1
MOVIES AND SONGS IN TRANSITION
THE PROBLEMS OF FILM MUSIC THAT AROSE DURING the transition to electric sound were cast in a language with a long history. The project of reconciling the formal patterns of popular music with the demands of dramatic narrative long preceded sound movies.1 For centuries dramatists had debated options for bringing together music and theatrical narrative, to meld the two into a single form, and this history of music-drama theory informed discussion and debate on the aesthetics of sound films.2 The main problem for producers of feature-length fiction films was the tendency for the typical song sequence to come across as “an isolated interlude [isolierte Einlage] in the film,” as a critic in Germany put it.3 In coming forward as self-contained events, song performances “hindered the natural course of the plot,” observed a journalist in London.4 In the entertainment press, much was made of the need to weave songs into a film’s narration. “The introduction of songs seems to be the most acceptable when they further the plot motivation,” stated an American journalist.5
Narrative integration remains a core desideratum today, when, as in the 1930s, song sequences that can be taken to advance the film’s story draw critical praise. Narrative considerations inform the American Film Institute’s list of “one hundred greatest film songs,” chosen because they “set a tone or mood, define character, advance plot and/or express the film’s theme.”6 The narrative motivation practiced by filmmakers, however, is not the formal integration promoted by music-drama theorists, and the latter proved difficult to achieve in the domain of the feature film. Attempts in the early 1930s to extend the tight link between music and image in song sequences to a full-length film through techniques such as the synchronization of actors’ movements to the music and the use of verse-like dialogue were relatively rare. For scenes with spoken dialogue, “the abstract time of the musical progression,” proposed a critic in Germany, “must be adjusted to the real time of the dramatic action.”7 The difficulty of this task is suggested in the frequency with which dialogue scenes in early sound films play with no music whatsoever.
Rather than try to render the whole film musical, filmmakers typically interpolated songs as distinct units, different from the film’s other scenes. The songs normally have a narrative motivation. A singer’s performance is almost always also a story event—an audition, a rehearsal, a revue’s opening night, a neighborhood talent show. But once the performance begins, it comes across as somewhat separate from the rest of the film, functioning more as an interlude in the story than a continuation. Filmmakers became adept at assigning the style difference a thematic function, as when the song sequence’s formal transcendence suggests a utopian rejoinder to the personal and social deprivations invoked in the film’s dramatic scenes.8 This is reflected in the critical writing on the American film musical, where the divide between “narrative” and “number,” between the music-driven song sequence and the dialogue-based dramatic scene, defines the genre’s basic syntax, its characteristic way of organizing scenes and sequences into an overall structure.9
The willingness of filmmakers to ignore in practice the ideal of formal integration can be attributed to multiple causes. Later chapters explore the special production conditions needed for song sequences and the commercial pressures of the larger entertainment culture. The focus of the current chapter is on inherent formal differences between songs and narrative films and how these differences factor into how song sequences are constructed. The chapter begins with an overview of how music in sound cinema differed from that in silent films, with an emphasis on the problems and possibilities associated with songs specifically. Illustrating the range of sound-era options are two extremes: the operetta, in which formal consonance between music and image occurs during not only the songs but also the dialogue scenes, and the variety show mix of acts in the revue film, in which intermedial dissonance is allowed full reign.
Film Songs after Recorded Sound
Songs accompanied motion pictures from the beginning.10 In the late 1920s, however, new technical conditions made them unusually conspicuous. For one, the songs now came exclusively in the form of recordings rather than live performances, which enabled a massive escalation in the showcasing in films of star vocalists known to the film audience through multiple media other than cinema, including the robust new medium of broadcast radio. The singers’ profiles in the media culture at large shaped how song sequences were designed, as is discussed further in chapter 3.
The second technical change concerns the sound-on-film systems from companies such as Fox, Tobis-Klangfilm, and RCA, whose permanent bonding of music and image onto a single celluloid strip allowed visual action and music to work together in powerful new ways. The exact synchronization of music and image had been rare in cinema prior to electric sound, when projection speeds and musical talent varied from one exhibition venue to the next and theater musicians could not be counted on to match music and visual action with precision. In movie houses in the 1920s, observed critic Emile Vuillermoz, “the mismatch [décalage] of music and image was the norm and coincidence the exception.”11 The uneven conditions of silent-era film exhibition, film-music scholar Kathryn Kalinak explains, made composers reluctant to “encourag[e] synchronization between the tempo of the action and the tempo of the music.” What was favored instead was “the more loosely defined concept of mutual correspondence, in which the structural properties of music (its tempo, rhythm, or harmonics) or its associated powers were loosely matched to the implied narrative content.”12 The Hollywood film scores of the mid-1920s, Kalinak adds, rarely include cues requiring an exact match of music with a film’s visual representation.13
Film-music practice changed radically with the sound-on-film systems of the late 1920s. In making music-image synchronization not only possible but inevitable, the new systems led to the proliferation of what media scholar Amy Herzog calls “musical moments,” which occur “when music, typically a popular song, inverts the sound-image hierarchy to occupy a dominant position in a filmic work.”14 Musical moments were pervasive in cinema around 1930, when songs, for the first time since the nickelodeon era, become the dominant form of film music.
The Ideal of Formal Unity
How did producers of feature-length narrative films respond to the opportunities and challenges of music-image synchronization? One option sought to take the close match of music with visual action characteristic of the period’s animated shorts and extend it to an entire film, the dialogue scenes included.15 The goal, in effect, was to utilize the tight synchronization that the Disney team had achieved in the Silly Symphonies in a feature-length production.16 This project of rendering musical a full-length film was promoted by theorists and pursued by a small number of ambitious filmmakers. Ernst Lubitsch at Paramount, René Clair at Tobis Films Sonores, and various German directors, such as those working for Erich Pommer’s unit at Ufa, are salient examples. But they are also exceptional. The majority of the period’s producers were inclined to construct a film’s dialogue scenes differently from the song sequences rather than try to unify them within a common, music-based style.
If the maximum integration aimed for by the operettas represents one extreme, then the period’s revue films, with their mix of acts featuring diverse singers, comedians, dance troupes, and other entertainers, can be said to illustrate the opposing possibility. Dispensing with narratives altogether, The Hollywood Revue of 1929 (dir. Charles Reisner, 1929), The Show of Shows (dir. John Adolfi, 1929), Paramount on Parade (dir. Edmund Goulding et al., 1930), King of Jazz (dir. John Murray Anderson, 1930), and Elstree Calling (dir. Jack Hulbert et al., 1931) instead present a series of variety show performances introduced by a master of ceremonies. A variant involved framing the performances with a minimalist plot about the making of a variety show, as in On with the Show (dir. Alan Crosland, 1929), Happy Days (dir. Benjamin Stoloff, 1930), and Wir schalten um auf Hollywood (dir. Frank Reicher et al., 1931). Such films fall into the category of narrative cinema, but only barely.
The dissonance inherent to the variety format was most pronounced in the revue films intended for foreign release, where the original American scenes were sometimes supplemented with footage shot in studios outside the United States. Seven foreign versions of Paramount on Parade, for example, were assembled at Paramount’s Paris studio, where new sequences, added to those made in Los Angeles, featured celebrities and musical acts from France, the Netherlands, Romania, Czechoslovakia, Poland, Italy, Sweden, and Denmark.17 The four foreign versions of The Hollywood Revue of 1929 were made in a similar fashion, with, for example, the French version incorporating extra French-language material shot at the Pathé-Natan facility in Joinville near Paris.18 Acts were added in or taken out depending on how the distributors had assessed the acts’ popularity in the version’s target market. Continuity, as it were, was provided by an on-screen impresario who introduced the individual acts, “speaking in the language of the country in which the film [was] being shown.”19 Bela Lugosi, soon to star in Universal’s Dracula (dir. Tod Browning, 1931), played this role in the Hungarian version of King of Jazz.
Reproduced below is a promotional still for Der Jazzkönig, the German version of King of Jazz, in which character actor Arnold Korff intr...