Ways of Looking at Ethnographic Film

In 1970, at the American Anthropological Association meetings in San Diego, Robert Gardner screened 'The Nuer' to a large anthropological audience for the first time. Probably everyone in the audience knew the Nuer from Evans-Pritchard's ethnographies. When it was over and the lights came up, we sat stunned. People said, 'Evans-Pritchard had never told us that the Nuer were like this!' A picture is worth a thousand words. What can that mean? Can one second of a movie (at 24 frames per second) be worth 24,000 words? No. But one could easily use 1,000 words to describe and analyze what is happening in a picture. The notion of 'seeing anthropology' is important, for both still photographs and moving images can complement the words anthropologists use to describe how culture works. Just as the words need to be studied, so the images must be thought about carefully.

Look at the picture to the right, which is taken from a book published in 1879. You can see at a glance that it refers to Europeans' schemas of their role among humanity. Now spend some time with the picture. Look at it more carefully. Describe and analyze it. What about the faces, clothing, body language, genders, and actions? What do all these details mean? Now try this exercise again: What is the Europeans' image of themselves, of those Others, and of their interrelationships? You can easily use more than 1,000 words to describe this picture. It is in this way that 'visual anthropology' implies the critical visual analysis of human behavior. The visuals mentioned in this article - both the still photographs and the video clips - are offered here for critical visual analysis.

This article is a very brief introduction to ethnographic film, intended to highlight some concepts and some ideas to keep in mind while studying the films. By now, a great deal of literature has focused on ethnographic film. My own book Ethnographic Film (1976) is a convenient starting place, but there are many more recent publications. For general overviews, there are excellent books by Peter Loizos, such as Innovation in Ethnographic Filmmaking (1993), and by Ilisa Barbash and Lucien Taylor, including Cross Cultural Filmmaking (1997). There are also books by or about some of the most important ethnographic filmmakers: Jean Rouch (Stoller 1992), John Marshall (Ruby 1993); Robert Gardner (National Rhythms by T. Cooper 1995 and Kapfer, Petermann, and Thomas 1990); David and Judith MacDougall (MacDougall 1998); and Timothy Asch (Lutkehaus 1995). Here we will touch on a few of the most important issues in ethnographic film. First, though, an abridged history.

As soon as camera and film technology became available, filmmakers looked for subjects. Different exotic peoples, with their costumes and dances, were considered good subject matter. People who were brought from afar to be exhibited in European expositions were recorded. Explorers and anthropologists went abroad to film Australian Aborigines, Native Americans, and others in their own lands. The first major milestone was in 1922, when Robert Flaherty's Nanook of the North was released to tremendous public acclaim. Edward Curtis' film In the Land of the Head-Hunters, later retitled In the Land of the War Canoes, was made eight years earlier among the Kwa Kwaka' Wakw (Kwakiutl), but it was not commercially successful. It fell into complete obscurity until a copy was found in the 1940s. Flaherty is known to have seen this film, however, and it likely influenced the making of Nanook.

Anthropologists were slower to see the uses of film in their research. Franz Boas had shot footage of the Kwa Kwaka' Wakw before World War I, but it was not until Gregory Bateson and Margaret Mead did fieldwork in Bali between 1936 and 1939 that major research really incorporated film. Those films were not edited and released until the early 1950s. The beginning of the modern era of ethnographic filmmaking came in the mid-1940s, with the films of Jean Rouch in France, and in the mid-1950s in the United States, with the films of John Marshall. Marshall began shooting extensive footage of the Jul'hoansi of the Kalahari Desert in 1951 at the age of 17. The release of The Hunters in 1957 opened a productive era for ethnographic filmmaking in the United States. John Marshall returned to the Kalahari throughout his life to make films and advocate for Jul'hoan self-determination. His ethnographic film achievements culminated in a six-hour, five-part series, A Kalahari Family, chronicling the lives of a band of Jul'hoansi over the span of fifty years, completed in 2002. Jean Rouch made many films in West Africa that are now classics. Rouch first visited West Africa in 1941 and returned frequently throughout his life. The films that we will see range from early footage by John Marshall to decidedly more modern films, often made from a more emic, or native, point of view (for example, the Vietnamese film, How to Behave (Chuyen Tute).

Words are linear; pictures are nonlinear. That is, words are arranged in sentences that are read in the order intended by the author. Those words refer to one thing at a time. 'A well-dressed European man, holding a Bible in his right hand . . . ' So could you begin the description of the picture we studied above. When you look at the picture, however, your eye has great freedom to roam around, to notice or ignore details, to make connections, and to interpret meanings. Your reading of the picture is hardly a neat, determined linear act. This idea is even more applicable to films. To complicate things even further, in films there is sometimes a wordy narration that tries to draw attention to only one aspect of the visuals. One skill of critical visual analysis of these films is the ability to see what is happening in the background, or at the sides of the frame, rather than only what the camera has focused on.

Although photographs and film have an apparent veracity, there are at least five ways in which this truthfulness of photographs can be affected:

1. The presence of the camera, or the film crew, can make people self-conscious, altering their usual behavior, perhaps evoking mugging for the camera.
2. The filmmakers can stage scenes, or pose people, to get more effective or more aesthetic shots, in the process altering reality.
3. The very act of photography is extremely selective, pointing the lens in one direction and thus not filming all other possible directions, as well as turning the camera on for only a few moments and so not capturing the rest of the action.
4. Editing, in its acts of including or cutting shots, is selective.
5. Editing, in juxtaposing shots from different times and places (the process called montage), creates an implied unity of action that may not have existed.

And let's not even get into the digital editing possibilities of adding or removing people or things from shots.

Imagine that an anthropologist does fieldwork somewhere. She takes notes and shoots film or video footage. She comes home and prepares a film and a book. For the book, she works over her notes, writes, rewrites, and produces a smoothly flowing manuscript. As she prepares the film, she edits her footage. She throws out some shots, shortens and rearranges other shots. Although she cannot create new footage, the anthropologist can manipulate the original footage - cutting, pasting, and discarding, as well as using narration to cover gaps of shooting and to provide understandings that she has developed late in the game.

In the film Dead Birds, for example, there are battle sequences edited from footage shot at several different battles. As the narration does not mention this point, only the most attentive observers would pick it up. (I have discussed it in print - see Heider, 1972, 1976, 1997.) The narration says that women are making the trip to the brine pool on the day of the battle. It is probably obvious that Robert Gardner, with his camera, was not simultaneously filming both the battle and the salt trek. Both of these constructions of reality seemed reasonable in the early 1960s, when the film was made, for they gave an accurate picture of Dani battles and division of labor. Today, a desire for a different, more literal accuracy might have resulted in different editing.

Another example of creative editing occurs in John Marshall's classic 1957 film about the Jul'hoansi, The Hunters (some shots from it are included in Marshall's film N!ai). Four men are followed on a long and eventually successful giraffe hunt. The film presents a basically accurate picture of Jul'hoansi hunting knowledge and techniques. But the hunt itself has been edited together from footage of different men, shot in different years. These conventions of editing have been criticized in recent years. In reality, it would be virtually impossible to make a film that did not, to some extent, use them. Nevertheless, ethnographic film, in so far as it is ethnographic, is held to more scientific standards of truthfulness than other types of film. One solution is to have a study guide of some sort accompany the film to explain the editing decisions. (As models, see Heider, 1972; Rundstrom, Rundstrom, and Bergum, 1973; Connor, Asch, and Asch, 1986.)

Words can make abstractions and generalizations, but film cannot. Think of trying to visualize 'the Nuer have patrilineal clans,' or Vietnamese 'kindness,' or Mayan cargo rankings. Each of these concepts is easy to explain in words but virtually impossible to show in visuals alone. That is, in part, why narration seems so necessary for ethnographic films.

In earlier ethnographic films, the voice, or point of view, was that of the filmmaker speaking through the narration. This conceit was not so much a choice as the result of the available technology. There was no way to record sound simultaneously with the visuals ('synch sound'). The machinery was bulky, and the camera made noise that would be picked up by the tape recorder. At best, one could record appropriate sound ('wild sound') and try to match it up with the visuals during the editing process. So, for example, when he was shooting footage for Dead Birds, Robert Gardner would ask Michael Rockefeller (our sound man) to record men's feet running on dry ground, or women splashing and chatting at the brine pool. In editing, this wild sound was laid in carefully to give the illusion of synch sound reality. By the early 1970s, it was possible to have real synch sound with 16-mm cameras, but it was still expensive. Most of the footage for The Nuer was done with older cameras and wild sound. When Robert Gardner visited Hilary Harris on location (Gardner had been filming the Hamar, another group), however, he brought his own camera and filmed several interviews in synch sound.

Later in the 1970s, synch sound equipment became widely available and ethnographic filming began to incorporate it routinely. This technological advance had a major effect on ethnographic filmmaking. Hearing the voices of the people themselves was not just an audio flourish. It changed the very nature of authority on film. Instead of being instructed by the all-knowing narration of the (usually male) outsider, we could listen to the people explain themselves. Even when the conversation was idle chitchat, a sense of greater reality was present. Most of the films featured in this article convey key ideas and interpretations through such peoples' voices. Of course, outsiders made each of these films - even the Vietnamese film crew of How to Behave carefully defines itself as different. Nevertheless, synch sound does much to mitigate the 'outsiderness' of film and to give some touch of the native point of view. Even in Dead Birds and Cows of Dolo Ken Paye, the omniscient narrators at times quote or paraphrase people's comments. (The only exception, alas, is my own film, Dani Sweet Potatoes.)

A second important change that synch sound made is to give us not just 'the' voice of the people, but multiple, often discordant, voices. The narration style had often simplified matters with statements like 'the Dani say. . . ' or 'the Kpelle believe...' In films that let us hear more than one 'voice of the people,' we learn that Vietnamese have varied understandings of the concept 'kindness,' as suggested in the film How to Behave, that not all Indian wives agree about arranged marriages and dominant mothers-in-law (Dadi's Family), and that Malay villagers have many explanations for latah (Latah: A Culture Specific Elaboration of the Startle Reflex). This plurality of voices is not trivial, but speaks to the issue of culture as shared.

None of the films referenced in this article are purely visual. Each has a soundtrack, and each soundtrack carries not only general noise but also words, sentences, and utterances. One - Dani Sweet Potatoes - has only words on the soundtrack. At best, the words complement and inform the visuals; at worst, they present information quite unrelated to the visuals and so distract the viewer's attention. For the viewer, it is already a challenge to search shots for complex visual information. If one is being fed data on population density, or some such topic that has little to do with the action on screen, one is distracted from both visual and verbal data. Ethnographic films do not always use narration. In the 1960s, Asen Balikci and Robert Young made a series of 30-minute films of Netsilik Eskimo life that had only natural sound and no English explanation or translation at all. These films were part of a grade-school curriculum project, and their goal was to show Netsilik technology in such detail that American schoolchildren could follow and understand the making of an igloo, the creation of a sled, or seal hunting.

One way of thinking about behavior is in terms of multiple sets of acts that have beginnings, come to peaks of intensity, and then end. A major challenge of ethnographic filming is to capture these whole acts. It takes considerable knowledge of the culture to begin shooting even before an act begins, so as to get its prehistory. Then it takes great patience (and lots of film) to keep shooting as an act winds down and the actors move on to other acts. Anyone can see to shoot the peak moments, but it is the rare photographer who can capture whole acts. Whether the act lasts moments, minutes, hours, days, or years changes the magnitude of the problem but not its basic shape. Even when shooting whole acts, the filmmaker often zooms in for close-up shots. Certainly, there are some moments when close-up detail is needed. I have taken the position that for the most part the lens should be open wider. (This position is not universally applauded, as evidenced by Paul Hockings' 1997 review of my book in the journal Cultural Anthropology). There are several reasons for using a wide angle.

First, when filming a person, there is a temptation to zoom in on the face. This choice cuts out the rest of the body, and many cultures employ hands and arms to carry important nonverbal information in communication (as illustrated, for example, by the film Box of Treasures). Many or most ethnographic films have been shot by filmmakers from northern Europe or North America, where nonverbal patterns typically emphasize facial communication and the rest of the body is relatively ignored. In filming people of other cultures, these northern Europeans are not aware of the importance of hands and arms and simply do not think to include them in the frame. The principle of whole bodies refers to the need, in much communication, to be able to see more than just the face.

A second reason to avoid close-ups in filming communication is that when two people are talking, both are communicating constantly. Perhaps it was Gregory Bateson who said, 'Nothing never happens,' meaning that no one is ever just sitting there, not communicating. The work of conversation analysts like Charles Goodwin, in his 1981 book Conversational Interaction, shows that although two people take turns speaking (with the verbal channel), the person not speaking is still sending messages (through the nonverbal channels) that contribute to the forming of the conversation. Goodwin speaks of 'the emergent form of the utterances,' meaning that the speaker is constantly reacting to those nonverbal signals being sent by the 'listener' and that the two work steadily to shape the utterances. A convention in American fiction films calls for representing a conversation by showing first one person full-frame speaking, then cutting to a full frame of the other answering. This practice does give fine detail of one speaker's face alternating with the other speaker's face, but it completely denies the interactional character of the conversation. The solution is to open the lens wider to include both (or all) speakers at once, thereby capturing the (nonverbal) messages sent by the 'listener.'

A third reason for using a wide lens is to show the behavioral context, or setting, of an act. Behavior takes place in settings. There are appropriate and inappropriate settings, and settings influence behavior. Thus, especially when looking at unfamiliar cultural behavior, it is important to be able to see it in context. This argument does not imply that all close-ups are always bad. Sometimes the importance of detail warrants being deprived of the view of the whole body or of the greater context. Ethnographic films rarely show too much context, however, for the zoom lens has a siren song that few photographers can resist - it says, 'Zoom me! Zoom me!'

We have seen that ethnographies are becoming more reflexive, incorporating information about the ethnographer and the effect of the ethnographer on the situation described. The films used in this book show little of such reflexivity. For example, Dead Birds gives no hint that the entire Harvard Peabody Expedition was living in the neighborhood and that we were often present, just outside the frame, at many events. Cows of Dolo Ken Paye shows a brief glimpse of the anthropologist, James L. Gibbs, Jr., serving for the moment as soundman. Other, more recent films actually show the anthropologist interviewing people. None of these films really deals with the effect that anthropologists may have on the subjects' behavior, however.

A problem for many filmmakers is how to make people sympathetic and understandable when they are doing things bound to disgust the film's audiences. In Dead Birds, the killing of a pig at the boy's funeral often turns Americans off and convince them that the Dani are savage. Never mind that most of those Americans are not vegetarians - they are just not able to watch animals die. As a Nuer man cleans his teeth in the morning in The Nuer, the narrator dances away from the fact that he is doing it with cow dung ashes. Surely, these films must at the very least create empathy and understanding across cultures. Must one therefore omit any scene that might have the opposite effect? I made two films about the Dani - Dani Houses and Dani Sweet Potatoes - to explain the simplicity and sophistication of Dani gardening and architecture. When I showed them to my introductory anthropology class at Andalas University in West Sumatra, Indonesia, I was shocked at the students' reaction. When they saw the films, they roared with scornful laughter, convinced that those people in Irian Jaya were indeed primitive orangutans ('people of the forest'). It was a low moment in my teaching career.

'Taking photographs,' 'shooting film,' and 'captured on film' are not totally inadvertent metaphors for a process that raises daunting ethical issues. There is inevitably an element of theft and of aggression in filming. One way to avoid the problem is to get 'informed consent' from everyone in the film. But even camera-sophisticated Americans cannot be really made aware of all of the possible consequences of appearing in a film. Even if they view the final version of the film, they will not know in advance how audiences will react to them. And what of the people shown in our various film clips? It would have been impossible to obtain truly 'informed consent' from each of them, and signed release forms are not the answer.

Speaking as both an anthropologist and a filmmaker, I suggest that there are two possible solutions to this problem. One is to stop making films. The other is to take the greatest care to make films that represent people with understanding and respect, and during the filming to inform and involve them. I choose the second way, even though no magic formula is available for achieving these goals. Nevertheless, these standards allow ethnographic films to be made ethically, doing more good than harm.

Finally, here are some questions to ask yourself when you watch ethnographic films:

1. Whose voice(s) or point(s) of view does the film present?
2. How much information is carried by the visuals and how much by the sound (talk or narration)?
3. Does the narration complement of distract from the visuals?
4. How visual is the film? Does it show things that would be difficult or impossible to describe in words?
5. What is the balance between close-ups and long shots? What does each contribute?
6. Are whole bodies shown? To what effect? (Is there too much or too little zooming?)
7. Are whole acts shown? To what effect? (Are shots or sequences of shots too short or too long?)
8. Are the anthropologists acknowledged? To what extent?
9. Does the film create empathy? Are there scenes that create antipathy?
10. How does the film establish generalizations or state abstract concepts?
11. Does the film contradict itself? Is this a strength or a weakness?

[This article is extracted and edited from Seeing Anthropology: Cultural Anthropology Through Film by Karl G. Heider (Pearson Education, 2007, pp. 434-440).]