramp will beg from other tramps. If you’re panhandling you can expect another tramp to give you money or a cigarette if he has it. You realize that sometime he will need something and then it will be your turn.”
LEVEL SIX: It was a dull Tuesday afternoon and a slight mist of rain was blowing gently in from the Puget Sound. Joe had become a kickout an hour earlier; several minutes ago he walked off the elevator on the first floor of the Public Safety Building and found his way to the street. Pulling the collar of his worn tweed jacket up around his neck, he hunched his shoulders slightly and headed downtown, wondering where he would find money for a drink or even a cigarette. He might have to make a flop under the bridge on Washington Street tonight to stay out of the rain. He saw a man approaching him as he headed slowly down James Street, obviously another tramp. Looked like a home guard tramp, but he couldn’t tell for sure. “Can you spare a quarter for a jug?” he asked. “I just got a kickout.” “No, I’m flat on my ass myself,” the other man said, “but how about a smoke, all I got are Bull Durhams.” After taking a light too, Joe started on down James Street looking for a tourist or businessman to panhandle.
Ethnographic writing includes statements at all six levels on the continuum from the general to the particular. Effective writing, that which serves to communicate the meanings of a culture to the reader, is achieved by making all these statements, but doing so in a certain proportion. Professional journals, in which the author writes primarily for colleagues, tend to consist of statements at Levels One and Two; that is, the description is made in general terms, the author avoiding specific incidents. Those outside a narrow professional group often find these articles dense, dull, antiseptic, and inadequate translations. Some ethnographic writings, whether articles, papers, or books, adopt a formal style using Levels Three and Four. Most dissertations and theses are written at these middle levels of abstraction, although they may contain a great deal of information also at Level Five. They tend to present the bare bones, the skeleton of knowledge, without the flesh of examples and specific incident statements of Level Six. At the other extreme, some ethnographic novels and personal accounts consist entirely of statements at Level Six, with, perhaps, a few statements from Level Five thrown in now and then. This kind of writing holds the reader’s attention hut may fail to communicate the overall structure of a culture or the nature of ethnography.
It should be clear that mixing the various levels in the desirable proportion depends on the goals of the ethnographer. In You Owe Yourself a Drunk: An Ethnography of Urban Nomads (1970), 1 made a great deal of use of Levels Three through Six, ranging back and forth among statements about tramps generally, to specific incidents. Many of the incidents were contained in quotations from informants. In The Cocktail Waitress (Spradley and Mann 1975), we sought to communicate to a wider audience and included many more statements at Level Six, the most specific level. We also tried to relate the culture of Brady’s Bar to the universal level of writing more frequently. In retrospect, we tended to scale down the middle level of generalizations. In Deaf Like Me (Spradley and Spradley 1978), an in-depth study of a family coping with a deaf child, we moved almost entirely to the most concrete level in order to communicate with the widest possible audience. Although much of the data were gathered by ethnographic interviewing and other ethnographic techniques, we recounted specific incidents in order to communicate more effectively to the reader. We sought to show what the culture of families was like, how they coped with a deaf child, what strategies they used, and the consequences for communication. Although statements appear in this study at all the other levels of generalization, they are woven into the particular so thoroughly that they do not stand out. We attempted to communicate more general statements through the use of particular statements.
Each ethnographer will have to determine his or her intended audience. I believe that ethnographic research holds important values for all people and that ethnographers should write for those outside the academic world. I urge students and others to use the middle levels of generalizations sparingly. Emphasize the most general and the most specific. In ethnographic writing, the concern with the general is incidental to an understanding of the particular for an important reason. It is because generalities are best communicated through particulars. And the second half of all translation involves communicating to outsiders the meanings of a culture.
STEPS IN WRITING AN ETHNOGRAPHY
Like doing ethnographic research, writing an ethnography can appear to be a formidable task if seen as a single task. All too often, the beginning ethnographer conceives the writing as simply writing. You sit down with blank paper and all your fieldnotes and begin writing the ethnography. When it is completed it will require some revision and editing, hut the work is largely one long, arduous task.
Underlying the D.R.S. Method of research is the assumption that breaking a large task into smaller ones and placing these in sequence will simplify the work and improve one’s performance. This assumption applies equally to writing. However, because each of us has developed patterns of writing from years of experience, it is far more difficult to create a series of steps that have wide applicability. The following steps must be considered as suggestions only. Readers will want to create their own series of steps to organize writing in a manner that best fits patterns developed through past experience. However, the underlying premise, that it is valuable to divide up the writing of an ethnography into tasks, does have wide applicability.
Step One: Select an audience. Because the audience will influence every aspect of your ethnography, selecting an audience is one of the first things to be done. All writing is an act of communication between human beings and in that sense it is similar to talking. When speaking to someone, there are innumerable cues that remind us that our audience is present. The writer needs to select an audience, identify it clearly, and then keep in mind throughout the writing who that audience is.
When writing for a specific journal or magazine, the ethnographer must carefully scrutinize past issues of the journal to discover the style of writing. You are, in fact, discovering the audience that such a journal is written for. If one intends to write a book-length ethnography, then the audience may be scholars in the field, students, the general public, or some other group.
The best advice I have ever received for selecting an audience came from Marshall Townsend, the editor at the University of Arizona Press:
A basic concept we stress at the University of Arizona Press is that of the “target reader.” What we urge YOU as an author to do is to pick out a “target reader” and write in book form for only one reader. Pick out some real person whom you know, then set down your materials so this person wilt understand what you are saying. When you have a “target reader,” you effect a single level of presentation. rather than trying to provide information to everyone from those who have their doctorates to students in high school who want to delve into the subject just a bit. Choose your level of communication and stay with it—by addressing yourself in your writing to only this one person. We believe you will find this concept a highly workable one.
When you as an author write successfully for one, we as a publisher may he able to take your book and sell thousands of copies because each person feels “this was meant for me.” On the other hand, if you try to write for thousands, and embrace all of their varied interests and viewpoints, we may not he able to sell a single copy. Stick to your one-level approach, and we as publishers will take care of informing readers at all levels of interest and of understanding how the hook will fit Into their realm.
Step Two: Select a thesis. In order to communicate with your audience, you need to have something to say. All too often, ethnographic descriptions appear to be like meandering conversations without a destination. Although of interest to the ethnographer and a few colleagues, such writing will not hold the attention of many more. A thesis is the central message, the point you want to make. There are several sources for finding a thesis.
First, the major themes you have discovered in ethnographic research represent possible theses. For example, a major theme in the culture of tramps was that being in jail affected one’s identity, even made a man want to go out and get drunk. In jail a man learned to “hustle,” and this reinforced his identity as a tramp trying to “make it” on the street. This theme became the thesis of the ethnography: that jailing drunks, rather than being therapeutic, actually played an important role in creating the identity of tramp. This thesis was summed up in the title of the ethnography, which came from an informant who said, “After thirty days in jail, you owe yourself a drunk!”
Second, a thesis for your ethnography may come from the overall goals of ethnography. You may, for example, state your thesis in the following way: “To most people, a bar is a place to drink. But to the cocktail waitress, it is much more complex. It is a world of varied cultural meaning that she learns in order to carry out her work and cope with difficulties. In this paper I want to show just how complex the cultural knowledge of the cocktail waitress is, in contrast to the casual impressions of the outsider.” Your thesis can simply be to show that cultural meaning systems are much more complex than we usually think.
Another way to formulate this type of thesis is in terms of a set of recipes for behavior. Culture can be viewed as a set of instructions for carrying out life’s ordinary activities. Your thesis would be to show the reader the recipe for being a tramp, a cocktail waitress, or some other kind of person. Frake in a series of articles, has made effective use of this kind of thesis. For example, he has written on “How to ask for a drink” (1964c) and “How to enter a house” among the Yakan in the Philippines (1975).
Still another way to formulate this type of thesis is to show the tacit rules for behavior. This thesis argues that much goes on in social life that we do not see, that there are tacit rules of behavior that people have learned but seldom discuss. The point of your paper is to make those tacit rules explicit.
Third, a thesis may come from the literature of the social sciences. In one paper on tramps I reviewed the literature on the concept of “reciprocity.” Then, I formulated a thesis that linked the patterns of reciprocity among tramps to these more general concepts (Spradley 1968).
When a thesis has been selected, it is useful to state it briefly, perhaps in a single sentence, and place it before you as a constant reminder as you write. This will help organize your paper and integrate it around a single major idea. It will also help the reader to grasp the meanings of the culture ma way that a simple listing of domains and their meanings will not.
Step Three: Make a list of’ topics and create an outline. Any ethnography will necessarily deal with only selected aspects of a culture. Furthermore, you will use only part of the material you have collected. This step involves reviewing your fieldnotes and the cultural inventory you made and listing topics you think should be included in the final description. Some of these topics will be things like “introduction” and “conclusion.” Once listed, you can then make an outline based around your thesis. This will divide up your actual writing into sections, each of which can be done as a separate unit. If you have been writing short descriptive pieces throughout the project (See Appendix B), many or all of these may fit into the outline.
Step Four: Write a rough draft of each section.. A rough draft is intended to be rough, unfinished, unpolished. One of the great roadblocks for many writers is the desire to revise each sentence as it goes down on paper. Constant revision not only slows the entire writing process but also takes away from the free flow of communication. Constant revision seldom occurs in speaking; we may now and then restate something, but usually we talk without revising. Write as you talk is an excellent rule to follow in composing a rough draft of each section.
Step Five: Revise the outline and create sub/heads. Almost always the outline from which one writes becomes changed in the process of writing. Once a rough draft is completed for each section, it is a good idea to make a new outline, rearranging sections as appropriate. You may want to use subheads to give your reader a clue to the structure of the paper and also to act as transitions from one part to another. Native folk terms can often be used as subheads in an ethnography, helping to create a view that reflects the cultural knowledge of your informants.
Step Six: Edit the rough draft. At this point in the writing you will have a rough draft of your paper, a fairly clear outline, and a number of subheadings you want to use throughout the paper. Now it is time to go over it with an eye to improving the details of writing. Work through each section and at the same time keep the entire description in mind. Make changes directly on the pages you had previously written. When you want to add a paragraph or sentence, write them on the back of the page or on a separate piece of paper with instructions as to where they will appear. At this stage, it is often useful to ask a friend to read over the manuscript and make general comments. An outside perspective is especially useful for making improvements that will enhance the communicative power of the description.
Step Seven: Write the introduction and conclusion. By now the description has taken on substantial form and you can write these two parts of the paper in a more effective manner. Some writers find that they write better if they write a rough introduction at the start of the writing but save the conclusion until the end. In either case, now is the time to review both the introduction and conclusion and revise them to fit the paper.
Step Eight: Reread your manuscript for examples. Examples involve writing at the lowest level of abstraction. Because of their importance in communication, a special reading of the paper to see if you have used enough examples is highly desirable. Look for places where general statements have made your writing too “dense” and see if you can insert a brief or extended example at those places.
Step Nine: Write the final draft. In some cases this will merely involve typing the paper or turning it over to someone else to type. In other cases. you will need to go carefully over the manuscript again, making the final editorial changes. Using steps such as this means you have been over the entire manuscript numerous times during the course of writing. Instead of a single, first-draft-as-final-draft, your paper has gone through a series of developmental stages.
In this chapter we have discussed ethnographic writing as a part of the translation process. Writing is a skill learned slowly and one that shows great variation from one person to another. The suggestions in this chapter are offered only as general guidelines, not as hard-and-fast rules that every writer should try to follow.