In Sound Theory/Sound Practice, Rick Altman, ed. New York: Routledge, Chapman, and Hall, 1992, 217-234.

Conventions of Sound in Documentary

This essay draws comparisons between various examples of sound practices and narration in the documentary tradition, focusing primarily on synchronous sound observational films from the 1960s and 1970s, in particular the 1973 PBS series An American Family. While documentary sound tracks typically include voice-over, dialogue, music, and effects, the hierarchy and distribution of these sounds differ in important ways from classical Hollywood conventions. In fact, Hollywood's standard division of sound into discrete tracks obscures the extent to which these are integral parts of documentary sound. In a series of articles, Rick Altman has described the conventions of sound in classical Hollywood cinema as an interplay between intelligibility and fidelity, a system in which fidelity is sacrificed in favor of the more narratively central dimension of intelligibility. (Altman 1984, 1985a, 1985b, 1986, 1990)

Similarly, Noel Carroll has argued that the hallmark of Hollywood movie narration is clarity and comprehensibility. Popular movies offer experiences of places, events, characters, and drama more clearly delineated than our ordinary lives. In Carroll's words, "The flow of action approaches an ideal of uncluttered clarity. This clarity contrasts vividly with the quality of fragments of actions and events we typically observe in everyday life." (Carroll 1988: 180) Hollywood filmmakers use cinematic techniques of image and sound to focus the attention of the spectator on the salient elements that further the narrative action. Carroll suggests that it is not the purported realism of the cinematic apparatus that millions of viewers find compelling, but rather the heightened intelligibility that is the hallmark of Hollywood cinema. If audiences were truly interested in greater fidelity to the real world, then presumably documentary films

Documentary films rarely demonstrate the degree of clarity that these writers see as the standard of classical Hollywood cinema. Location sound work in documentary films occasionally makes discrimination among sounds difficult, if not impossible. The intelligibility of documentary rarely approaches that of popular movies; characters lack clear motivations, speech may be inaudible in parts, lighting haphazard and variable, camera movements follow actions with difficulty, sound spaces differ radically between scenes, microphones accidentally appear in the image, jump cuts disrupt continuity, and questions remain unanswered.

History of Observational Cinema

In the late 1950s, innovators in television journalism worked to apply different principles of story-telling to the documentary format, in an attempt to move away from illustrated lectures. Following the tradition of the photojournalism of Life magazine, producers like Robert Drew wanted to give the impression of lived experience by being there on location as events happened. During a Nieman Fellowship at Harvard University in 1954, Drew studied the narrative structure of the 19th century realist short story, a form he wanted to apply to documentary. (O'Connell 1988: 88) Returning to New York with funding from Time-Life to create a new kind of actuality film, Drew assembled a team of talented young filmmakers to form Drew Associates--Richard Leacock, D. A. Pennebaker, Albert and David Maysles, Hope Ryden, James Lipscomb, and others. New portable 16mm equipment, developed during World War II for use by the military and expanded by the needs of television in the post-war era, made observational synchronous sound filmmaking a possibility for the first time in the late 1950s.

Observational filmmakers were not to intrude on the lives of their subjects, not to ask questions, conduct interviews or otherwise direct, stage, or influence the events for the camera; they were to be as flies on the wall. These filmmakers wanted to eliminate overt narrational devices like voice-over in favor of stories that begin in medias res and unfold seemingly without a narrator. Drew Associates opted for stories that had inherent drama and were structured around crisis events with clear beginnings, middles, and ends. (Mamber 1974: 115-38) Similar experiments at the National Film Board of Canada by Tom Daly, Colin Low, Michel Brault, Roman Kroiter, Terrence Macartney-Filgate, and Wolf Koenig

were broadcast on the series "Candid-Eye" that ran from 1958-61 on Canadian television. The Canadian filmmakers were directly influenced by the street photography of Henri Cartier-Bresson, which mixed formal composition and spontaneity. (Jones 1988) Early observational films focused heavily on biography for their narrative unity, with titles such as David (1961), Eddie (1961), Lonely Boy (1962), Nehru (1962), Jane (1962), and Susan Starr (1962). As Robert Allen and Douglas Gomery point out in their case study of Drew Associates in Film History: Theory and Practice, these innovations in television journalism were never fully adopted by the networks and probably had more influence on fiction film than on commercial television. (Allen 1985: 239)

Modernist variations on the theme of observational cinema quickly emerged. The Maysles brothers constructed open-ended episodic narratives in films like Showman (1962) and What's Happening! The Beatles in the U.S.A. (1964), while Andy Warhol moved towards minimalist experiments in Sleep (1963) and Empire (1964). Jim McBride and L.M. Kit Carson parodied the search for truth in the mock-autobiography David Holtzman's Diary (1967), while Allan King looked at private life in A Married Couple (1970). Frederick Wiseman, eschewing the earlier concentration on celebrities as subjects, introduced a multiple-focus narrative structure for his series of films on everyday life in American social institutions--Titicut Follies (1967), High School (1968), Law and Order (1969), Hospital (1970)--claiming in each case that the institution itself is the star. During this period, Craig Gilbert made celebrity portraits like Margaret Mead's New Guinea Journal (1968) and The Triumph of Christy Brown (1970), mixing observational footage with re-enactments, archival footage, and voice-over narration.

Documentary filmmakers in the 1970s turned increasingly to more private subject matter in autobiographical forms. Joyce Chopra's Joyce at 34 (1972), Amalie Rothschild's Nana, Mom, and Me (1974), and Jill Godmilow's Antonia: Portrait of a Woman (1974) explored personal issues in the growing women's movement. As Craig Gilbert commenced work on An American Family, filmmaker Ed Pincus embarked on the autobiographical Diaries, 1971-76 (1981), adopting a loose chronological first-person narrative style, based on chance and the everyday, in which the filmmaker appears as the main character. An American Family represents a compromise among these different tendencies.

Produced by veteran National Educational Television director Craig Gilbert, An American Family, a twelve-hour series on the Loud family of Santa Barbara, California, captured the imagination of the American viewing public in the spring of 1973. Under Gilbert's supervision, filmmakers Susan and Alan Raymond filmed the everyday lives of Pat and

William Loud, and their children Lance, Kevin, Grant, Delilah, and Michelle for seven months. The extensive filming gave the crew and family ample time to get to know one another so that the family members could perform their everyday lives in the presence of a camera crew and the filmmakers could become temporary members of the family. Gilbert deliberately chose an upper-middle class white family whose standard of living approximated that of the suburban families shown living the American dream in television situation comedies such as Father Knows Best, Leave It To Beaver, Make Room For Daddy, The Adventures of Ozzie and Harriet, The Dick Van Dyke Show, and The Brady Bunch.

The series focused on the small details of the everyday lives of the seven family members, while relying on the overall crisis structure of the Loud's divorce to organize the story. The concentration on individual biography is mediated by the ability to shift the narrative focus from one family member to another. Gilbert initially considered organizing An American Family around episodes devoted to individual members of the family before eventually choosing a chronological multiple-focus narrative framework for the series. A major innovation of the WNET series was that the film unfolded in twelve separate episodes over the course of twelve consecutive weeks, allowing for on-going viewer involvement in the lives of the characters. Critics responded to the series format and subject matter by referring to An American Family as "real-life soap opera." Although unusual in documentary, serial structure was, of course, standard practice with situation-comedies and soaps well before the 1970s. While Gilbert worked as a documentary producer for telelvision in New York throughout the 1960s, he acknowledged that he was never really part of the vanguard independent community gathered around Drew Associates, although he knew their films well. Significantly, An American Family was made for public television and never received theatrical distribution. Although this style of observational filmmaking was standard in the world of independent film by the 1970s, never had it attained such a wide audience. To this day, An American Family remains the most widely seen and debated example of observational cinema in the United States.

Although observational filmmakers in the 1960s justified their new narrative style through references to Gustave Flaubert and to fidelity to the real world, the classical Hollywood cinema, whose roots may also be found in 19th century realism, provided the clearest example for the new documentary. However, documentary filmmakers who argued that they were brushing up against the truth could hardly cite the classical Hollywood cinema as their model; this was a period when many documentary filmmakers defined Hollywood as their enemy. In 1983, Robert Drew admitted that "You don't need Dan Rather in the middle of a fiction motion picture to tell you what's going on." (Hindman 1983: 47) Similarly, David

MacDougall stated, "Many of us who began applying an observational approach to ethnographic filmmaking found ourselves taking as our model not the documentary film as we had come to know it since Grierson, but the dramatic fiction film, in all its incarnations from Tokyo to Hollywood." (MacDougall 1975: 112) These observational filmmakers abandoned the established Griersonian tradition of direct address in favor of a style which used techniques of storytelling and continuity editing conventionally associated with fiction films, although documentary pioneers like Robert Flaherty had also worked in this narrative tradition.

Location Sound in Documentary

One of the major stylistic characteristics of documentaries that use exclusively sounds recorded on location is the lack of clarity of the sound track. Ambient sounds compete with dialogue in ways commonly deemed unacceptable in conventional Hollywood practice. A low signal-to-noise ratio demands greater attention from the viewer to decipher words spoken in situ. Slight differences in room tone between shots make smooth sound transitions difficult. Indeed, listening to many of the scenes of observational films without watching the screen can be a dizzying experience. Without recognizable sources in the image to anchor the sounds, we hear a virtual cacophony of clanging, snippets of dialogue and music, and various unidentifiable sounds, almost an experiment in concrete music. Freed of their associations to objects, the sounds resurface in their phenomenological materiality. Because scenes in observational films are not usually shot under optimal conditions, such as those found in a Hollywood studio, the sound track lacks the clarity and directness signifying that the sound was created for the listener. While Hollywood sound tracks are typically easier to understand than sounds in everyday life, documentary sound tracks are potentially more difficult to follow than sounds in everyday life.

The history of industry practices indicates that film production in the United States moved inside studios around 1908 to avoid the kinds of uncertainties encountered in actuality and location filmmaking. (Izod 1988: 13) Mass production techniques and a precise division of labor ensure the maximum efficiency and technical quality of Hollywood productions. Repeated takes are done until satisfactory sound has been recorded; if necessary, dialogue will be post-produced through dubbing techniques to ensure clarity. Hollywood directors shoot individual shots one at a time under optimal conditions, while documentary filmmakers often shoot entire scenes in one long take in unpredictable situations. In Hollywood films, the degree of direct sound, as opposed to reflected sound, indicates the level of control exercised over all aspects of production. The brilliance of sound practice in the classical Hollywood cinema derives from a combi-

nation of direct sound, closely miked in order to reduce reverberation and increase clarity, with an overall system of impersonal narration. (Altman 1990: 25)

The clarity of sound in documentary usually depends on the degree of control that the filmmaker has over the profilmic events. Voice-over narration allows for maximum control over sound quality. Voice-over has long been one of the stylistic signatures of documentary sound. Recent documentaries like Jill Godmilow's Far From Poland (1985), Ross McElwee's Sherman's March (1985), Tony Buba's Lightning Over Braddock (1988), Lise Yasui's Family Gathering (1988), and Michael Moore's Roger and Me (1989), have rediscovered the possibilities of voice-over narration, using personal, ironic, and interpretive commentary to counterpoint the synchronous images and sounds. In these documentaries, voice-over narration is more than just a necessary concession to the needs of story-telling. In Hollywood cinema, voice-over is still considered "the last resort of the incompetent," as Sarah Kozloff points out in Invisible Storytellers: Voice-over Narration in American Fiction Film, a view shared by many observational documentary filmmakers. (Kozloff 1988: 21) When Leacock and Pincus taught documentary filmmaking at the Massachusetts Institute of Technology in the 1970s, voice-over was not considered an acceptable technique. (Ruoff 1988) When voice-over narration appears in fiction films, it often serves as a marker of documentary realism, as in the "News on the March" sequence in Orson Welles' Citizen Kane (1941) and in John Ford's How Green Was My Valley (1941), Anthony Mann's The T-Men (1947), and Jules Dassin's The Naked City (1948).

Speech in Documentary

Characters in documentary films typically demonstrate a wider variety of accents, dialects, and speech patterns than those found in fiction films. In their discussion of the films of Frederick Wiseman, Thomas Benson and Carolyn Anderson see this breadth as a marker of truth to reality, "No other filmmaker has more to say to us about the American language than Frederick Wiseman. In film after film he has shown us the structure and uses of the American idiom, inviting us to listen, at length, to conversational passages that most other filmmakers would have left on the cutting-room floor." (Benson 1984: 31) While this breadth promotes a rich diversity, it presents obstacles for the viewer's understanding. Regional accents, slang, and idiosyncratic syntax make documentary representations of speech more difficult to understand than their fictional counterparts. In order to assure comprehension, Pincus had to subtitle the conversations of some of the children who appear in Black Natchez (1967), his film about civil rights struggles in Mississippi in the mid-1960s. However, subtitling may imply deviance from an assumed linguistic norm.

Observational films often do not succeed outside their national boundaries because of the difficulties presented for viewers who are not native speakers of the language. Part of the delight comes from hearing the material texture and richness of unrehearsed speech, the grain of the voice. (Marcorelles 1973: 63)

Speakers in everyday life typically fill in the gaps of their phrases with various exclamations and sounds that maintain the flow of verbal communication. In conversation, we interrupt one another, digress, ask questions, hem and haw. Telephone conversations exemplify these characteristics of spoken language. The absence of non-verbal cues necessitates a constant use of verbal signals to indicate that the listener is in fact awake and listening. Much of verbal communication consists of what sociolinguist Dell Hymes calls the phatic function of speech, the banal pitter-patter that signifies sociability, "talk for the sake of something being said." (Hymes 1972: 40) Anyone who has ever transcribed interview tapes recognizes the differences between the conventions of spoken and written language. Characters in Hollywood films typically speak scripted versions of spoken language and are careful not to interrupt one another's lines. In addition, from one take to another, actors must be capable of maintaining virtually identical volume, pitch, tone, and inflection in the delivery of their lines for continuity purposes, a talent for which they are handsomely paid. Dialogue in observational documentaries overlaps considerably as characters interrupt one another, speak at the same time, and affirm their listening stances. As Michel Marie remarks of synchronous sound recording techniques, "Direct is really a manifestation of a new modality of voice recording in film." (Marie 1979: 39) Interview films attempt to circumvent the fullness of ordinary speech in various ways. Staged to be filmed, interviews may be miked for maximum intelligibility of speech.

Documentary makers learn how to stage interviews so that the interviewee will appear to speak directly to the viewer. Michael Rabiger instructs in Directing the Documentary, "During the interview, you should maintain eye contact with your subject, and give visual (NOT verbal!) feedback while the interview goes on. Nodding, smiling, looking puzzled, signifying agreement or doubt are all forms of feedback that can be relayed through your expression." (Rabiger 1987: 59-60) In an article in the New Yorker, Errol Morris, director of The Thin Blue Line (1988), describes the importance of providing these non-verbal cues: "'Listening to what people were saying wasn't even important,' he says. 'But it was important to look as if you were listening to what people were saying. Actually, listening to what people are saying, to me, interferes with looking as if you were listening to what people were saying.'" (Singer 1989: 48) Interview films increase the clarity and directness of speech through editing techniques and shooting conventions.

Roger Silverstone makes note of this process in the shooting of a BBC documentary, in which the director instructs the interviewee to answer in full sentences so that the questions may be left out of the soundtrack,

M: Say that again because you spoke while I was speaking.
L: Stability is the key word in terms of what he is, would be, receptive to. [....]
M: So he's a tougher judge than scientific colleagues almost?
L: Exactly.
M: Say that again from the start.
L: What, about the . . .?
M: Yes, as a sentence. (Silverstone 1985: 69)

Interview films permit a mise-en-scene of speech, a trimming of the materiality of conversational speech in favor of clarity and comprehensibility.

In the 1970s and 1980s, independent documentary filmmakers returned to the direct address style of interview films in part because they allowed for greater control over what was happening in front of the camera. Films like Julia Reichert and Jim Klein's Union Maids (1976) and Seeing Red (1984), Peter Adair's Word is Out (1976), Connie Field's The Life and Times of Rosie the Riveter (1980), and Noel Buckner's The Good Fight (1984) allowed for more thorough preparation during the pre-production phases of fundraising and writing. Connie Field describes this process for Rosie the Riveter: "We did extensive preinterviews--seven hundred women were interviewed over the phone, two hundred in person on audio tape, thirty five were videotaped; and we filmed five." (Zheutlin 1988: 237) This passage also suggests the importance of casting in documentary. The use of a string of interviews permits a stronger sense of textual voice, not unlike that of a voice-over dispersed across multiple characters. (Nichols 1983) While voice-over narration and interviews allow for more direct sound in documentary, they remain marginal techniques in observational cinema. "Interviews" that appear in observational cinema are carried out by characters who appear in the films, such as the psychiatrists in Wiseman's Titicut Follies who interview the incoming patients about their medical histories, thereby introducing us to the characters and the institutional procedures in the film.

Location sound recording in observational documentaries does not clearly differentiate foreground and background spaces; rather, all sounds compete together in the middleground. The lack of clarity of the sound undermines the communicative intent of these films. Shotgun microphones are frequently used in documentary productions precisely because they allow for a choice of narrative information and raise the ratio of direct to

reflected sound, thus isolating sounds in the environment. Directional microphones enable recordists to place certain sounds in the foreground while relegating other sounds to the background. Instructions for location sound recordists include standing as close as possible to the speaker without appearing in the frame. Observational cinematographers prefer viewfinders which allow them to see beyond the frame of the film; the perfect space for the roving microphone. In this way, the cinematographer has a constant view of the microphone and the location of the soundperson, while the viewer never sees them. In Alan Raymond's words, "The camera/sound team must develop a kind of choreography where both parties are aware of each other all the time. The cameraman must listen to the dialogue and the sound recordist must watch what the cameraman is shooting." (Raymond 1973: 594) The development of microphone technology has been guided by demands for clarity as well as fidelity. Wireless lavalier microphones fulfill similar stylistic requirements. Extremely small and unobtrusive, they are designed to be worn on the chest of individual speakers, reproducing the human voice with great fidelity at the expense of the ambient sound environment.

Occasionally in observational films, poorly recorded scenes are included because of their central importance to the story. In An American Family, a conversation between the Loud couple at a crowded restaurant is virtually inaudible due to the presence of competing ambient sounds. A determined viewer may overcome the marginal quality of the sound track to catch snippets of the Loud's argument, an important indication of the downtrun of their marriage. Pat's comments are the harshert heard throughout the series: "I think that the things that you do are shitty. And perhaps you think that the things I do are shitty, that's your problem. But I think that you're a goddamn asshole." Narrative considerations necessitated the use of less than technically satisfactory footage. Although the sound is muddled, the point is clear. Anything that contributes to our understanding of the drama of the divorce finds its way into the narrative. The filmmaking team admitted that such technically difficult, and ethically sensitive, scenes would have been impossible to film without unobtrusive wireless lavalier microphones. (Raymond 1973: 605)

Observational shooting and editing techniques conform to continuity conventions established in the film industry during the silent era. Hollywood studios developed elaborate means of maintaining continuity through supervision of scripts, props, lighting, performance, and shooting style, means typically unavailable to documentary filmmakers. David Hanser, series editor of An American Family, describes how continuity conventions dictate the shape of the material, "If somebody says what they're gonna do that night or the next weekend, maybe that's useful, knowing what else you have. Then you try to make sure that you include

that in the editing, cause we're trying to tell the story without any narration." (Ruoff 1989b: A7) Continuity conventions make it difficult to edit together actuality material shot at different times, because characters will likely be wearing different clothing or they may change their appearance from one day to another. For these reasons, observational films usually follow the chronological order in which they were shot. Observational filmmakers do not necessarily believe that chronology best represents actual experience, but a particular system of narrative causality and continuity leads them to tell stories in chronological order. Reflecting on the difficulty of maintaining continuity in An American Family, the coordinating producer exclaimed, "The problem with public television is there are no commercials." (Ruoff 1989b: A14) As this comment suggests, commercials provide a convenient way to make transitions between scenes.

Music in Documentary

Music plays an important part in the soundscape of documentary films. Many of the classic documentaries of the 1930s, like Harry Watt's Night Mail (1936) and Pare Lorentz' The River (1937), were scored by famous composers such as Benjamin Britten and Virgil Thomson. Highlighting the importance of sound in their titles, Dziga Vertov's Three Songs of Lenin (1934), Basil Wright's Song of Ceylon (1935) and Humphrey Jennings' Listen to Britain (1942) make extensive use of music. Stephen Mamber suggests that documentary filmmakers in the 1960s moved away from these techniques, "In line with this commitment [to an observational ethic], some of the standard devices of fiction film and traditional documentaries fall by the wayside, especially music and [voice-over] narration." (Mamber 1974: 4) While the rhetoric of observational cinema demanded the exclusive use of images and sounds recorded during the synchronous shooting events, the new conventions did not preclude the use of music in documentary. On the contrary, music was fine as long as it was diegetic, and throughout the 1960s, there was plenty of music in observational films. In fact, in this period, the documentary musical emerged as a distinct subgenre, focusing primarily on the sounds of the new youth counterculture of rock music.

Just as the coming of sound fostered the growth of the Hollywood musical in the 1930s as a genre of spectacle and pleasure, innovations in location sound recording technology led to the rock documentary in the 1960s. The Maysles' What's Happening! The Beatles in the U.S.A. and Gimme Shelter (1970), Leacock's A Stravinsky Portrait (1964), Pennebaker's Don't Look Back (1966) and Monterey Pop (1968), and Michael

Wadleigh's Woodstock (1970) offer views of performers which incorporate music into the structure of the films. The rock documentary brings together the documentary cinema's traditional focus on actuality and the fictional cinema's emphasis on stars and spectacle. Wiseman's Titicut Follies uses a musical revue performed by mental patients as a framing device, a grotesque inversion and parody of the music documentary, in which the patients sing, "Have You Ever Been Lonely," "I Want to Go to Chicagotown," and "So Long For Now." Still today, music documentaries are among the most popular of non-fiction forms, with commercially successful works such as Martin Scorsese's The Last Waltz (1979), Michelle Parkinson's Gotta Make This Journey: Sweet Honey in the Rock (1985), George Nierenberg's Say Amen, Somebody (1985), Jonathon Demme's Stop Making Sense (1985), Stevenson Palfi's Piano Players Rarely Ever Play Together (1985), Susan and Alan Raymond's Elvis '56 (1986), Bruce Weber's Let's Get Lost (1988), Charlotte Zwerin's Straight: No Chaser (1990), and Les Blank's French Dance Tonight (1990). The canonization of rock-documentaries is only more apparent through successful generic parodies like Rob Reiner's This Is Spinal Tap (1984).

Recorded music appears frequently in observational documentaries. As in the early days of sound film, a shot of a radio or record player often signals the diegetic source of recorded music. While the filmmakers want to indicate that the music was found on location, this practice is also the result of legal and financial concerns. Filmmakers believe that if they can prove that they are using a musical segment as a social document, they will not be obliged to pay users' fees to the copyright owners. In a recent appearance at the Ohio Film Conference in Athens, Wiseman argued that his extensive use of location-recorded popular music by bands like the B-52s in Model (1980), for which no fees were paid, would be defensible in court. Negotiating for the rights to use contemporary music in film is a notoriously difficult and expensive process, forcing some filmmakers to avoid such scenes altogether. Tony Buba addresses this issue in an amusing scene in Lightning Over Braddock in which a song by the Rolling Stones is not heard on the soundtrack. As we watch a mock performance of the song by local teenagers in a bar in Braddock, Pennsylvania, Buba tells us in voice-over that the rights to the song, which was played on the jukebox, would have cost $10,000. He remarks that if he paid such an extravagant amount of money, from his low-budget film about the economic downturn in the rust belt during the Reagan years, St. Peter wouldn't allow him into Heaven.

In An American Family, after Pat Loud has instructed her husband to move out of the house, we see her languishing by the poolside. On the soundtrack we hear the strains of Carole King's "Will You Still Love Me Tomorrow?," which Delilah is listening to in the adjoining bedroom. The bitterness of the breakup after twenty years of marriage is not lost on the

viewer. This scene neatly echoes that of Bill and his new lover dancing on New Year's Eve as a piano man croons King's "You've Got a Friend," from the same album "Tapestry." Although documentary filmmakers often imply in interviews that such incidents simply happen and are just happy coincidences, their use clearly demonstrates an authorial intention on the part of the makers, a sense of aesthetic and thematic unity, and an implicit point of view. The music comments on the action, providing an editorial perspective for interpreting the images, as Claudia Gorbman has noted of the function of narrative film music in general. (Gorbman 1988)

While the conventions of observational film require that music be recorded on location, the function of music in the narrative structure of these films appears quite similar to that of music in classical Hollywood cinema. Music provides continuity, covers up edits, facilitates changes of scenes, provides mood, offers entertaining spectacle, allows for narrative interludes and montage sequences, and comments on the action. Eleven of the episodes of An American Family open with musical passages, while ten episodes end with music over the credits; these musical passages bracket the programs. Wiseman begins High School with a car radio playing Otis Redding's "Sitting on the Dock of the Bay" and the chorus about "wasting time" quickly comes to stand for the experience of the students at Northeast High. For the most part, Wiseman avoids such commentative uses of music in his later films. The virtual absence of music in Hospital no doubt contributes to its oppressive atmosphere of suffering and pain. The young man who overdoses on mescaline in that film begs his attendants to "play some music or sing" to relieve his anxiety. By and large, documentary filmmakers have become as rigorous as their Hollywood counterparts in finding musical passages that contribute to the narrative and thematic concerns of their films. Barbara Kopple's Harlan County U.S.A. (1976) profits from a rich and moving sample of folk songs that shows music to be a repository of community and memory in the miners' struggle for their civil rights.

The new conventions of observational documentary in the 1960s required that filmmakers find their musical selections in the original scenes that they filmed. Not surprisingly, the ordinary people who populate observational documentaries often happen to play music themselves. In Marc Obenhaus' The Pasciaks of Chicago (1976) from the television series Six American Families, the son's dedication to rock music creates tensions within his Polish working-class family; his mother wants him to play traditional Polkas on the accordion and his interest in contemporary rock music reflects his attempt to assimilate into the mainstream of American popular culture. Nick Broomfield and Joan Churchill's Soldier Girls (1981) perform a lively rap tune in the barracks during a break in their basic training sessions in Georgia.

In An American Family, young Grant diligently plucks away on his guitar and sings in a budding garage band. During the course of the twelve-hour series, we hear the band perform the Rolling Stones' "Jumping Jack Flash" and "Brown Sugar," and the Who's "I Can See For Miles" and "Summertime Blues." These regular appearances culminate in an amusing parody of rock concerts staged by Grant's band at a pep rally at Santa Barbara High School. With his friends cheering in mock hysteria, Grant arrives on a motorcycle wearing black leotards, a skin tight black shirt and cape, to sing "I'm Gonna Get You In My Tent," a song of his own composition.

This sequence shows how thoroughly Grant, Kevin, and their friends had absorbed the ecstatic performance style of Mick Jagger, canonized in the Maysles brothers' portrait of the Rolling Stones, Gimme Shelter. The sequence of Grant's band also suggests that the value of music as entertainment and spectacle was recognized by the Raymonds, Gilbert, and the editors of An American Family. Later, Grant serenades his mother in the living room with acoustic versions of the Kinks' "Ape Man" and the Beatles' "Mother Nature's Son," joking self-consciously, "It's entertainment time at the Loud house. Will you get situated, please." After the series was broadcast, in a turn of events in which life may be seen imitating art, the Loud children performed several songs as a group in a televised fundraising event for PBS. Meanwhile, Lance attempted to capitalize on his newfound celebrity by forming the Mumps, a punk rock group that played original music in clubs in New York throughout the 1970s. In the Raymonds' re-make An American Family Revisited: The Louds Ten Years Later (1983), Grant pursues a career as a lounge singer in night clubs in Southern California.

Conventions of Observational Film

While An American Family may be the most famous example of the observational cinema, it deviates significantly from the proscriptive rules of that style. These stylistic ruptures shed light on the conventions of documentary expression in broadcast television and independent film. At the outset of each episode, a split-screen montage sequence and a musical theme song introduce the family members one by one. Gilbert commissioned Elinor Bunin to make this one-minute series title film. The sequence ends with the title shattering like glass, an obvious indication of the break-up of the family that forms the basis and narrative unity of the whole series. The series title film indicates to the viewers that they are entering the frame of a narrative, the world of a story. The musical introduction was not recorded during the shooting, an important violation of the observational aesthetic. In this passage, Gilbert can be seen conforming to the standards

and conventions of broadcast television, rather than independent film, a balancing act that becomes more apparent over the course of the series.

The opening episode uses on-camera narration as well as recorded images and sounds to introduce the characters, the time and place of the setting, and the dramatic tension. An American Family opens with a shot of producer/director Gilbert standing on a hillside above a city. He speaks directly to the audience, "During the next hour, you will see the first in a series of programs entitled An American Family. The series is about the William C. Loud family, of Santa Barbara, California." Gilbert's opening monologue, in which he attempts to frame audience expectations and deflect possible criticisms, is a central deviation from the observational model. He apparently believed that the novelty of the form and subject matter necessitated this introduction. He denies that the Loud family is either average or typical, thereby contradicting the thrust of the series title. Viewers nevertheless understood the significance of the title; still today it is commonly referred to as "The American Family." Gilbert also admits that the filmmaking inevitably influenced the Louds, although no evidence of this may be seen in the series. Furthermore, he asserts that the series was a cooperative venture "in every sense of the word," although few traces of this interaction appear. In this way, the tensions of the textual system are displaced onto the opening monologue, a displacement which also occurs in John Huston's The Battle of San Pietro (1945) in which an introductory statement from a representative of the army attempts to deflect the main point of view of the film. (Nichols 1981) Similarly, the on-camera narrator who introduces Jennings' Listen to Britain, on the 16mm print in distribution in North America, deliberately calls attention to "the first sure notes of the march of victory, as you and I listen to Britain."

Gilbert continues his description of the setting of An American Family, "The population of Santa Barbara is somewhere around 73,000. Located on a slope of the Santa Ynez Mountains, Santa Barbara faces south on the Pacific Ocean, ninety miles north of Los Angeles--driving time an hour and a half, flying time twenty minutes. The average daytime temperature is 78 degrees in summer and 65 degrees in winter." Throughout this introduction, Gilbert sounds curiously like the Stage Manager in Thornton Wilder's Our Town, who sets the stage for a re-enactment of family life both particular and universal. Like the Stage Manager, Gilbert remains somewhat detached from the everyday affairs of ordinary mortals. After introducing the characters and the setting for the series, he disappears. Indeed, Our Town presents a model for understanding An American Family as social commentary; both make appeals to family, community, and nation. The camera and tape recorder serve as witnesses to the events, an internal audience like the Stage Manager or Emily who, after her death,

returns to observe the living going about their everyday lives. In Emily's eyes, they are painfully unaware of their singualarity and existential importance. Many of the reviewers of An American Family, in the subsequent controversy generated by the series, echoed Emily's angst in their critiques of the family and the film.

In both play and film, viewers are invited to see everyday events with their ordinary significance heightened. Gilbert's voice-over narration gives away the drama of the series within the first thirty minutes, so that the viewer is primed to read all of the events of the Loud's unfolding lives as telltale signs of the inevitable decline of Pat and Bill's marriage, "This New Year's will be unlike any other. This New Year's will be unlike any other that has been celebrated at 35 Wooddale Lane. For the first time the family will not be spending it together. Pat Loud and her husband, Bill, separated four months ago after twenty years of marriage." As in Greek drama and popular film genres, we know the ending already and the suspense lies more in the telling than in the tale. Here, Gilbert embodies the voice of God, knowing the end and the beginning.

At another similar point in the film, Gilbert presents new story information in the voice-over narration, a trade-off, as the coordinating producer Jacqueline Donnet described it, between the comprehensibility of the narrative and the conventions of observational film, "There's [a scene] in the show where she is telling her brother that she's gonna divorce [Bill] and she wants to separate and get a divorce. That was the other time that we wrenched the convention of not telling you, was by telling you up front that she was driving down to speak to her brother to tell about the divorce. Mainly the reason for that was the soundtrack is so garbled; it's very, very hard to hear and our feeling was you miss half of it and all of a sudden realize what she's saying and you haven't heard the front half, so that was the other time we pinpointed something." (Ruoff 1989a: B7) Voice-over presents essential story information and restores a potentially incomprehensible scene. Throughout the rest of the twelve-hour series, Gilbert's voice-over occasionally resurfaces to provide details about the time and place of the action.

Gilbert, however, is not the only one to speak in voice-over. Both Pat and Bill Loud narrate voice-over passages. In her voice-over narration, Pat speaks her words haltingly, as if reading from a script or answering off-screen questions. Her voice lacks any specific spatial signature and the sound is uncharacteristically clear. She is literally not speaking in her own voice; a photograph published in Studies in Visual Communication shows Pat Loud and Craig Gilbert composing the narration for this sequence, a stopwatch and film projector within view. (Gilbert 1982: 49) The rehearsed style of her speech contrasts vividly with the spontaneity of her voice in the observational scenes of her daily life; her inexperience as a voice-over narrator becomes a touching signifier of the authenticity of her routine

appearances in front of the camera. In this scene, Gilbert at once undermines and reinstates the claims of observational filmmakers to observe human behavior without unduly influencing it. Later in the series, Bill reads in voice-over a letter that he wrote to his son Lance shortly after the parents' separation; his delivery shares many of the same qualities of awkward direct address as Pat's voice-over. There is no ambient sound at all in this scene as we see Lance bicycling through his home town.

When Lance travels to Paris, we hear the sound of accordion music; the clarity of the music and the complete lack of spatial signature suggest that it was recorded not on location where he was vacationing, but rather came directly off of a record. Needless to say, accordion music provides the most conventional associations with France, a convenient way of announcing the setting. Gilbert also uses canned music with the home movies which Pat Loud describes in voice-over. The scenes from childhood are accompanied by the sounds of a toy musical box. Another sequence of home movies, originally shot in Brazil where Pat lived as a young girl, unfolds with generic samba music in the background. Again Gilbert fulfills conventional expectations about the narrative television soundscape, which apparently cannot tolerate silence. The use of home movies and family photographs also represents an important deviation from the observational focus on images and sounds recorded in the present. One of the criticisms of observational film is that it fails to deal adequately with the past, an important limitation which Gilbert simply ignores in favor of other means of expression--interviews, voice-over narration, canned music, family photographs and home movies.

Interview material also occasionally appears in the series. When we see Lance for the first time, he delivers a long monologue about his family, entirely in voice-over, as we see him sorting his clothes "alone" in his room at the Chelsea hotel. Lance describes his family in ways that sound like a series of answers to questions not heard on the soundtrack, "I have two brothers and two sisters; Kevin, Grant, Delilah, and Michelle. I don't know any of their ages or any of their birthdays or anything like that. I can never remember anything of those private things of anybody, except my own." Lance goes on to comment that Michelle is really "selfish and snotty" and that Grant is "unfortunately" the one most likely to succeed. Lance's witty descriptions and criticisms help to introduce the personalities of his siblings, while underlining his own marginal status in the family. This uncharacteristic voice-over technique suggests that the film endorses Lance's point of view. In another instance, Lance recreated his half of a long-distance telephone conversation with his mother, "When they were editing the show, they took me in the recording studio, and they had me do overdubs. [...] They would just give me rough outlines of areas they wanted me to talk about." (Ruoff 1990: A4)

Although I have spoken little of sound effects in An American Family, one anecdote provides some illuminating information about them. After escorting his mother to a taxi, Lance returns to his room at the Chelsea hotel. Following him in a virtuoso long take, the camera climbs up four flights of stairs and enters his room, where Lance flips on the television in time to catch the beginning of Abbott and Costello's Buck Privates Come Home. The mood of this scene is melancholy, conveyed in some measure by the weary sounds of footsteps fading away in the empty hallways, appropriate enough for the life of a boy from Santa Barbara living alone in New York. Unfortunately, no synchronous sound was recorded to accompany this scene. Editor David Hanser returned to the hotel some time later and re-created the cavernous sound of footsteps by walking up the stairs, carrying a Nagra recorder, at exactly the same pace as Lance had done. For the sound of the television show, Jacqueline Donnet stated, "we knew exactly the day and the time it was shot and went to the New York Times and found out what film it really was, rented the film--and that was the exact same print that they had on the air--and just added it." (Ruoff 1989a: AA11) No viewer ever noticed this betrayal of the conventions of synchronous sound recording. This supports Michel Chion's contention that we have virtually no way of knowing for sure, short of confessions from the filmmakers, whether sound effects in documentary were actually recorded at the time of shooting. (Chion 1985) Donnet admitted that she "wouldn't trust anyone with audio in terms of saying that was really on the track or that wasn't on the track." (Ruoff 1989a: AA11)

Throughout this essay, I have attempted to characterize some of the conventions of documentary sound and narration, concentrating in particular on the style of observational filmmaking and An American Family. In both fiction and documentary, clarity of sound and image derives from the degree of control that filmmakers have over profilmic events. Interviews and voice-over narration in documentary provide exceptionally clear and direct sound, although observational filmmakers try to avoid those techniques. Observational documentaries still make extensive use of music, even if non-diegetic music falls by the wayside. Synchronous sound observational documentaries borrow conventions of storytelling and continuity editing from fiction films, without, however, offering the clarity of image, sound, and story that are the hallmark of classical Hollywood cinema. Close analysis of An American Family indicates that conventions of observational cinema were circumvented in numerous ways in order to make the narrative more comprehensible, suggesting the director's commitment to other forms of documentary address. Forms of direct address occur in on-camera narration, voice-over narration, canned music, and interviews. This twelve-part PBS series represents a unique

mixture of conventions of broadcast television and independent documentary film. The ruptures in the sound track suggest not that strictures for making documentaries were violated, or that audiences were necessarily deceived, but rather that all films are constructions, meaningful assertions about the world made by directors and those with whom they collaborate.

Bibliography

Allen, Robert C., and Douglas Gomery
1985 Film History: Theory and Practice. New York: Alfred A. Knopf.

Altman, Rick
1984 "Towards a History of Representational Technologies," Iris, 2(2): 111-126.

1985a "The Evolution of Sound Technology," in Film Sound: Theory and Practice, Elizabeth Weis and John Belton, eds. New York: Columbia University Press, pp. 44-53.

1985b "The Technology of the Voice I," Iris, 3(1): 3-20.

1986 "The Technology of the Voice II," Iris, 4(1): 107-118.

1990 "Sound Space," Unpublished.

Benson, Thomas and Carolyn Anderson
1984 "The Rhetorical Structure of Frederick Wiseman's Model," Journal of Film and Video, XXXVI: 30-40.

Carroll, Noel
1988 Mystifying Movies: Fad and Fallacies of Contemporary Film Theory. New York: Columbia University Press.

Chion, Michel
1985 Le son au cinéma. Paris: Editions de L'Etoile.

Gilbert, Craig
1982 "Reflections on An American Family," Studies in Visual Communication, 8(1): 24-54, Winter.

Gorbman, Claudia
1988 Unheard Melodies: Narrative Film Music. London: BFI.

Hindman, James and Victoria Costello, eds.
1983 The Independent Documentary: The Implications of Diversity, A Conference Report. Washington, DC: American Film Institute.

Hymes, Dell
1972 "Models of the Interaction of Language and Social Life," in Directions in Sociolinguistics, John J. Gumperz and Dell Hymes, eds. New York: Holt, Rinehart and Winston, Inc, pp. 38-56.

Izod, John
1988 Hollywood and the Box Office, 1895-1986. New York: Columbia University Press.

Jones, D. B.
1988 "The Canadian Film Board Unit B," in New Challenges For Documentary, Alan Rosenthal, ed. Berkeley: University of California Press, pp. 133-147.

Kozloff, Sarah
1988 Invisible Storytellers: Voice-Over Narration in American Fiction Film. Berkeley: University of California Press.

MacDougall, David
1975 "Beyond Observational Cinema," in Principles of Visual Anthropology, Paul Hockings, ed. The Hague: Mouton, pp.109-124.

Mamber, Stephen
1974 Cinema Verite in America: Studies in Uncontrolled Documentary. Cambridge: MIT Press.

Marcorelles, Louis
1973 Living Cinema: New Directions in Contemporary Filmmaking. London: George Allen & Unwin Ltd.

Marie, Michel
1979 "Direct," in Anthropology-Reality-Cinema: The Films of Jean Rouch, Mick Eaton, ed. London: BFI, pp. 35-39.

Nichols, Bill
1983 "The Voice of Documentary," Film Quarterly, 36(3): 17-30, Spring.

O'Connell, P.J.
1988 "Robert Drew and the Development of Cinéma-Vérité in America," Unpublished Ph.D. Dissertation, Penn State University: Department of Speech Communication.

Rabiger, Michael
1987 Directing the Documentary. Boston: Focal Press.

Raymond, Alan and Susan Raymond
1973 "An American Family," American Cinematographer, 54: 590-1+, May.

Ruoff, Jeffrey K.
1988 "'Nothing in our films is ever not in sync': The MIT Film Section." Paper presented at the Society for Cinema Studies and University Film and Video Association in Bozeman, MT.

1989a "Interview with Jacqueline Donnet." Unpublished.

1989b "Interview with David Hanser." Unpublished.

1990 "Interview with Lance Loud." Unpublished.

Silverstone, Roger
1985 Framing Science: The Making of a BBC Documentary. London: BFI.

Singer, Mark
1989 "Predilections," New Yorker, February 6, 1989.

Zheutlin, Barbara
1988 "The Politics of Documentary: A Symposium," in New Challenges For Documentary, Alan Rosenthal, ed. Berkeley: University of California Press, pp. 227-244.