Stay informed: Sign up for eNews Subscribe
Read Part 1

Part One

Playing the Pictures

Music and the Silent Film, 1895-1925


That film historians have only recently begun to recognize film not as a uniquely visual art but as a highly integrated one, one that unites the previously separate mediums of image, sound, and music, stems no doubt from a peculiarity rooted in the beginnings of film history itself. For nearly the first three decades, film was not a fully mechanized art; instead it relied on a strange simultaneity of technically disparate parts, merging mechanically reproduced moving images with live performances of music and sound. Unfortunately, this mix of real and reproduced media led instantly to a critical inequity. Because it was mechanized and represented a new technology, the visual part of the film came immediately to define the film proper. Then as now, film was prized primarily as a visual technology or art.

The history of silent film music is important not only because it challenges this visual-centric model of film, but also because it offers a new and deeper understanding of the term silent. Certainly, if film had been truly silent from the beginning, this section would not exist. If the following documents reveal anything, it is the irony of the most common term used to refer to early film. For in fact, the "silent" period was full of sounds-noise, music, even dialogue and narration. In qualifying the silence of early film, therefore, the documents in this section redefine that silence not as a lack of sound, but as a lack of integration. The undoing of film's silence, in other words, will come not with the inclusion of music and sound, but with their mechanization, the technological innovation that allowed music to be represented alongside the images.

The history conveyed by these documents of music in the silent period also challenges the assumption that the sound of film music was standardized or "classicized" by Hollywood composers of the 1930s and 1940s. Far from being the beginning of a classical tradition of wall-to-wall orchestral music, Max Steiner's "Golden Age" scoring model was, rather, the culmination of three decades of silent film music experimentation.

Like any art form, film has an extensive prehistory. Beginning as early as the sixteenth and seventeenth centuries, inventors and artists had been seeking ways not just to represent reality, but to animate it. It wasn't until the nineteenth century, however, with the invention of photography and the cinema-like experiences of the diorama and panorama, that those efforts took off, and as a result audiences were now treated to a flourishing of optical experiments. While some of these early experiments were more popular than others, none was completely successful at both animating the photograph and projecting it onto a large screen for mass viewing.1 That distinction was achieved only at the end of the century, on December 28, 1895, when the brothers August and Louis Lumière, owners of a film and photographic plate manufacturing business in Lyon, France, projected a series of short films, or cinématographes as they called them, onto the wall of the salon indien at the Grand Café in Paris.2

In terms of basic technology, these films were similar to the films we watch today, although they were silent and radically shorter. Each was limited to the length of the reel, which at first was only about 25 meters, or one minute, long. These first films were also limited in subject matter. As a surviving playbill for the Paris exhibition reveals, the Lumière films were actualités, or proto-documentaries.3 While many featured the Lumière family and the city of Lyon, one of the Lumière brothers' first production initiatives was to place trained cameramen in major cities and exotic locations all around the world to document life outside of France. The result was short scenics, of places like Venice, Milan, Naples, and even Melbourne. They made these short documentaries not to entertain audiences, but to advertise their film equipment company to a global market.4

The cinema experience for the audience at the Grand Café was more than just a visual spectacle. In an important footnote at the bottom of the playbill, musical accompaniment by the pianiste-compositeur Emile Maraval was announced. Little is known today of Maraval, nor does the announcement reveal what kind of music he played, whether he penned new music or simply improvised an accompaniment for each short film. Perhaps he changed the style and tempo of his music to suit the topic of each film, giving Venice and Australia different treatments. It is impossible to know5 Nonetheless, the presence of a musician at this first film exhibition is significant. Not only did the Lumière brothers find a way to animate photographs and project them onto a screen in larger-than-life-size form, but they also thought to integrate those images with music.

When a similar group of Lumière cinématographes was shown to Queen Victoria several months later, the musical part of the experience was somewhat different. As the program for this special occasion announces, the Windsor Castle exhibition was accompanied not by a pianist, but by an orchestra, specifically the Empire Theater Orchestra conducted by Leopold Wenzel.6 Although, again, little is known of Wenzel,7 the exhibition program yields much more information about what the musical part of the experience may have sounded like. On this occasion, not only was the accompaniment orchestral, but it relied to some extent on preexisting music, including selections previously written by Wenzel for the ballet and works by other composers such as Karl Millöcker, Olivier Metra, Ernest Gillet, and Charles Gounod. Wenzel also seems to have differentiated filmic action and subject matter through use of varying musical styles and tempi. "Hussars Passing through Dublin," for instance, was paired with a march by Métra, and the comedy "A Joke on the Gardener" with a waltz by Wenzel, while the final travelogue, "A Moving Train near Clapham Junction," was accompanied by "Metropolitan Galop" by Charles Hubans.

Not all of the earliest film exhibitions followed the Lumière exhibition format, of course, or even used the Lumière technology. In Europe, for instance, early cinema audiences were treated to Max Skladanowsky's Bioskop films;8 in the United States, audiences watched "vitascopes," films projected by Thomas Edison's Vitascope machine.9 While individual technologies varied, all these films appear to have been shown with musical accompaniment.

In the United States at the turn of the twentieth century, film's first home was the vaudeville theater. In this setting, films were not shown as "programs," or uninterrupted collections of short films with different subjects, as in Europe. Instead, the film was simply part of the parade of individual acts that defined the vaudeville program. A short film of a modern dancer or two men boxing, for instance, might have been sandwiched in between a juggler and a comic routine.

The isolated appearance of film in the vaudeville setting fuels the first challenge to the argument that early silent films may in fact have been silent. According to the contemporary literature, it is unclear whether music was indeed heard during the film portions of this kind of program. Some reports make no mention of music, while others indicate its clear presence. In April 1896, for instance, when Koster and Bial's Music Hall in New York City hosted the first Vitascope projections, the hall's band was said to have provided "a musical accompaniment." When the Vitascope was premiered the following year in Philadelphia, not only was a musical accompaniment provided, but it was valued as an essential aspect of the new cinema experience: "The soldiers marched to the stirring tune of the 'Marseillaise' and the scene stirred the audience to such a pitch of enthusiasm that has rarely been equaled by any form of entertainment. The playing of the 'Marseillaise' aided no little in the success of the picture. In the sham battle scene the noise and battle din created also added to the wonderful sense of realism."10

George Beynon, an early historian of film music, however, describes a different understanding of music for film in the vaudeville tradition. In the introductory chapter to his 1920 instruction manual Musical Presentation of Motion Pictures, "Evolution of the Motion Picture," Beynon insists that in the vaudeville setting music was provided for everything but the motion picture. If music was heard during the film sections of the program, he asserts, it was there by default rather than by design. "The film was run in silence except for the beating of the big drum outside for the purpose of drawing the crowds ... [and] during the 'packing process' [when] the pianist regaled the seated ones with some music, mostly apropos of nothing." It was only by accident, Beynon concludes, that a conscientious musician "forgot himself as to play soft music for a particularly touching death-bed scene."11

Although a few current film historians, including Rick Altman, argue that early film might have been truly silent, at least on some occasions or in some venues,12 the majority of the evidence indicates the opposite. In fact, "silent" films appear, generally speaking, to have had downright noisy. As the Philadelphia critic at the Vitascope premiere observed, in addition to the musical accompaniment, films had sound effects, and attempts were even made to include spoken dialogue. As film historian Charles Musser sums it up, "Modern-day film producers distinguish four basic kinds of sounds: music, narration, effects, and dialogue. Of these, all but the fourth were commonly used during the first year of moving pictures. But even dialogue was employed within a short time as actors or singers were placed in back of the screen."13 Early silent film, in short, was hardly silent but instead hosted a number of additive sound features. And of all the initial accompaniments, music appears to have been most consistently included, for it gave film broad and democratic appeal, helping it to appeal to different classes of filmgoers at a variety of venues.

Over the course of the next decade, the transition of film from novelty act to independent art form brought a number of changes to both film and its exhibition. One of the most significant changes was the acquisition of an independent exhibition space devoted solely to the display of moving pictures, starting in about 1905. In the United States, storefront theaters soon became known as nickelodeons because of their nickel admission price.14 As new spaces for film exhibition began to flourish, and a system of film rental was standardized through the establishment of film "exchanges," film production also changed. Even before the nickelodeon model began to dominate, production focus was shifting from documentary-minded films to fictional or "story" films.15 With this shift in focus came an increase in film length. Whereas the first narrative films and actualities typically featured a single shot or "scene," emerging narrative films expanded to include multiple scenes (typically six to twelve), occupying anywhere from two or three hundred to a thousand feet of film. Longer films took up one reel and typically lasted from ten to eighteen minutes, depending both on their length and on projection speed, which were not standardized until 1909-10. Shorter films were typically half as long so that two could occupy a single reel (known as a "split reel")16

As films grew in length, they began to be organized by genre. The earliest films tended to fall into three categories: actuality (travelogues, newsreels, reenactments, etc.), comedy, and drama. By 1910, those genres had been expanded to include a host of subgenres: western, Indian film, war picture, detective serial, melodrama, trick film, farce or slapstick, fairy tale, biblical passion, and science fiction. It was this initial flowering, between 1905 and 1910, that saw the rise of such pioneer filmmakers as George Méliès, Edwin Porter, D. W. Griffith, and Cecil B. DeMille and early production companies like Pathé, Edison, Biograph, Vitagraph, Selig, Lubin, Kalem, and Essaney.17

Just as the expansion of narrative filmmaking encouraged the specialization of directors and genres, it also led to the standardization of exhibition practices, especially with regard to musical accompaniment. For instance, in a 1909 advice column for fellow nickelodeon owners entitled "Plain Talk to Theatre Managers and Operators" (Document 1), featured in the trade magazine Motion Picture World, theater manager F. H. Richardson describes the standard instrumentation of the musical accompaniment as piano and drums. A successful accompaniment, he notes, rested on the musicians' ability to differentiate musically not only individual film actions and episodes but film genres as well. Although Richardson does not describe particular musical selections-what one would have heard during a Civil War picture, for instance, or a western-he did say that the pianist needed to have good improvisation skills. This suggests that music for the cinema before 1910 included renditions of popular tunes and simple impromptu melodies. He also describes the placement of the musicians, at the front of the theater next to the screen, which not only allowed them to see what they were accompanying, but also helped create the illusion that their sound was emanating from the screen. A simple methodology is also available in Richardson's description. A successful accompaniment rested on the musician's ability to attentively follow and differentiate individual film actions and also larger film genres through musical style and tempo.

The musicians did more than set an appropriate mood for each film, however. As Richardson notes, they were in charge of providing all the sound for the film, which they did primarily by way of sound effects. In the very early days of film, the standards for "realistic" sound had been set high by traveling shows like Lyman H. Howe's High-Class Moving Pictures and Hale's Tours and Scenes of the World.18 Because these road shows toured with a single, unchanging program of films, they had an elaborate array of sound effect devices, everything from train, truck, and tractor sounds to gunshots, chimes, electric door bells, baby cries, roosters crowing and dogs barking, the clip-clop of horses' hoofs, and wind, rain, thunder, and ocean waves. Some of these devices, like thunder sheets and wind whistles, were borrowed from the theater; others were developed just for moving pictures.19

Since the film fare at the storefront nickelodeon typically changed three times a week, theater musicians did not have time to compose elaborate sound accompaniments, nor did they have as elaborate an array of sound effect devices as the road shows had.20 The nickelodeon drummer typically had some standard percussion instruments: drums, bells, gongs, woodblocks, and whistles. By 1905 or so, he or she would also have a number of additional percussive devices, or "traps." Most traps articulated common sounds, animal noises like hens' cackling; and the sounds of mechanical devices like winches, ratchets, and blacksmith anvils; and signals such as chimes and steamboat whistles. Traps were widely advertised in music and film journals of the early 1900s, and were used well into the 1920s, the golden age of silent film. For those exhibitors who could not afford the expense of lots of traps or multiple persons to articulate them, sound-effect "cabinets" became available starting around 1907. Semiautomated machines, like the Ciné Multiphone Rousselot, the Excela Soundograph, and the British Allefex, consolidated many of the most popular traps or film sounds in a large tabletop device that could be operated by a single percussionist. The fact that these traps and cabinets were conceived of as percussion instruments, to be played by musicians, is a reminder that in the early cinema, sound and music were not separate. In silent film, film music was film sound.21

By 1910, singers had become part of cinema music, though as Richardson points out, they did not typically accompany films. Instead, they provided interstitial material-specifically, the "illustrated song," a musical interlude or sing-a-long that took place between films or while the reel was being changed on the projector. During this portion of the program the singer would stand to the side of the stage and sing while a series of pictures illustrating the text of the song, or the sheet music itself, was projected on the screen.22 Sometimes the song was coordinated thematically with one of the films in the program, but most often the illustrated song simply promoted the sale of a new popular song. It was one of the earliest examples of using film to market a product, in this case song recordings and sheet music for home consumption.23

Film production companies, of course, were interested in the standardization of the musical part of films as well. By 1910, most film companies were providing written synopses of new films as part of their rental service, and as early as 1909 Edison's film company took this practice one step further, distributing "musical suggestion sheets" in its bimonthly magazine, the Edison Kinetogram, under the title "Incidental Music for Edison Pictures" (Document 2). While one column condensed the plot into numbered episodes, a corresponding column suggested an appropriate type or tempo of music to play during each episode. These early cue sheets allow us a glimpse of the specific repertoires that pianists might have played. Although the description is often general, calling simply for waltzes, marches, or "popular airs" without identifying specific compositions, titles are occasionally given as well. The early cinema was clearly full of the popular music of the day-Tin Pan Alley songs, folk songs, ballads, rags, and Sousa marches.24

These early cue sheets also describe a "compilation" approach to underscoring the film. After breaking a film down into individual scenes, the pianist would translate the action of each scene into a musical tempo, which was then used to specify an accompanying musical selection. Hurried actions on screen, for instance, were given fast music-jigs, allegros, marches; quieter actions like love scenes called for slower music-plaintive melodies, andantes and adagios. That each scene was distilled primarily to the tempo or rhythm of the action, with the moods or emotions of the characters performing those actions only a secondary consideration, is significant. Although the earliest film music was certainly used to enhance the screen narrative, it seems primarily to have been a rhythmic feature, emphasizing both the pace of the action on the screen and the structure of the film as a whole.

Theater musicians also took part in these early efforts to standardize the live, musical part of the film. Their participation in the discussion came primarily by way of two new critical venues. The first was the "music advice column." Film industry magazines and journals had already been encouraging some discussion of music, as exemplified by Richardson's essay. Around 1910, that effort intensified, becoming more specifically musical. That year, for instance, in Film Index, one of the main weeklies that serviced the motion picture industry, Clyde Martin launched a column entitled "Playing the Pictures." A few months later, Moving Picture World introduced a similar column called "Music for the Pictures," edited by Charles E. Sinn, the seasoned musical director for Chicago's Orpheum Theater, and Martin added an additional column, "Working the Sound Effects." Each week, in response to readers' letters, Martin and Sinn discussed the problems musicians encountered with specific films. Both also usually ended their columns with suggestions or cue sheets for newly released films (Document 5). In 1912 Ernst Luz, the musical director for the Loew's Theater chain, launched a column in Moving Picture News that, bearing various titles from "Music and the Picture" to "Musical Plots," routinely featured cue sheets but no dialogue with practicing musicians. By the mid-1910s, most of the industry serials, as well as major music journals like Metronome, the Musical Courier, and the American Organist, had weekly or monthly advice columns for the moving picture musician. Metronome, a journal devoted to the practical musical occupations like wind and dance bands, was one of the first musical publications devoted specifically to music to take notice of film music, as evidenced, for instance, by Frank Edson's long-running column "The Movies," launched in 1916. The addition of regular columns on film in more prestigious publications such as Musical America and the New York Dramatic Mirror, both of which surfaced around 1917, was a clear sign that film was beginning to be viewed as a serious art form.25

In the early 1910s, one of the most frequently discussed advice column topics was the execution of sound effects. In a Motion Picture World column from 1911 entitled "With Accompanying Noises," a guest writer, musician Emmett Campbell Hall, reveals not only how complex the musicalization of sound had become but also how troublesome.26 Part of the problem, Hall acknowledges, stemmed from inattentive musicians. A missed or poorly timed sound could have an unintended comic effect on the appreciation of a film. Other sounds were discovered to be mechanically unreproducible (a fusillade of cannon or gunfire, for instance). A subtler part of the problem, however, lay in the conceptual equality of sounds, with no one sound, whether foreground or background, having priority over another. Likewise, there was no hierarchical distinction between sound and music. In a film with a pastoral setting, for instance, musicians might give equal priority to imitating the sound of the sheep seen in the background as to musically evoking the broken heart of the shepherdess in the foreground. Hall doesn't provide any prescriptive solution, other than to exhort musicians to use caution and taste ("cut out the sound effect or use it with brains" are his final words); nevertheless, his remarks describe the general confusion of sound and music that dominated the early cinema.

Similar observations fuel Louis Reeves Harrison's sublime satire "Jackass Music" (Document 3), published in Moving Picture World in 1911. Harrison's humorous distillation of the inattentive and unthinking drummer, the fictitious Percy Peashaker, is another reminder of the flat quality of early live sound, and of the importance of properly executed sound effects to a good musical accompaniment. As the bad example of Percy reveals, drummers were making little or no distinction between distant sounds and sounds that were at the center of the film's action. Both were typically executed at the same dynamic level, as if they existed in the same spatial dimension.27

Film sound was not the only aspect of the musical accompaniment in need of reform. Harrison's two other "Jackass Music" caricatures, Lily Limpwrist and Freddy Fuzzlehead, address the equally significant problem of repertoire. In both examples, we see a new criterion for selecting or improvising the music for individual scenes taking shape. Whereas before, the tempo of the screen action was the prime musical motivator, here Harrison points to another pressing consideration: the mood or emotion of the characters on the screen. Jackson couches his suggestion in the satire of the emotionally challenged Lily Limpwrist, a pianist oblivious to the appropriate tempo and mood of death, and in the example of Freddy Fuzzlehead, a "funner" pianist who often intentionally counterpoints scenes with unsuitable music or songs featuring contradictory texts.28 Behind Harrison's artful satire, however, is an innovation of significant proportions: the idea that not just the tempo, but also the emotion and atmosphere of the scene, should drive musical selection.

Not all issues could be discussed adequately in the space of the music advice columns, of course. Some important topics, like the question of appropriate repertoire for the pianist, were dealt with in another new critical venue, the playing manual, a sort of advice column writ large. The earliest playing manuals, in fact, texts like W. Tyacke George's Playing to Pictures (1912), George Ahern's What and How to Play for Pictures (1913), and Lyle True's How and What to Play for Motion Pictures (1914), surfaced around the same time as the columns and featured a similar mix of methodology and analysis of individual films.29 In general, though, they provided more theory and aesthetics than the advice column, no doubt because they had more space.

As a frequent column contributor to Moving Picture News, George Ahern was no doubt aware of cosmopolitan theater practices. But considering that his manual was published in the decidedly unmetropolitan town of Twin Falls, Idaho, his observations about repertoire or "appropriate music" become even more noteworthy. Like Harrison, Ahern urges a closer reading by theater pianists of both the emotions and actions of the characters on the screen. Under the chapter heading "Appropriate Music" (Document 4), he also advocates a subtle hierarchy in the selection process. In choosing or improvising music, he says, tempo and sound effects should yield to the more important question of the mood or atmosphere of the scene. Moreover, every mood or event should be recognized musically. This more detailed approach, which involved not only more nuanced musical interpretation but also attention to a greater number of filmic variables, in fact pointed to a minor revolution of sorts in motion picture accompaniment. Musicians were no longer just "accompanying" the film; they were now "illustrating the picture" or "playing the moods," to use two catchphrases that began to surface with greater frequency in the early 1910s. Ahern's suggestions are all in the service of building not just a theater pianist but a new breed of motion picture musician, the musical illustrator.

As for the repertoire of the early cinema, Ahern gives an idea, if only theoretically, of what pianists most likely played. Most film accompaniment, he admits, consisted of a continuous parade of popular "hits." Although he doesn't define what he means by "hits," the sample cue sheets he includes in the manual indicate that he means Tin Pan Alley tunes, Broadway songs, popular dance hall numbers, rags, and folk songs. In 1913, in other words, popular music still dominated the cinema's musical soundscape.

Not all films, however, were treated with "hits," Ahern continues. "Really good pictures," he says, are given "better" music. This category of repertoire was defined not by Mozart and Beethoven, as one might expect, but rather by lesser-known composers of operetta and light classical fare.30 The works of the classical masters, in fact, appear to have been deliberately avoided. Most contemporary audiences, Ahern warns, are not familiar with Mozart, Beethoven, and Chopin and would not want to hear such "high-brow" selections. "People don't go to a picture show to hear a concert," he concludes, "but to see the pictures accompanied by good music." Not all audiences would have agreed with Ahern, of course, especially his murky category of "better" music. Certainly some classical music was being played, though it is true that even in 1913 the practice was not widespread.31 Given that the cinema was still attended primarily by lower- or working-class audiences that were much more familiar with popular than classical music, Ahern's assessment about the nature of the music heard in movie theaters makes sense.32

In addition to the advice column and the playing manual, a third venue catering to theater musicians emerged in the early 1910s as well. While not critical or theoretical, film music "repertoire collections" and "encyclopedias" were nonetheless an important resource. These books brought together musical scores, classified and organized by mood, tempo, or geography for easy manipulation during improvisation. Gregg A. Frelinger's Motion Picture Piano Music: Descriptive Music to Fit the Action, Character, or Scene of Moving Pictures (1909), The Emerson Moving Picture Music Folio (1910), the Orpheum Collection of Moving Picture Music (1910), F. B. Haviland's Moving Picture Pianist's Album (1911), the Carl Fischer Moving Picture Folio (1912), and Gordon's Moving Picture Selections (1914) were some of the earliest such collections, along with the multivolume Sam Fox Moving Picture Music (1913), in which the editor, J. S. Zamecnik, arranged the musical selections according to film topics or musical styles.33 While most of the repertoire featured in these collections was newly composed, later repertoire collections like Erno Rapee's Motion Picture Moods for Pianists and Organists (1924) also contained a good deal of preexisting music by well-known classical composers. These film music collections, helped musicians organize their material by mood and tempo, but they were also useful for film accompanists who wished to acquire repertoire and build their own personal libraries.

The encyclopedias in particular contained little or no discussion of selection practices, but their contents raise an important aesthetic consideration. Most relied heavily on preexisting music. Certainly, new music was being improvised and written specifically for the pictures. Zamecnik, for instance, a composer and theater pianist in Cleveland who originally studied with Dvořák, wrote a number of original pieces, or "moods," for the Sam Fox anthology.34 However, an equal amount of music previously written for entirely different situations was also being recontextualized by film accompanists. Film may have relied on music's preassociations to make itself understood, but it was also, from the very beginning, repurposing existing music and giving it new associations.

By the early 1910s, cinema audiences were being treated to a fairly consistent film music experience, one that integrated the sound of live piano and percussion with visual imagery to closely "illustrating" the actions and emotions of the characters on the screen. In the next decade, this illustrative approach would prove formative in several respects. Popular music and light classics, for instance, continued to be an important part of the musician's repertoire, for comedies and newsreels especially. The practice of using music to highlight the mood on the screen remained dominant as well. However, several of the illustrator's practices were rejected by the next generation of film accompanists, and new, significant changes were proposed, especially in terms instrumentation. This next generation also distanced themselves from the illustrators by expanding the repertoire of the cinema and by labeling their new music the "better" music.

In many respects, the reforms that began to reshape film accompaniment during the 1910s were triggered, as they had been before, by changes that film itself was undergoing. The average length of a narrative film was increasing, for one thing. Before 1910, theaters typically screened a half-hour program of one-reelers, short films that were differentiated (in both production and distribution) by general typology: comedies, dramas, newsreels (weeklies), travelogues, melodramas, slapsticks, and so forth. Between 1910 and 1915, however, the "feature" film-typically meaning a drama from two to eight reels long-began to surface with increasing regularity.35 Initially, these films were exhibited like touring "roadshow" films-that is, they were given special, stand-alone screenings with an admission cost of between ten and twenty cents instead of a nickel-but by the mid-1910s they were being incorporated into film programs. This exhibition format, with the longer film "featured" among the shorts, provided a new incentive for audiences to patronize the movie theater.

Feature films were distinctive not only because they were longer but also because they were more narratively complex. Many offered treatments of dramatic masterworks from literature or important historical events. Some imported well-known theater actors and actresses from the stage. Sarah Bernhardt, for instance, one of the most famous dramatic actresses of the early twentieth century, starred in several dramatic silent films of the early 1910s. Her appearances, together with the new multireel length of the feature film, gave the film drama in particular a new sense of maturity and artistry.36

The seriousness that attended the production of these new feature films also carried over into their exhibition, including their musical accompaniment. In a 1911 "Music for the Pictures" column, for instance, Charles Sinn argues that a similar "elevation" or "improvement" should mark the music as well. The only music suitable for a great actress in a classical drama, in his view, was that of the concert hall and the opera house, the great classical masters such as Wagner, Mendelssohn, Puccini, Tchaikovsky, and Grieg. This new repertoire was not to be used for all the films on the theater program, however. Comedies, slapsticks, and newsreels, for instance, still required the use of popular music. Feature films, according to Sinn, were special not only because they presented serious drama, but also because they should rely on classical concert-hall and operatic literature for the accompanying music.37

In another column from 1911 (Document 5), Sinn agitates not only for the more dedicated inclusion of classical repertoire, but also for a radically different organizational method. Drawing on the compositional techniques of opera composer Richard Wagner, Sinn proposes that the dramatic musical accompaniment be constructed around the concept of the leitmotif. In his operas, Wagner associated certain characters, ideas, and events on the stage with specific themes or melodies in the orchestra, so that each reiteration of the theme or melody would add to the meaning of the drama by recalling a previous visual context. Using such a technique with motion pictures, Sinn notes, would allow the musical accompaniment to parallel a film's larger narrative structure instead of just illustrating the moods of each scene. Yet, Sinn admits, such a technique would also be virtually impossible to execute without advance screenings of the film, a luxury rarely available to musicians in 1911.38 Although at this point it is only a suggestion, the idea of improving dramatic film music with compositional techniques that emphasize thematic unity will shortly prove significant.39

While Sinn may have been one of the first to advocate "classicizing" film music-importing classical music and the compositional practices of classical composers into film accompaniment-one of the first composers to successfully execute these ideas was Joseph Carl Breil. Breil began his career writing for the theater and later wrote several operas. But by 1914, when D. W. Griffith approached him in 1914 to provide the film accompaniment for his new Civil War epic, Breil had also established a reputation as a film musician on two high-profile European films.40 His third score, for Griffith's Birth of a Nation (1915), was a high-water mark both for him as a composer and for film accompaniment in general. Many of the score's distinctive features were dictated by innovations D. W. Griffith was making to the film proper. At over three hours in length, the twelve-reel Birth of a Nation was the longest American film yet made. Breil's accompaniment, which also clocked in at over three hours, was thus equally singular in terms of length. Birth of a Nation was also one of the first blockbuster films, playing in many metropolitan theaters for well over a year. As a result, Breil's accompaniment became one of the best-known and most widely circulated orchestral film scores in early film music history.41

Although he employed traditional compilation techniques, supporting individual scenes with well-known patriotic and Civil War tunes, for instance, Breil departed from the illustrator's approach in two significant ways. First and most noticeably, his accompaniment drew heavily from the great symphonic and operatic literature. In his compilation, Breil quotes the music of Grieg, Weber, Beethoven, Tchaikovsky, and Wagner, among others.42 He gave Wagner's famous "Ride of the Valkyries," in fact, not only a new military context but also, considering how widely the film circulated, greater exposure than it had received in American concert halls and opera houses to that date.43 In this regard, the Birth of a Nation accompaniment was an audible challenge to the traditional nickelodeon practice of using popular and light operetta repertoire to illustrate filmic action. It showed that special-event films were special to some degree because of the serious classical masterworks used in their musical accompaniment.

Breil also departed from the old approach by importing the compositional technique of the leitmotif into his score. Although a majority of the score consisted of preexisting music, in the parts Breil composed himself he used Wagnerian-type leitmotifs to refer to individual characters and events.44 As Breil put it, addressing the specific needs of the dramatic film required awaking to the possibilities of a musical accompaniment that, like the new drama itself, emphasized "uniformity of design and construction."45 While Breil's score fell short of this goal because of its patchwork use of preexisting music, his attempt represented an innovative emphasis on thematic unity, as opposed to episodic structure, in film accompaniment.

The lofty intentions and inspirations for this new classical and thematic approach were also reflected in the terminology Breil used. Even after acknowledging that he had composed less than half of the music for Birth of a Nation himself, Breil preferred to call his specially prepared accompaniments "musical scores." The term itself signaled a shift in the aesthetic sensibilities of film accompaniment. Although Breil may not have been the first to import classical sounds and techniques into film music, even in its day the Birth of a Nation score was seen to epitomize these improvements.46

This "classicization" project could not have been completed, however, without the help of exhibitors. Musicians like Breil could import classical symphonic repertoire into their "specially prepared scores," but without a specially prepared orchestra to execute them in the theater, the improvement would not have been heard. Although having an orchestra accompany the pictures was by no means unheard of in the mid-1910s, it was also not standard.47 In most cities, the pianist still held sway, although the organ, too, was beginning to find its way into movie houses. With the inclusion of more classical repertoire and techniques, however, came the move to make the orchestra the main purveyor of cinema music. As an interview in Moving Picture World entitled "The Art of the Exhibition" (Document 6) suggests, this crusade was the territory of one exhibitor in particular, S. L. Rothapfel. Rothapfel was manager of the Strand Theater in New York City, one of the first movie theaters to employ a full-time orchestra to accompany the pictures instead of a pianist.48

The new dramatic epic not only required more serious, classical music, but it required the classical symphonic orchestra to play that music. Although Rothapfel was not the first to expand the instrumentation of film music beyond pianos and drummers, previous "orchestras" were chamber sized, consisting, for the most part, of two to twelve instruments.49 Rothapfel was one of the first metropolitan movie palace managers to assemble a concert-sized, forty-piece orchestra to play for films. This instrumentation allowed movie theaters to feature specially prepared scores and more of the classical symphonic or operatic literature than before. Rothapfel's mission to "symphonicize" the film experience, to change the standard instrumentation of film music from piano to orchestra, addressed the practical aspects of the "better music" or "classical music" project. His mission to make theater music orchestral even led to a structural innovation, the symphonic prelude or orchestral overture, used to introduce the feature film.

This push to reform and elevate film accompaniment included efforts to provide films with an entirely original, often symphonic score. Much as serious actors and actresses were imported from the theater into film, directors and exhibitors sought to lend an air of sophistication to the musical accompaniment by importing well-known classically trained composers. In 1908, the great French composer Camille Saint-Saëns composed an original score for the film L'assasinat du duc de Guise, produced by an esteemed group of French theater actors. The film itself proved unsuccessful, however, so Saint-Saëns's efforts went unnoticed. Over the next several years, the practice of composing entirely original scores was pursued only intermittently, and then primarily with big-budget historical or religious films. In 1912-13, for instance, productions of Cleopatra, The Life of John Bunyan, The Prisoner of Zenda, Hiawatha, and Quo Vadis were all advertised as having original special scores. The 1913-14 season included more historical epics, but only a few of these "special scores," by composer Manuel Klein, music director at the New York Hippodrome, and one by George Colburn, are known to survive.50 Between 1911 and 1913, the Kalem film company also tried distributing some of its dramatic films with original scores by composer Walter Cleveland Simon, but, due to time constraints and the variations in individual theater orchestras, the effort was soon abandoned.51

In 1916, the highly regarded operetta composer Victor Herbert attempted to write an original score for The Fall of a Nation. Music expressly written for a specific picture, Herbert argued, would solve the problems of the "patchwork character" of current film accompaniment. "When the orchestra plays bits of 'Faust,' or 'Tannhäuser,' or 'Carmen' or 'Traviata,' the hearing of the music flashes pictures from those operas on the minds of the spectators, and attention is distracted from the characters in the [film] story."52 Preexisting popular and classical music came loaded with extramusical associations, and while some compilers were skillful at manipulating those associations to good effect, some types of music-those with texts in particular, such as songs and arias-were inherently problematic. New or original music, Herbert asserted, would circumvent these problems. In the end, Herbert, hampered by a constantly changing script and last-minute editing, did not complete the score, but his experiment laid important groundwork for further attempts at original scoring in the late 1920s, the golden age of silent film.53

While the compilation method remained the most practical solution for accompaniment in the teens, not all voices in this period sang its praises. Theater organist Blanche Greenland, for one, approached the practice with some skepticism. In a 1916 article titled "Faking in Movie Music Corrupting Public Taste," Greenland describes how the new and admirable practice of using "better" or classical music was being compromised by the widespread practice of "faking." "Do we realize," she asks, "that a juggernaut is bearing down upon the public of our moving picture theaters? Emblazoned over the front is its name, 'Faking.'" Faking, she continues, "is the deliberate mutilation of harmony by a performer with the intention of deceiving the ear of the listener. For instance, in such well-worn melodies as Rubinstein's 'Melody in F,' Mendelssohn's 'Spring Song,' and Dvořák's 'Humoresque,' horrid modulations are substituted, absurd inventions inserted, wrong chords introduced, producing something entirely wrong, which leaves its effect on the listener.... It is a vicious mingling of wrong combinations of notes perpetrated by an unlearned performer."54 Greenland's concern was not that the use of classical music would fail to elevate film, but that the butchering of great music by the ubiquitous and unskilled small-town musician was cheapening classical music. Instead of edifying public musical taste, the movies were corrupting it with sloppy, simplified arrangements of concert-hall classics.

Greenland was not alone in her assessment of film's "better music" project. Concerns about the use of the classical repertoire surfaced frequently in the critical literature of the time. Some doubted that any films, dramatic or otherwise, were worthy of Bach, Beethoven, and Chopin. "There have been relatively few subjects shown on the screen which call for classical accompaniment," pronounced the author of one playing manual, the Stolley-McGill Ten Lesson Course in Moving Picture Playing (1916). Some saw the coming of movies as representing the destruction of serious, concert-hall literature.

Ultimately, what these repertoire concerns really reveal is how idiosyncratic film music accompaniment was in the mid-1910s. The use of the classical repertoire and its close synchronization to film was still only a wish and not a fact. Because Breil's, Griffith's, Herbert's, and Sinn's innovations continued to share space with the less sophisticated practices of small-town illustrators and funners, their suggestions would not be fully realized until the 1920s and the golden age of silent film music.

If the 1910s was a transition period that brought a new aesthetic to film accompaniment, the decade that followed saw those innovations made general practice by a new generation of reformers. By the 1920s, the use of classical music and thematic compositional techniques and the replacement of the pianist by a classical orchestra were becoming quite standard. As Rick Altman has pointed out, as the orchestra came increasingly to substitute for the piano, film music began to focus more and more on published scores and classical repertoire.55 The intensification of this classical campaign is reflected in a quartet of publications by theater musicians from the mid-1920s (Documents 7-9 and 11) that document the successful transition from the improvised piano accompaniment of the nickelodeon era to specially prepared, concert hall-inspired compilation scores.

As before, developments in the parameters of the musical accompaniment were influenced by changes happening in film production and exhibition. Edith Lang and George West's Musical Accompaniment of Moving Pictures (1920) and Erno Rapee's Encyclopaedia of Music for Pictures (1925) show the film program, especially as it was practiced in large metropolitan movie palaces, to have developed significantly since the teens, particularly in terms of organization. Lang and West were Boston theater organists, and Rapee was one of the most prominent New York city movie palace conductors, working first at the Rivoli and Rialto Theaters, then, from 1920 to 1923, at the five-thousand-seat Capitol Theater with its 77-member orchestra.56 Together their texts reflect much of the range of exhibition practices that were being utilized at movie theaters around the country, with Lang and West focusing on the challenges facing the theater organist, while Rapee describes the variety of duties assumed by the music director or conductor of a large movie palace orchestra.

As part II of Lang and West's Musical Accompaniment of Moving Pictures, "Musical Interpretation" (Document 7), outlines, by 1920 an evening's program routinely consisted of a variety of short films-a comedy, a scenic, a newsreel, a documentary or educational film, and a longer feature film, typically a drama. Each type of film had a different structure-narrative versus nonnarrative, say-requiring different organizational approaches, such as thematic versus nonthematic. Each film genre, as Lang and West point out, also had different repertoire requirements. Newsreels and comedies, for instance, typically allowed inclusion of popular music or texted music. Scenics or travelogues, and newsreels to a certain extent, often relied on ethnically coded music, repertoire that consumed a large portion of the published anthologies of film music. Classical selections and newly composed music were reserved for dramas or feature films.

As Lang and West's manual reveals, anticipating the needs not just of individual films but of entire film programs was an essential aspect of film accompaniment.57 The musical needs of the film program as a whole steer Rapee's Encyclopaedia of Music for Pictures as well, which devotes complete chapters to the scenic (chapter 3), the newsreel (chapter 4), and the comedy (chapter 7), for example. From the early 1910s, these shorter films were joined in the film program by the dramatic film, which was typically the longest entry and the most substantial in terms of content and budget. As the dramatic film evolved into the feature film, it demanded more separation from the rest of the film program. According to the discussion of the feature film in the section of Lang and West's manual entitled "Musical Interpretation," typically, if not uniformly, the feature film was set apart by means of a special thematic treatment, resulting in a leitmotif compilation score that required some level of rehearsal. But as Rapee points out in his chapter on the feature film (Document 9), another means of providing emphasis, especially in the deluxe movie palaces that maintained in-house orchestras, singers, and even dance companies, was through the musical prologue.

In his chapter entitled "Vocal or Dance Artists" (Document 9), Rapee further describes how a musical performance can prepare the audience for the feature film. Many theaters, he observes, have adopted the practice of employing "vocal or dance artists to build prologues to the feature film." In building these prologues, which are ordinarily inserted between the newsreel or scenic and the featured dramatic film, Rapee advises: "The question as to whether a vocal or dance prologue should be used for any particular picture depends largely upon the atmosphere and the main characters of your picture."

When using music to set off, or "feature," the long dramatic film at the center of the program, Rapee notes, lighting and sets can be employed as enhancement-for example, to create special effects such as "rain, snow, floods, waterfalls, clouds, moon," effects that are central to the atmosphere or narrative of the feature film. The fact that this essentially nonmusical task fell to the music director or conductor, which explains why Rapee included it in his manual on film music, is testimony to the importance of live musical performance in the film program by the early 1920s. Indeed, people now went to the movie theater not just for films but to hear live music as well. Most of the film listings in the weekly trade journals described not just which films were playing at which theaters in major cities around the country, but also what prelude, postlude, and interstitial music was being offered in between the films. In most metropolitan cities, in fact, the cinema was second only to the symphony and the opera in terms of articulating significant musical repertoire.58

Like the special-event films and historical and religious dramas of the teens, the feature film, Rapee states plainly in his manual, needed "better" music. Contradicting Ahern, who twelve years earlier was sure his patrons would not tolerate concert-hall fare, even for dramatic films, Rapee, in a chapter entitled "The Missionary of Good Music and the Motion Picture Theatre" (Document 9), asserts exactly the opposite. The works of the classical masters should form the mainstay of the film accompanist's repertoire. Rapee takes this idea a step further, speaking not just of the importance of including classical repertoire but of a "movement for better music" in film accompaniment in general.

Opera had long been a part of silent film, but with the growing standardization of the orchestra as the primary musical instrument of the movie theater came greater use of the symphonic repertoire. By the 1920s, the orchestral overtures, symphonies, and symphonic works of Beethoven, Tchaikovsky, Schubert, Schumann, Dvořák, Grieg, Liszt, Rimsky-Korsakov, Mendelssohn, MacDowell, Verdi, Wagner, Gounod, Massenet, Puccini, Offenbach, and von Suppé, among many others, were routinely heard not just in program overtures and interludes but within film accompaniments themselves. Rapee himself loudly trumpeted the first performances of Richard Strauss's complex orchestral tone poems Till Eulenspiegel, in 1921, and Ein Heldenleben, in 1922, when they were presented as "overtures" on the movie theater musical program. Just as significant, but garnering less attention, was Rapee's use of selections from Strauss and Debussy tone poems as well as modernist fare from Stravinsky and Schoenberg in the compilation score for the great German expressionist film The Cabinet of Dr. Caligari at the Capitol Theater in 1921.59

In their manual, Lang and West describe a shift in the primary repertoire of the film musician, especially for the feature film. Audiences now, they note in their chapter entitled "Mental Alertness," are "capable of much more education and cultivation than they are generally given credit for," a process that includes classical music, a new and "inexhaustible treasure trove for all who seek diligently and patiently."60 In his 1921 manual Musical Presentation of Motion Pictures (the longest and least technical of the three discussed here), George Beynon, a well-known accompanist and compiler, puts it another way, saying that classical music should be used to counteract the cinema's long and close relationship with popular or, as he calls it, "cheap" music. In 1915, Beynon had been part of an early experiment on the part of movie producers and music publishers to standardize film accompaniment by providing complete specially prepared scores through the film exchange service.61 His manual was a further effort, not just toward standardization, but also toward improvement of musical accompaniment.

As all three of these manuals reveal, unlike the nickelodeons that a decade earlier were featuring primarily light and popular selections, the picture palaces of the 1920s were serving up a steady diet of operatic and symphonic literature, especially for the long dramatic features. Popular music-Tin Pan Alley, Broadway songs, early jazz, ragtime, and folk music-was still heard in the movie theater, but it had become almost exclusively the domain of the shorter, non-featured parts of the program, the comedies, scenics, and newsreels, and so received less air time than classical music. Beynon, in his chapter entitled "Proper Presentation of Pictures: Songs as Themes" (Document 8), in fact cautions against excessive use of popular or texted music, while noting the value of the original song-theme often used to accompany the main hero or heroine of the feature.

In addition to changing the standard repertoire of the compilation score, the general move to "classicize" and reform film accompaniment was affecting the structure of the specially prepared score. Film accompanists were still choosing their musical selections based on the emotions and actions of the characters on the screen, but now they were relying on specific musical themes or melodies to convey key information. "The kernel of the musical illustration of a picture is the main theme," Lang and West state in their manual. "This should be typical in mood or character of the hero or heroine.... [It] should be announced at the introduction ... [and it should] be emphasized a at the first appearance of the person with whom it is linked."62 Beynon devotes three chapters of his manual to thematization of the score. In "Proper Presentation of Pictures: The Theme and How to Use it" (Document 8), Beynon not only discusses ways of implementing a thematic approach, but he also acknowledges the aesthetic implications of doing so. A thematic approach to film accompaniment does more than underscore mood and tempo, he asserts; it reinforces narrative unity and structure as well. A repeated melody allows the images on the screen to be interpreted by recalling previous imagery.

The systematic deployment of musical themes in film accompaniment was not a new idea. As early as 1911, Clarence Sinn, for one, had encouraged theater musicians to use Wagnerian leitmotifs to create a sense of film unity (see Document 5). By the 1920s, what he could only theorize about was being executed on a regular basis. While neither Beynon nor Lang and West mention Wagner in this regard, they do describe the deployment of themes as part of a larger improvement project. In a section of their manual entitled "Thematic Development," Lang and West note that thematically coherent scores are the result of film accompaniment having imported the models of classical "symphonies and sonatas." For Beynon, the ramifications of this new technique are even bigger. In his view, the "proper presentation of pictures" involved not just conveying individual moods or tempos through compiled snatches of music; rather, it required a single composition that brought together both thematic and "incidental" music to create a structured viewing experience.

Executing the new thematic score was possible only with better synchronization techniques and tools, which the playing manual authors make frequent mention of, Beynon especially. In the initial decades of the silent film, musicians had addressed synchronization informally, usually through advice columns in trade publications. Already in the early teens, "musical suggestions"-alternately called "plots," "synopses," or "musical programs"-for selected weekly films were commonly included at the end of these columns, tied to specific tempo markings. Starting in the mid to late teens, the "music" columns in film journals like Moving Picture World and Motion Picture News, and the "photoplaying" or "movies" columns in music journals like the American Organist and Metronome, began to include not just summaries of titles and actions with corresponding repertoire selections, but also stopwatch timings, down to the quarter-minute. In the mid-teens, "stand-alone cue sheets" issued by studios as part of their film exchange service also began to surface. Although the musician Max Winkler did not invent the cue sheet as he claimed, the cue sheets he began producing for Universal in 1915 were some of the earliest produced independent of a trade journal column.63 Winkler, who was eventually contracted by Paramount and Bluebird studios to make cue sheets for their films, was quickly joined by Beynon, who prepared cue sheets for the films made by the Famous Players, Lasky, and Morosco studios, and S. M. Berg, who for worked for the Metro, Vitagraph, World, Selznick, and Triangle studios.

These studio-produced or film exchange cue sheets were significant, but because their repertoire selections were often guided by financial arrangements the compilers made with music publishing companies, it was not until the early 1920s and the advent of the commercial cue sheet service that cue sheets became a common part of film music accompaniment. Each service was launched with great fanfare and announcements of which celebrity conductors and composers had been secured to prepare the service's cue sheets. These included many of the most prominent New York movie palace music directors and conductors of the 1920s, Rapee, Rothapfel, and Hugo Riesenfeld among them, as well as popular advice columnists such as Ernst Luz, James Bradford, J. C. Zamencik, Edward Kilyeni, and Carli Elinor and veteran cue sheet makers Winkler, Berg, and Beynon.

The most successful of the independent commercial services was Cameo Music Service Corporation. Cameo had a distinctive patented style developed by M. J. Mintz, the general music editor of the service. Mintz's "Thematic Music Cue Sheet" format included not only stopwatch timings, but also a musical incipit (usually ten to twenty measures long) of the melody of each suggestion. Document 10 presents a good example of this popular style, in the first page of the cue sheet James Bradford prepared for Cameo for the 1924 hit film The Thief of Bagdad. This format also included special "Notes" to the performer that covered atypical structural requirements-instructions to repeat a section of the melody, for example, or to play a tune arhythmically to suggest "drunkenness," or prompting a specific sound effect. Cameo also issued entirely verbal cue sheets too, as shown in the second example in Document 10, the cue sheet for the 1926 film Dame Chance. Although this cue sheet accomplishes the same thing structurally as the first one, its non-incipit format made room for a concern that became increasingly important in the preparation of cue sheets: the use of taxable or nontaxable musical repertoire-the subject of Document 15.64

Despite the increased accuracy of cue sheets, there remained one significant impediment to true synchrony, something that was beyond the control of the musician. As Beynon observes in chapter 12 of his manual, "Synchrony" (Document 8), the science of timings was still plagued by variations both in the speed of the film and in the "load," or electrical current, being supplied to the projector. As long as those two features remained irregular, so did film accompaniment. While a solution to this problem eluded musicians and theorists alike, the fact that such technical matters were being discussed was tacit acknowledgment of how sophisticated the improvement project had become.

Synchronization was also a concern of Rapee's. Although he does not address the issue in terms of the electrical current running the projector, he does point out how variations in the film's projection speed can affect the structure of the prepared musical accompaniment. Achieving a consistent and predetermined feet-per-minute ratio on film projectors, he notes in the chapter of his manual titled "Projection" (Document 9), is crucial to being able to synchronize the musical accompaniment with the film.

While much of their effort was devoted to elevating film accompaniment through the importation of classical repertoire and compositional techniques, some of the manualists of the 1920s found room to address the difficult musical problem of film sound. Lang and West's discussion of the "unit" organ is noteworthy in this regard. In the 1920s, in addition to the standard sound of an orchestra, theaters began to feature "unit" organs, an electric keyboard instrument made specifically for motion picture accompaniment.65 As outlined in "Special Effects and How to Produce Them" (Document 7), part of the growing appeal of these enormous electronic organs, aside from their volume, was the wide range of special sound effects they could produce. Most unit organs had a variety of stops that could imitate everything from clock chimes, train whistles, and car horns to babies crying, dogs barking, and hurricanes raging. While traditional "traps" and sound effect devices were still being used, theaters big and small now favored the unit organ over the piano because it consolidated sound and music in a single instrument.

As the equipment for sound production became more sophisticated, however, the concept of sound itself saw an important refinement. Whereas the film "funners" and "jackass" illustrators of the previous generation had imitated all sounds equally, embedded in Lang and West's discussion of the unit organ is a tacit acknowledgment of foreground and background sound. Through the use of dynamics and selective silence, theater musicians were now adding a consideration of distance and perspective into their definition of sound. Still, a practical or even theoretical distinction between film sound and film music was not yet available. Like most of their contemporaries and predecessors, Lang and West, too, describe a fundamental conceptual equality between sound and music in the theater.

One final improvement to the film program, especially as it was practiced at the major metropolitan movie palaces, resulted directly from improvement and expansion of the theater orchestra. Part of what made a movie palace a "palace" or "deluxe" was that it maintained a full-time orchestra, in addition to employing a full-time house organist. Although, as Rapee acknowledges in his chapter titled "How to Organize and Rehearse an Orchestra" (Document 9), movie theaters around the country were still using a variety of instrumentations and ensembles, increasingly the New York City movie palace model, exemplified by the Capitol, Rivoli and Rialto theaters, with their orchestras of sixty pieces or more-differing from a "symphony" orchestra only in the size of the string section-was what most metropolitan theaters aspired to.66 Moreover, the movie theater's symphonic orchestra was being allowed to flex its muscle more, not just in terms of accompanying the feature film or preparing the feature film with a musical prologue, but in preparing the entire film program with a specifically musical number: the overture. As Rapee suggests again in his discussion of this device (Document 9), the movie theater by the mid-1920s was rivaling the symphony hall and the opera house as a venue for serious musical performances.67

As the first-generation reformers had foreseen, the "better music" or classicization effort was supported not just by changes in the sound, construction, and content of the specially prepared score, but also by developments in a number of exhibition and production practices. Although he never wrote a handbook or playing manual, Hugo Riesenfeld played a significant part in shaping the sound of the 1920s cinema. Born in Vienna and trained as a violinist, pianist, and composer at the Vienna Conservatory of Music, the famed music director brought the great German symphonic tradition with him to New York's best movie houses. Over the course of his career, Riesenfeld was the musical director of three of the most prestigious movie palaces-the Rialto, Rivoli, and Criterion theaters-and he was renowned for his specially prepared scores for feature films. 68 In many ways, Riesenfeld continued the discussion begun a decade earlier by S. F. Rothapfel, for he, too, was interested not only in creating complex thematic scores that drew primarily from the "better," classical music repertoire, but also in having those scores performed by large, sixty- to eighty-piece symphony orchestras that rivaled in skill the most prestigious concert-hall orchestras in the country.

By 1925, Riesenfeld and music directors like him at "deluxe" movie palaces in major metropolitan areas had access to significant exhibition resources. As his article "Music and Motion Pictures" (Document 11) outlines, the movie palaces had enormous music libraries, repertoire collections that were being continually refreshed with new music from an international array of publishers. Central to the accumulation of repertoire was the music librarian and a library staff who categorized new and existing music according to tempo, style, or form. If the maintenance of repertoire required a large staff, so did the compilation of each film score. Music directors at the deluxe palaces were no longer solitary compilers and composers, but coordinators of teams of supernumeraries-timers, copyists, arrangers, rehearsal conductors-who contributed to the score creation process. The division of labor that Riesenfeld describes, in fact, is remarkably similar to the Hollywood studio system that would emerge less than a decade later.

Riesenfeld also describes the improved synchronization practices that were making the complex, symphonic compilation scores possible. Where previously music directors had little time and opportunity to prescreen a film, Riesenfeld constructed scores based on multiple screenings of each film. The use of the stopwatch was still key in integrating live music with mechanized images, but what really allowed the specially prepared score to become commonplace in the 1920s was the ability of the music director to see a film several times in advance and to analyze it, stopping and replaying it at will. At Riesenfeld points out, the compilation formula had also progressed to a position where directors had several days to rehearse and refine their scores.

A few other matters of note in Riesenfeld's essay not only reflect the state of the mature compilation score, but describe the state of new music in the late 1920s as well. The author's passing reference to jazz-"for the time being jazz predominates in our film theatres"-reflects both the country's general obsession with the new style of popular music and the widespread view that jazz was not an artistic form of music. For Riesenfeld, jazz was not only of lesser quality, but also ephemeral. "It is only a matter of time," he observes, "before the wheel of public favor again turns, bringing the better type of music to the foreground again." He leaves little room for guessing at what "better" still means to compilers and accompanists, noting that "a jazz selection is old and discarded in a single season. A Beethoven overture or Chopin nocturne is eternally new."

Riesenfeld also describes the establishment of not just "better" repertoire, but better performers on the film theater's musical program. He and the other directors at the New York movie palaces now routinely engaged the services of renowned concert-hall soloists, such as pianists Jan Paderewski and Joseph Hoffman, violinists Sascha Jacobson and Efrem Zimbalist, and conductor-composers Percy Grainger and John Philip Sousa, to play during the interludes between films. The addition of esteemed concert-hall performers added another measure of seriousness and artistry to the pursuit of music for film. This improvement went hand in hand with the general increase in both skill and pay of the orchestral musicians and organists working in the movie theaters' pits. That the very existence of the pit musician might shortly be in peril is not hinted at in Riesenfeld's discussion. In fact, he perceives the new and experimental sound film, mentioned briefly near the end of the essay, not as a threat to live film music practices but instead as a possible solution for rural communities with less sophisticated musical programs and accompaniment. Although "it is not probable that the Vitaphone will ever entirely replace the orchestra," Riesenfeld concludes, it does make possible "the finest musical accompaniment ... where there is no orchestra available."

Something that remains unavailable in either the small-town or the deluxe cosmopolitan theater, even in the golden age of silent accompaniment, Riesenfeld notes, is the newly composed score, due to both compositional time constraints and silent film music's fundamental ephemerality. Composers in the late 1920s would "rather starve," he asserts, "with the hope of creating a great symphony that will live through the ages, than grow fat off the proceeds of an excellent but short-lived film score."

That is not to say that original scores were not being written. Several of the greatest films of this period did indeed have original scores, many by emerging concert-hall composers. European directors seemed particularly interested in testing the possibilities of collaborating with composers. In France, director Abel Gance commissioned several young composers to create scores for his films: Darius Milhaud wrote the music for La roue in 1923 and Arthur Honegger composed a score for his epic Napoléon in 1927.69 Established eccentric Erik Satie wrote an original score for the short avant-garde film Entr'Acte, directed by René Clair, which premiered as the interlude in Francis Picabia's 1924 ballet Relâche. The daring and "noisy" American composer George Antheil, working in Paris, wrote original music to accompany the 1924 French surrealist film Ballet mécanique, by Fernand Léger.70 Perhaps the most successful original score from the period was one that German composer Edmund Meisel produced for Soviet director Sergei Eisenstein's 1926 masterpiece Battleship Potemkin. Meisel's thematic, orchestral score was so dissonant and compelling in the last reel that it was banned in parts of Germany and England for fear that it, even more than the images of revolution in the film, would incite the audience to riot.71 A young Dmitri Shostakovich, who would go on to write more than forty film scores throughout his career, composed his first score in 1929 for the Russian silent film New Babylon.72

While filmmakers in Europe and Russia tentatively explored the possibilities of the original score, few U.S. directors made similar attempts. Other than actor Douglas Fairbanks, who entrusted several of his silent blockbusters-The Thief of Bagdad (1924) and The Black Pirate (1926), for instance-to the young American composer Mortimer Wilson, silent film music production in the United States focused primarily on the compilation not the original score.73 Several impediments still stood in the way of the original score becoming standard practice. A general lack of time was arguably the most problematic. Providing music for a three-hour epic was equivalent to writing an opera, yet film composers were usually given several weeks at most to complete their work. Often the director compounded the problem by editing the film literally moments before its exhibition, leaving the composer little or no time to make adjustments to the score. Thus Herbert's experiences and frustrations with The Fall of a Nation (1916) continued to be revisited-literally, in fact, in the case of Honegger's score for Gance's Napoléon a decade later.

Although the reform efforts to improve or "classicize" film music failed to produce consistent collaborations with concert-hall composers, they did succeed in standardizing the use of classical music in actual film production. Although accounts vary as to when the practice was started, music-particularly classical music-was frequently played on the set during film production.74 On-set music helped supply emotional inspiration for the actors as they filmed, and the public announcement of actors' on-set musical preferences, suggestive of their familiarity with the classical repertoire, also enhanced their reputations.75

The embrace of "better" music was also encouraged by copyright laws, which dogged the use of new music in the movie theater. Through the 1910s, composers, especially popular song composers, became increasingly aware that their music was being used in film accompaniments without their permission and without royalties being paid. Several composers attempted to address this problem, most visibly and ultimately most successfully in a legal action that took place in 1917. That year Victor Herbert served a lawsuit against theaters owners, arguing not only that original film music was worthy of copyright protection, but that royalties should be paid when any copyrighted music was performed in movie theaters. This suit had consequences even before it was settled. Beginning in the early 1920s, the entire body of working film music literature began to be separated into two categories: copyrighted (and taxable) and noncopyrighted (or free) music. Since the majority of the classical literature fell under the public domain and was as a result nontaxable and free to use, film accompanists and music directors acknowledged including classical music more frequently in their compilations.76

The greater inclusion of classical music in film music was no doubt also stimulated by another lawsuit, this one from 1924. As Document 12, a New York Times article titled "Publishers Win Movie Music Suit," recounts, songwriter Irving Berlin and nine others successfully sued theater owners for compensation for all the air time their popular tunes had been getting free of charge. Although enforcing copyright protection through bans and taxes was difficult, if not impossible, especially in small-town theaters, these court cases no doubt encouraged composers to use less disputed repertoire. In any case, the neat concurrence of this string of litigations and the rise in popularity of the classical repertoire in the movie theater is striking.77

The development of film between 1896 and 1926 was attended by great variety and experimentation. Within a short period of time, film had developed a broad range of genres and styles-from newsreels to narrative masterpieces, commercial advertisements to scenic travelogues, serious educational films to silly cartoons, low comedies to high drama, surreal montages to moralizing melodramas, magical tricks to unadorned realism, as well as a range of temporal lengths, from ninety-second actualités to fifteen-reel historical epics. The visual vocabulary and structure of silent film, from the experimentation of the nickelodeons in the 1910s to the standardization of the movie palaces in the 1920s, are well studied, primarily because it is the visual portion of these films that has survived intact. Film image-tracks can, with variations in projection speed aside, be viewed largely as they were originally intended to be. The same cannot be said of the sound that accompanied those images. The evolution of silent film sound and music is less well understood in large part because sound was an aspect of the film that was not mechanized. Music in the silent film, rather than being a fixed production element, was an ephemeral exhibition practice, and in many ways it was valued for its very lack of mechanization, for its ability to be living, flexible, and personal. Music and sound made film not only realistic but also artistic.

The documents in this section reinforce the centrality of this missing, or "silent," unmechanized part of early film. Just as directors experimented with and established visual standards for a range of film genres and lengths, so did musicians work to establish standards for the structure, synchronization, and sound of film music. They experimented with a variety of repertoire-the allusive properties of popular music and texted songs, and the less specific emotional qualities supplied by newly composed music. They explored instrumentation, anywhere from single upright pianos to eighty-piece orchestras. They tested the filmic effect of different tempos and meters on the perceived speed of moving images, and the way musical structure enhanced film structure and continuity. A stylistic change or abrupt musical silence could momentarily disrupt or emphasize the visual imagery; a single musical selection could have a suturing effect, smoothing over a visual cut or scene change. Similarly, a recurring musical theme could bring large-scale continuity to familiar characters in unfamiliar scenes. Silent film musicians even experimented with turning music into ambient sound and mechanical noise.

That the parameters for most of these elements of film sound were tested and standardized to a very high degree by the 1920s reveals just how formative this earliest period was to the history of film and film music. In fact, many of the techniques and practices pioneered by the musicians of the golden age of silent film music will be revisited by future composers, and some will continue virtually uninterrupted. Musicians like Max Steiner simply transferred many aspects of the mature compilation score-its thematic construction especially-to the sound film score. In many ways, Steiner's "wall-to-wall music" model of the early Hollywood score was a literal extension of the specially prepared orchestral scores of the silent period. Other silent music practices, however, like the rendering of sound effects, will be severely limited (to occasional mimicry or "mickey-mousing") and transferred to the newly created department of film sound. The revolution that ends the silent period will not just be the technical innovation of including dialogue; it will include as well the separation of film sound from film music, and of composers from sound engineers. This separation will also lead to a redefinition of the placement of music in film. Although the classic Hollywood score will be sonically and structurally familiar, its evolving status as an "underscore" defines an entirely new concept of filmic space, one that is not visualized. Therein lies one of the clearest indications that the musical world of the silent film has been left behind