4
Making Movies

Breaking into the Business
New York and Hollywood
Writing Pictures
The Producer System
How Much to Make a Picture?
Scenic Art
Behind the Camera
Film Color
Performance
Editing

Breaking into the Business

By 1915 the motion-picture industry had achieved a certain sophistication in the way it produced, marketed, and exhibited its pictures. Significant economic and industrial forces now acted to standardize these procedures, although filmmaking in the silent era could hardly be thought of as routine. What remained in a primitive and even chaotic state was the most elementary aspect of the filmmaking process: recruitment and training of new personnel. This is not to say that there was any shortage of activity in this area. Many volumes of self-instruction were published, schools abounded (some of them even legitimate), and magazine advertisements enticed men and women out to the studios. But the results of all this activity were negligible. Some actors and directors successfully came from the legitimate stage (many more were unsuccessful), and a few short-story writers and romantic novelists found new careers as photoplaywrights.

If memoirs and oral histories are to be believed, accident and bizarre coincidence seem to have been responsible for a disproportionate number of major careers. Allan Dwan, an electrical engineer, arrived to install Cooper-Hewitt lamps at the Chicago Essanay studio and stayed to write and direct. Raoul Walsh had most recently been working as a cowboy. Lois Wilson was a beauty-contest winner. Robert Flaherty was an explorer. The tremendous expansion of the industry within a single generation had created employment for thousands, but without a usable tradition there seemed no rational way to master the various professional skills involved. Relying on the theater as a model had limited application, and in the most technical areas, such as editing or cinematography, guideposts were nonexistent. Such fields soon came to be dominated by men and women who entered them just out of school and grew up with the medium, like Arthur Miller and Lee Garmes behind the camera, or Dorothy Arzner and William Hornbeck at the editing bench.¹

Organized schooling seems to have had almost no success in preparing young hopefuls for motion-picture work, despite a surprising amount of effort. Columbia University demonstrated a commitment to film studies throughout this period but without any notable acceptance from either the industry or the academy. In 1914 Columbia introduced film as a teaching aid in the Journalism School and extended its use the following year into economics, science, history, psychology, and English-literature

classes. Victor O. Freeburg (author of Disguise Plots in Elizabethan Drama) lectured on film at Columbia from the fall of 1915 to the spring of 1917 and eventually produced a pair of sophisticated texts, The Art of Photoplay Making (1918) and Pictorial Beauty on the Screen (1923). Frances Taylor Patterson began as "instructor of Photoplay Composition" in the summer of 1917 and continued into the talking era. While Freeburg essentially defined screen art in graphic terms, Patterson had a stronger interest in narrative, and her three books are almost completely concerned with the niceties of plot construction. The avowed purpose of her class was to help foster public appreciation of film and offer prospective screenwriters a "short cut to success."²

When America entered World War I, the U.S. Army Signal Corps established a training school at Columbia University to provide instruction in the making of medical and historical-record films. Among those who passed through this program, either as students or as instructors, were Josef von Sternberg, Alan Crosland, Ernest B. Schoedsack, Irvin Willat, and Lewis Milestone. These men, and others, later looked back on their Signal Corps service as a valuable shared experience, but what they might have gained directly from this schooling is unclear. Efforts to continue a course in film production after the war were defeated because Columbia refused to support it with proper equipment or facilities and because according to one faculty member, "the funds to pay competent instructors were inadequate."³

Other attempts to shape film as an object of university study in this period ranged from Vachel Lindsay's lectures at his alma mater, the University of Chicago, to a series conducted by Joseph P. Kennedy at Harvard's Graduate School of Business Administration. An increasingly sophisticated amateur film movement (especially after the introduction of the 16-mm Bell & Howell Filmo in 1923) did make use of many of these manuals and scenario models, but they seem to have had little effect on the industry as a whole.⁴

When established programs like the one at Columbia failed to satisfy the demand for film production courses, a rash of private technical schools arose, most derided by the industry as mere "sucker traps." Such instruction was aimed largely at four major crafts: acting, writing, cinematography, and projection. Directing, while always given lip service in the various "Breaking into the Movies" manuals, was so hard to conceptualize that little energy was devoted to promoting it as a way of making a living. For example, one 1921 manual limits its discussion of directing to this observation: "Next in line is the director, who takes the scenario and sets out to make the picture. There is a shortage of directors at present, and for that reason, salaries are particularly high in this line, but of course, direction is a profession which takes many years of study.…"⁵

Prospective actors and actresses appear to have been the most widely victimized of these hopefuls. Relatively harmless instructional manuals like Motion Picture Acting (1916) offered for study "191 posed photographs of motion picture stars, showing 499 different expressions and emotions." Readers of Mae Marsh's Screen Acting (1921) learned that circumstance, hard work, and native talent developed before the camera were the keys to screen success—a dictum broad enough to discourage no one.⁶

Publicity regarding the "effortless" rise of many movie favorites encouraged large numbers of men and women to see themselves as potential screen material. This route was perceived as far easier than stage stardom as well as more accessible to the young and impressionable (Lillian Gish, Mary Pickford, Norma Talmadge, and others became major film stars while still in their teens). Motion Picture reported various "fakes and frauds" directed at such young hopefuls. In one common "scam," magazine advertisements would solicit "directory entries" (for a fee) that would be circulated to the studios. Because films were not cast in this fashion, the use of such directories by unknowns was worthless.⁷

For a time, many film studios followed the pattern of popular literary magazines in accepting unsolicited story contributions. A torrent of books on the writing and selling of photoplays was patterned after similar volumes in more established fields. For example, A. Van Buren Powell's The Photoplay Synopsis (1919) was the latest in one series that already included Writing the Short Story, Writing for Vaudeville, The Technique of the Mystery Story, and others.⁸

Organizations like the Palmer Photoplay Corporation offered a "complete Course and Service in Photoplay Writing." For a fee of ninety dollars (less for cash), students received two textbooks (including the Photoplay Plot Encyclopedia, which discussed the thirty-six dramatic situations in terms of motion-picture style), three sample film scripts (including Lois Weber's For Husbands Only [1918]), a subscription to the Photoplaywright, the right to staff criticism of five submissions, and a series of booklets or "lectures" with titles like "Photoplay Elements of Situation Comedy" by Al Christie, or "The Necessity and Value of Theme in the Photoplay" by Jeanie MacPherson. Cecil B. DeMille, Thomas H. Ince, Lois Weber, and Rob Wagner served as Palmer's "advisory board."⁹

A rival operation, the Photoplaywrights' League of America, provided no written texts but offered to act as agent for its members in the sale of scripts. Criticism of manuscript submissions and legal and financial advice were also available. Sada Cowan and J. Grubb Alexander were among the "professional members" listed in its brochure.¹⁰

With the availability of good 35-mm motion-picture cameras on the open market, even the once-mysterious field of cinematography became accessible to ambitious amateurs. Ernest A. Dench advised "the ambitious young man or woman determined to break into the studios as a regular Motion Picture photographer" that many local opportunities already existed. He suggested establishing a service to film weddings, children, pets, and other domestic subjects, the production of advertising or industrial films, occasional work as a newsreel stringer, or even the establishment of one's own local newsreel. Creating an "amateur photoplay society … is no more expensive than is dabbling at ordinary photography on a moderate scale," Dench claimed. "Those with acting ability can figure in the cast, while the member possessing the most dramatic aptitude should be made the director. The talented weaver of stories would be the right man for scenario editor, provided he studied a book on photoplay writing and mastered the technique of photoplay construction. Last, but by no means least, you should do justice to the position of camera-man."¹¹

Camera skills might be acquired at the New York Institute of Photography, which offered a classroom program as well as a correspondence course. It promoted its own 35-mm cine camera, the Institute Standard, especially for such semiprofessional work, and issued various texts through its Falk Publishing Company.¹²

These were the glamorous areas of the motion-picture industry, but the ten thousand local projectionists (1920 figures) were also fair game for unscrupulous educators. Sink-or-swim training seems to have been the conventional route here, something that was not the case in cinematography or editing, where an apprenticeship system developed. It was not uncommon for the ill-trained graduates of "projection schools" to find themselves alone in a booth with no idea of how to put on the evening's show.¹³ Unfortunately, these were the men ultimately responsible for the quality of the show seen by the public, the weakest link in the chain apparently being the last.

New York and Hollywood

Southern California was clearly recognized as the major American production center by 1915, although the generic use of the term "Hollywood" to describe nearly all such activity had not yet developed. With studios scattered from Santa Monica to Edendale to Pasadena (and north to San Francisco and east to Phoenix), most commentators spoke of the area "in and about Los Angeles" as the "Mecca of the Motion

Picture." By 1924, however, a Hollywood mythos was clearly emerging. Perley Poore Sheehan, a popular screenwriter, issued a bizarre tract called Hollywood as a World Center, which combined elements of small-town boosterism, industry braggadocio, and occult transcendentalism (known locally as "new thought"). For Sheehan, "The rise of Hollywood and its parent city, Los Angeles, has world-wide significance. It is a new and striking development in the history of civilization.… This flooding of population to the Southwest has its origins in the dim past. It is the culmination of ages of preparatory struggle, physical, mental and spiritual. In brief, we are witnessing the last great migration of the Aryan race." Going beyond traditional American disdain for the eastern cities, Sheehan saw the birth of Hollywood as the dawn of the Aquarian age and described a New Jerusalem that would reveal to all mankind the "Universal Subconscious."¹⁴

In the absence of any traditional moral, intellectual, or religious framework, such philosophizing was quite popular in Hollywood in the boom years of the immediate postwar era, especially among the circle around Nazimova, Valentino, and June Mathis. Before it faded from fashion, one could see its traces in such curious films as the Mathis-Valentino Young Rajah (1922) or Sheehan's original story The Whispering chorus (1918) for Cecil B. DeMille. But its representation in specific films is far less important than the gauge it provides to help measure the emotional and intellectual distance between Hollywood and the traditional centers of American culture. The physical isolation of the place—five days by rail from the corporate home offices—very quickly inculcated a special "Hollywood" way of looking at life that generations of audiences would instantly recognize. European films tended to be produced in and around traditional centers of national culture—Paris, London, Vienna, Rome, Berlin—but this community of desert exiles made something very different of the American cinema. Of course, such films won the contempt of American critics for many years, but audiences worldwide (and some foreign critics, such as Louis Delluc) felt otherwise. Movies from Hollywood would become the first American cultural export to conquer the world.

Equating American film production in this period with Hollywood production is, however, somewhat misleading. That colony was a factory town, producing the motion-picture industry's major product, while executive operations, newsreel production, and even much of the animation industry remained in the East. Kevin Brownlow, in Hollywood: The Pioneers, gives a flavorful account of early West Coast units seeking and finding "sun, space, and somnolence." The fact that Los Angeles was renowned as "the nation's leading open-shop, nonunion city" also did not hurt, keeping wages to half what they might have been in the East.¹⁵

The concentration of feature production in Hollywood during the teens was widely noted at the time, especially because few could have predicted the rapidity of this migration. Although visiting companies had worked in Los Angeles since 1908, much of the outlying area was still largely a wilderness in 1915, lacking equipment houses, prop and costume shops, a steady supply of professional actors, or even basic sanitation and safety. Thomas H. Ince and a group from his studio were robbed "stage-coach style" by four masked gunmen on "the lonely road near Inceville " in March 1915. Three years later, Erich von Stroheim was still carrying a revolver under the seat whenever he and his fiancée drove the dangerous Cahuenga Pass to Universal City. In this light the stories told by Cecil B. DeMille, Allan Dwan, and others, regarding the need for sidearms to protect themselves from Patents Company thugs, might conceivably have a more mundane explanation.¹⁶

It was generally accepted that the West Coast studios were producing 125 reels of film per week in 1915, including everything from split-reel comedies to The Birth of a Nation. Knowledgeable estimates put this figure at between 62.5 and 75 percent of total American production.¹⁷ Despite the fact that the bulk of domestic production had already moved to the West Coast, it was only in 1914–1915 that the makeshift facilities used by most of the producers were replaced by permanent installations comparable to the eastern studios, most notably the American Studio in Santa Barbara and Carl Laemmle's Universal City.

The comparatively large segment of production remaining in the East maintained its position throughout the early feature years, when the proximity of Broadway plays and players caused firms such as Goldwyn and Famous Players to increase the output of their New York and Fort Lee studios. This situation survived until the winter of 1918–1919, when problems with coal rationing forced nearly all the companies then operating in the East to consolidate operations in their West Coast facilities. Paramount-Artcraft closed in Fort Lee but kept operating at their Fifty-sixth Street studio in Manhattan; Universal closed all their eastern studios except their Coytesville, New Jersey, operation; Goldwyn, Metro, and Fox moved everyone to the West Coast; Vitagraph kept only a small operation in the East. Only World Film remained, because "their coal is not only all purchased, but delivered."¹⁸

This forcible relocation was strongly resisted by many filmmakers who found California unsuitable as a production center, but not until after the Armistice could anything be done about it. As soon as wartime restrictions were lifted, a boom in studio construction swept the New York area (for various reasons this renaissance bypassed Fort Lee, which soon disappeared as a production center). D. W. Griffith, who had filmed Hollywood's most stupendous films, The Birth of a Nation and Intolerance, created a new studio for himself on the Flagler estate in Mamaroneck, just north of New York City. The East was where "the money and the brains" were, he said, as he proceeded to film Way Down East (1920), Orphans of the Storm (1922), America (1924), and other costly features in Mamaroneck.¹⁹

William Randolph Hearst transformed Sulzer's Harlem River Park and Casino into the vast Cosmopolitan studio, a beer-hall-to-movie-lot conversion that also occurred at the smaller Mirror studio in Glendale, Queens. At Cosmopolitan, Marion Davies starred in such lavish epics as When Knighthood Was in Flower (1922), Little Old New York (1923), and Janice Meredith (1924), an underrated Revolutionary War spectacle.²⁰

Vitagraph increased the pace of its New York operations, while Goldwyn and Metro reopened their eastern studios. Fox came back as well, opening a large new studio on West Fifty-fifth Street in Manhattan, where Pearl White and Allan Dwan worked from 1920. This facility continues in operation today as the Cameramart stages. The Talmadge sisters had their studio on East Forty-eighth Street, and various rental studios abounded, hosting such New York-based stars as Richard Barthelmess. By September 1920 the Exhibitors Trade Review headlined "Producers Say California Has Been Filmed Out—Are Looking for New Producing Centers." The paper reported that all the major companies were now making films in the New York area and that California's weather remained its sole compelling asset.²¹

That month the most important of the new eastern studios opened, the Famous Players-Lasky studio in Astoria. Over the next seven years, this studio would produce 127 silent features, serving as home base for those Paramount stars who preferred to work in the East, such as Gloria Swanson (nine films) or Bebe Daniels

(fourteen films). Such recruits from the stage as W. C. Fields and Louise Brooks began their "Hollywood" careers in Astoria, while Rudolph Valentino and D. W. Griffith saw the studio as a means of escaping the factory conditions on the West Coast. Astoria was at its most active in 1926, when the 26 features produced there constituted 40 percent of the entire Paramount program.

However, by 1922 Hollywood's share of American production stood at 84 percent, with 12 percent remaining in New York and 4 percent filming elsewhere. The East Coast had several large studios and ample rental space for the smallest companies. What it lacked was a significant group of middle-level producers, an area almost completely monopolized by Hollywood. New York boosterism reached a peak soon after, when Barrons predicted that "the motion picture business of the next decade will be mostly in sight of the tower of the Woolworth building"—a piece of advice that its readers would have done well to ignore. Such enthusiasm on the part of New York-based industry analysts reflected the positive feelings of those East Coast stars and executives whom they encountered while dining at the Colony or "21." It ignored the efforts of the larger and more powerful West Coast branches to centralize production under their own control, the increased costs of doing business in New York (Los Angeles was still relatively free of labor unions), and continuing problems with inclement weather. While the advantages of California sunlight were no longer crucial, problems with New York's rain and snow remained an issue, especially for a system that made heavy use of large standing sets. "Weather destroyed sets on the back lot time and again, ' wrote Jesse Lasky of the Astoria operation, which he temporarily closed in 1927. Other New York studios had already reduced operations, and by the close of the silent era, Hollywood was unchallenged as the center of American production.²³

But what of that 4 percent produced outside either Hollywood or New York? With 748 features released in 1922, this suggests that some 30 feature pictures (and a proportionate number of shorts) were filmed elsewhere. Even allowing for a handful of imports, the number is still significant. The bulk of these films were made by small local producers without access to national distribution. Many remained unseen. Some were sold on a states rights basis and never played in a key theater or won the attention of an urban reviewer. For historians, these regional productions are the terra incognita of the American film industry. Local historical societies treasure vague records of production in Ithaca, Providence, Ogden, and Augusta, for example, but so little information survives that most historians simply omit this activity altogether.²⁴

Writing Pictures

No matter what the location of the studio, the tremendous quantity of film generated each week required vast amounts of original (or at least semi-original) story material. The ad hoc practices of the earliest days of filmmaking had long been abandoned by most producers, although even as late as 1915 some directors could not resist a lucky opportunity. Henry Otto, directing for Flying A in Santa Barbara, noticed several hundred blackbirds sitting on telephone wires. He filmed them, then concocted a script in which the birds caused "wire trouble."

In general, though, it was well understood that regular release schedules demanded

a dependable flow of production and that scenario departments were needed to process scripts and synopses. A few writers, such as Ince's C. Gardner Sullivan or Thanhouser's Philip Lonergan, had steady positions generating large amounts of story material to order, but in 1915 the free-lance scenario market was still quite significant. Producers solicited manuscripts in much the same fashion as literary magazines. The New York Dramatic Mirror carried the "For Photoplay Authors, Real and Near" column, edited by William Lord Wright. Readers of the 3 February 1915 issue were told that the Universal editorial department was giving assurances that it would read all manuscripts, so long as they were typed, carried a synopsis, and came with a stamped, self-addressed envelope. Requests for submissions were often quite specific:

World Film Corporation, Fort Lee, New Jersey, is in the market for five-reel subjects running to not less than two hundred scenes. Stories must have original plots—not necessarily with what is known as "punch," but depicting a young innocent girl in country life. No costume plays considered, nor those dealing with crime or crooks. American stories preferable (William Lord Wright, "For Photoplay Authors, Real and Near," New York Dramatic Mirror, 3 February 1915, p. 30).

The Photoplay Author was a monthly publication brimming with articles on art and technique, profiles of photoplaywrights, and a tipsheet called "The Photo Play Market"

In one issue, the Holland Film Manufacturing Company of Dorchester, Massachusetts, put out a call for one- and two-reel comedies and comedy-dramas, the New York Motion Picture Company advised authors to send their material directly to scenario editor Richard V. Spencer, and Edison announced a contest for the most suitable conclusion to a prospective one-reeler.²⁶

Contests and similar schemes were constantly floated, in an effort to broaden the range of available materials, but without much success. One such contest promoted by Edison was directed at colleges across the country. Out of 337 scenarios submitted, only 8 were judged of produceable quality (the winner was "Jack Kennard, Coward," submitted by Harvard's William Marston). The conclusion of the Edison editors was that a dependence on amateur scenario writers was doomed to failure.²⁷

The small amount paid for original scenarios was hardly conducive to high-quality submissions. The Photoplay Author complained that writers accepting three dollars for a two-reel script were depressing the market: "At the present scale $35 is fair, $50 is better and $100 per reel good money. Most of the purchases are made at $50 or less per reel."²⁸

As late as 1923, Douglas Brown reported to the Society of Motion Picture Engineers that "the completed script of a feature picture costs the producer less than two thousand dollars." This sum did not, however, include the cost of any story rights involved, and beginning in 1919–1920 such costs began to soar for any property considered a sure success (essentially Broadway hits that already seemed to work in scenario form). "Apparently a season's run in New York automatically makes a play worth about $100,000 to the film producers," said the New York Times in 1920, with only slight exaggeration. Even before talkies, the existence of a usable Broadway playscript made a property far more interesting to film producers. For example, F. Scott Fitzgerald's The Great Gatsby reached the screen in 1926 via a 1925 stage version by Owen Davis, not directly from the original novel.²⁹

Prominent screenwriter Frances Marion reported in 1924 that the average price

for a successful play was $20,000. She offered the following list of high-priced properties:³⁰

Turn to the Right	$225,000
Way Down East	175,000
A Tailor-Made Man	105,000
The First Year	100,000
Tiger Rose	100,000
Daddies	100,000
The Gold Diggers	100,000
Merton of the Movies	100,000
The Virginian	90,000
Dorothy Vernon of Haddon Hall	85,000

This move toward the acquisition of pretested material clearly related to the collapse of the free-lance market. Submissions had grown so heavy that no quality control could be maintained. "When manuscripts come in they are handed over to the reading department," wrote an anonymous scenario editor in the Bookman in 1919. "This is a room where half a dozen women at an average salary of ten dollars a week, without the competence of a stenographer or salesgirl, sit all day making first choice of the material the editor is to see." According to this source, the women all had little scenarios of their own to promote, "consciously or unconsciously" stolen, which they schemed to place with their bosses, even to the extent of suppressing incoming material.³¹

Lawsuits by disgruntled authors were another problem, although one 1914 decision by a Los Angeles court dismissed charges of scenario-stealing against Hampton Del Ruth. "I have been given to understand that scenarios cannot be copyrighted," said the judge, incorrectly. "After looking into the question of scenarios I have decided they are of no value, and therefore dismiss the case."³²

The virtual elimination of the free-lance market was among the most significant production changes of the immediate postwar era. Writing credits went to such contract employees as Jane Murfin, Lenore Coffee, Charles Kenyon, Jeanie MacPherson, Waldemar Young, and Jules Furthman, skilled wordsmiths with backgrounds as reporters or short-fiction writers. The flood of Broadway playwrights that would engulf Hollywood in the talkie years was hardly in evidence before 1927, when dialogue skills were not a requirement.

The Producer System

Janet Staiger describes the central producer system as "the order of the day" by 1914 and invokes Thomas H. Ince's operation for the New York Motion Picture Company as the traditional example. Script material was recast into continuity form, which allowed careful preplanning of all production activities. Actors needed to appear only when required; props and costumes could be scheduled on a dependable basis, and the logistics of complicated location trips (or studio shoots, for that matter) might be clearly predetermined. By closely monitoring the scripting and editing process, a central producer like Ince—or Mack Sennett—could guarantee a uniform standard of quality without having to attend personally to the filming of each scene.³³

While this pioneering demonstration of organizational efficiency does mark Sennett and Ince as important innovators, their systems were primitive in comparison to those employed later in the silent period by more mature studios such as Paramount or MGM. In fact, the collapse of Ince's entire operation on the death of its central producer suggests that his studio was more an extended workshop than a true factory. Systems that could outlast their innovators reflected a higher level of organization and took several more years to develop.

In 1925 the new Metro-Goldwyn-Mayer studio produced a thirty-minute promotional film to demonstrate the power and scale of their factory operations. In true industrial-film fashion, they lay out the shape of their physical plant, boast of the capacity of their electrical powerhouse, and awe us with a staggering array of statistics. What is most interesting, however, is the way the film organizes and presents the studio workers. Dozens of cinematographers line up on a studio lawn, cranking away on Mitchells and Bell & Howells. They are matched by an equally formidable array of writers, directors, scenic artists, carpenters, electrical workers, cutters, even shippers packing MGM prints off to distant exchanges. Seen in control of this army of artists and technicians are three men, each busy behind a desk—Louis B. Mayer, Irving Thalberg, and Harry Rapf. Finally, we see a telegraph key that allows them to stay in constant touch with New York, where an unseen Marcus Loew and Nicholas Schenck call the ultimate shots.³⁴

This little film is especially revealing because it consciously deemphasizes the glamour of the studio's employees and underscores the tight, pyramidal control exercised by the top executives. MGM's stars are reduced to a few charming closeups. MGM's directors appear in a vast and nearly anonymous group. A title card announces, "Browning, Seastrom, Vidor, Niblo, von Stroheim, von Sternberg, …" but the names and faces do not really connect. All that matters is studio and system: clearly that is the message being communicated to the stockholders or theater owners who made up the film's original audience. To some later historians, this industrial self-consciousness must seem a simple admission of the way the business actually worked, the "real tinsel" underneath the usual phony tinsel.³⁵

Production of fodder for the nation's movie screens—many hundreds of pictures annually—was clearly generated by just such a system. Yet even Staiger suggests that the leading works of the age, the product of the most powerful stars and filmmakers, remained under individual control to a great degree.³⁶ The most memorable work of many of the key filmmakers discussed in chapter 8, Griffith, Weber, von Stroheim, Cruze, Lubitsch, DeMille, Neilan, and Ingram in particular, was created via a simpler director-unit system, where projects were developed from script level through editing by individual creative directors and their personal staffs. In the final analysis, these were the films that created the models for new styles or genres that were then mass-produced (often more lucratively) by the factory studios. While the centralproducer system certainly generated the bulk of American production in this period, those films that really mattered, to audiences of the time as well as to posterity, were often the dogged creations of an antiquated workshop system that somehow managed to survive well into the 1920s.

Just how efficient could this newer central-producer system be? One 1917 company was able to keep the costs of their five-reel features to three or four thousand dollars "by shooting players from one set to another with the speed of a Ford car assembly," but Taylorized efficiency was seldom as dramatic as this. A financial paper like Barron's looked for predictability as well as profitability, and here a studio like Famous Players-Lasky was at its best. "Their financial comptroller was formerly a cashier of the National Bank of Commerce. He has several young New York University statisticians associated with him," that paper reported approvingly. Of the last 164 films released by the company, only nine had failed to pay their expenses. Four of these were Arbuckle films, caught up in the scandal; one was an "English production"; and two others were only marginal failures.³⁷

Richard W. Saunders, the comptroller referred to above, outlined for the New York Times many of the recognized industrial practices that were, by 1926, beginning to attract large numbers of investors. These practices ranged from the traditional habit of forcing theater owners to pay in advance to a relatively new method of writing off negative costs in a standardized, monthly fashion.³⁸

When Saunders boasted to Forbes that Famous Players-Lasky had finally eliminated the seasonal peaks and slumps of production, he was making one of the manager's proudest boasts. These fits and starts had long been an embarrassment to those in the industry who had to deal with Wall Street, especially during extreme slumps, like that of the winter of 1923–1924. Then, Famous Players-Lasky stock dropped 12⅜ points when the management announced a total suspension of production, laying off nine hundred studio workers and putting contract players on half salary. Other studios cut back as well. By the end of the silent era, such economic swings no longer seemed an issue (fig. 4.1).³⁹

How Much to Make a Picture?

There was great interest, both inside and outside the trade, in establishing some sort of "average cost" for a standard program feature. In July 1917 the Motion Picture Classic published a "conservative average" budget based on information from three different production companies. The New York Dramatic Mirror, a trade paper, offered its estimate for a similar production that same month (table 4.1).

In retrospect, the budget proposed by the Dramatic Mirror seems to represent less an average price than a rock-bottom one. In any case, generalizations about average picture costs in this period are of little practical use because budgets rose so rapidly during the first dozen years of features, while even within a single studio or genre, costs could vary widely on different productions. As Samuel Goldwyn, who moved from Famous Players-Lasky to First National and then to United Artists during these years, told the New York Times in 1926, "In the old days the average

---

*The graph shows the average number of workers employed by sixteen "identical" companies. Data from Monthly Labor Review, Feb. 1927, pp. 66–69.

	Motion Picture Classic		New York Dramatic Mirror
Corporate salaries	$ 5,000		$ —
General manager	3,000		—
Director	2,500		3,000
Assistant directors	1,300	(two)	300	(one only)
Cameramen	1,200	(three)	300	(one only)
Assistant cameramen	600	(three)	100	(one only)
Stage carpenters	250	(three)	—
Other employees	1,800	(twenty)	—
Overhead expenses	900		—
Star	12,000	(two)	6,000	(one only)
Near stars	5,000	(three)	1,200	(one leading man)
Prominent characters	2,500	(three)	1,300	("other principals")
Other salaried players	3,000	(40)	—
Supers, etc.	1,100		300
Transportation	1,200		500	(transportation)
& location exp.			200	(location rental)
Scenario handling	—		1,000
Studio settings	—		1,000
Negative stock	—		1,000
Costumes	—		1,000
Incidentals	—		500
Total cost	$41,350		$17,700

negative cost of Famous Players was about $15,000. The distributors gave us an advance of $25,000. Today the average negative cost of Paramount productions is $300,000. The average negative cost of United Artists productions is from $750,000 to $800,000. The other two big companies are confronted with an average cost of between $240,000 and $260,000."⁴⁰

The average cost of Famous Players-Lasky releases had increased twenty-fold over thirteen years, but knowing a studio's "average " production cost may not, in fact, be very revealing. In 1921, for example, Universal spent $34,211.79 on The Way Back, a five-reel program feature. The cost of Foolish Wives, an unusual "special jewel feature," made that same year, was thirty times this amount (see tables 4.2 and 4.3).

Despite the tremendous differences in scale between the production of The Way Back and Foolish Wives, it is clear that staff salaries were a significant part of the budget of both pictures. In 1924 Barron's offered the following statistics to explain where the production dollar went, figures that were generally accepted throughout the industry:⁴¹

Actors' salaries	25%
Director, cameraman, and assistants	10
Scenario and stories	10

UNIVERSAL FILM MANUFACTURING CO.
Pacific Coast Studios
STATEMENT OF PRODUCTION COSTS
For Week Ending 2-1-22.
DIRECTOR: Paton	PICTURE NO: 3723 REELS: 5
TITLE: "THE WAY BACK"	DATE STARTED: 12-23-21
FEATURING: Frank Mayo	DATE FINISHED: 1-16-22
CHARGES:	PREVIOUSLY REPORTED	THIS WEEK	TOTAL TO DATE
Stock Talent Salaries	3205.00		3205.00
Pict. Talent Salaries	1979.20		1979.20
Extra Talent Salaries	895.25		895.25
Directors Producing "	1333.30		1333.30
" Writ'g & Edit'g "	1666.70	400.00	2066.70
" Staff "	1298.75	120.45	1419.20
Continuity Writers' "	525.00		525.00
Editors' & Cutters' "	267.45	102.85	370.30
Negative Raw Stock	1450.89	118.65	1569.54
" Laboratory Chgs.	322.13	31.73	353.86
Positive Raw Stock	992.25	38.43	1030.68
" Laboratory Chgs.	321.40	14.88	336.28
Wardrobe Pur. & Mfg.	98.20		98.20
" Hire & Expense	58.54		58.54
Rent Studio Ward. Equip.	252.41		252.41
Prop Pur. & Mfg.	44.92	3.75	48.67
" Hire & Exp:-Misc.	653.47	13.64	667.11
" " & " Horses
" " & " Special	40.00		40.00
Rent Studio & Prop. Equip.	606.90	17.09	623.99
Scenery	4744.24	256.12	5000.36
Auto Transportation	585.22		585.22
Traveling & Maintenance	330.70	2.00	332.70
Location Fees & Expense	25.00		25.00
Scenarios	2500.00	1000.00	3500.00
Light, Labor & Current	2153.31	226.90	2380.21
Rent Studio Elec. Equip.	1056.60	195.00	1251.60
Ranch Salaries & Exp.	1.00		1.00
Directors' Bonus
Arsenal Salaries & Exp.	45.40		45.40
Misc. Unclassified Exp.	179.02		179.02
New York Title Charges
Overhead Charges	4000.00		4000.00
Still Prints & Negatives	38.05		38.05
TOTALS,-	31670.30	2541.49	34211.79

Sets (manufacturing)	19
Studio overhead (including cutting, titling, etc.)	20
Costumes, gowns, etc.	3
Locations (and transportation)	8
Raw film	5

This negative cost factored into the total profit picture as follows:

Negative cost	40%
Distribution (U.S. and foreign)	30
Cost of positive prints	10
Administration and taxes	5
Profit	5

With salaries so large a part of production costs, much attention was devoted to limiting, or at least controlling, their growth.⁴² A major reason for the introduction of the continuity script was to better manage personnel resources, but competition for top talent continued to force these figures higher and higher.

Photoplay reported in 1916 that salaries of $1,000 per week had recently become common, while the highest figure for a single picture had reached $40,000 (Billie Burke's fee for Peggy). Leaving aside the well-known Chaplin and Pickford figures, they offered the following sampling of weekly star salaries:⁴³

William Collier	$ 2,500	Keystone
Raymond Hitchcock	2,500	Keystone
Sam Bernard	2,500	Keystone
Eddie Foy	2,000	Keystone
Weber and Fields	3,000	Keystone
De Wolf Hopper	125,000/year ($60,000 guarantee)	Keystone
Henry B. Walthall	500	Essanay
Blanche Sweet	750	Lasky
Marguerite Clarke	1,250	Lasky
Fannie Ward	1,200	Lasky (one picture only)
Valeska Suratt	5,000	Fox
Victor Moore	500	Lasky
Frank Keenan	1,000	Ince (one picture)
William S. Hart	300	Ince
Francis X. Bushman	750 (plus a percentage of the profits)	—
Beverly Bayne	350	—

Universal Film Man ufacturing Co.
Pacific Coast Studios
DAILY MEMORANDUM PICTURE COSTS
director Von Strohei picture no. 3322 episode no. date July 7, 1921
	Estimate	Amount Today	Total Amount to Date	Amount UNDER Estimate	Amount OVER Estimate
Overhead	$	$	$ 57,761.74	$	$
Stock Talent
Picture Talent
Extra Talent, Story and Continuity			257,283.20
Director and Staff		167.00	101,009.10
Film Editing			2,299.60
Negative			19,287.85
Sample Print			10,066.66
Wardrobe			20,258.79
Props (Arsenal, Drops, Props)			46,226.62
Scenery, Sets, Etc.		4.00	351,913.53
Transportation (Automobile)			33,178.88
Traveling			10,469.00
Maintenance (Lunches and Hotels)			41,209.54
Location Fees and Expenses			1,848.11
Lighting and Labor			77,010.23
Ranch and Zoo			9,347.97
Special Automobiles			814.90
Miscellaneous Supplies		6.00	13,305.08
TOTAL		$ 177.00	$ 1,053,290.80		$
Net Amount Under or Over Estimate	$

These figures reflect a relatively brief period when Broadway headliners were able to command salaries five times those of reliable film favorites such as William S. Hart. More typical of the era were weekly salary statistics offered by theater owner William Brandt following the announcement of the Famous Players-Lasky shutdown in 1923. Brandt's concern was that the exhibitors would bear the brunt of carrying these stars at half-salary while they were between pictures.⁴⁴

Norma Talmadge	$10,000	Conway Tearle	$2,750
Dorothy Dalton	7,500	Lewis Stone	2,500
Gloria Swanson	6,500	Milton Sills	2,500
Larry Semon	5,000	James Kirkwood	2,500
Constance Talmadge	5,000	Wallace Beery	2,500
Pauline Frederick	5,000	House Peters	2,500
Lillian Gish	5,000	Elaine Hammerstein	2,500
Tom-Mix	4,000	Bichard Barthelmess	2,500
Betty Compson	3,500	Betty Blythe	2,500
Barbara La Marr	3,500	Florence Vidor	2,000
May McAvoy	3,000	Elliott Dexter	2,000
Mabel Normand	3,000	Viola Dana	2,000
Priscilla Dean	3,000	Lon Chaney	1,750

Brandt's figures exclude United Artists' stars, and those, like William S. Hart and Harold Lloyd, whose incomes were tied to significant participation deals. By 1926 total earning figures to date for those stars had reached truly fabulous heights:

Harold Lloyd	$1.5 million per year
Charles Chaplin	1.25 million
Douglas Fairbanks	1 million
Mary Pickford	1 million
Gloria Swanson	Refused Famous Players-Lasky's offer of $1 million to sign with United Artists

The highest-paid star then on straight salary was Tom Mix, earning $15,000 per week at Fox.⁴⁵

Scenic Art

After salaries, the highest fixed costs were those related to set construction. Wilfred Buckland, Ben Carré, and Anton Grot were already established as "art directors" by 1915, but most settings were still designed and constructed by carpenters. Cameraman Arthur Miller remembered Grot as "the first art director I ever worked with who hadn't come up from the ranks of the construction department. " Grot created charcoal illustrations of the sets that displayed the scale and perspective of various motion-picture lenses. This technique enabled him to build only those segments of the set that would actually be used and resulted in substantial savings in construction costs. (He taught this skill to William Cameron Menzies.) Grot had received his training at the Akademie Sztuki in Cracow, Poland; Ben Carré came from the Paris Opera; and Wilfred Buckland was long associated with David Belasco.⁴⁶

In his very thorough 1918 study How Motion Pictures Are Made, Homer Croy was still giving all of the credit for set design and construction to gangs of carpenters:

With the scene locations determined upon, a list of the interior sets is handed the chief carpenter, who promptly starts building the necessary

woodwork. The list tells him in what order they will be wanted, with the date on which the first one will be needed impressed on his mind. He is held responsible for the finishing of the scene by the time specified…. From a bare wooden platform the carpenters, under the direction of their chief, may start to work (Homer Croy, How Motion Pictures Are Made [New York: Harper and Brothers, 1918], p. 110).

It should be remembered that, in this fashion, the master carpenter Huck Wortman constructed Griffith's Babylon in Intolerance, albeit under the supervision of theatrical designer Walter Hall. These two traditions—that of the graphic artist or scene painter on the one hand, and the practical carpenter on the other—remained dominant in American studios during the early feature period. But they shared an identical goal, one well articulated by Austin Lescarboura in Behind the Motion Picture Screen:

Realism is one of the main stocks in trade of the screen production. Compared with the speaking stage, with its highly artificial scenery which lacks correct perspective and general impressiveness, the motion picture makes use of backgrounds both natural and artificial which have depth as well as height and breadth …. Realism has made the success of present photoplays; and the screen artisans have made film realism what it is (Austin Lescarboura, Behind the Motion Picture Screen [New York; Munn, 1922], p. 107).

Lescarboura wrote just as von Stroheim's Foolish Wives appeared, with its full-scale reproduction of Monte Carlo's Casino, Hôtel de France, and Café de Paris, marking the apex of "realistic" set construction. Von Stroheims's mise-en-scène required such verisimilitude, but the vast majority of American films merely tried to stay one step ahead of the faultfinders. Scenic and technical accuracy became a fetish, and letters to the editors of various fan magazines prodded sloppy production teams.

But against this current, a feeble call for stylization defended that "highly artificial scenery which lacks correct perspective." As early as 1916 Edgar M. Kellar designed and produced The Yellow Girl, a self-described "decorative playlet" in the style of Aubrey Beardsley. One critic casually characterized it as "Futurist," but Kellar insisted, "I see no reason why we can't have a romantic rather than a practical background." More significant yet were Maurice Tourneur's The Blue Bird (1918) and Prunella (1918), both designed by Ben Carré. Invoking Gordon Craig, Max Reinhardt, Konstantin Stanislavsky, and Harley Granville-Barker, Tourneur declared, "The time has come where we can no longer merely photograph moving and inanimate objects and call it art…. We must present the effect such a scene has upon the artist-director's mind, so that an audience will catch the mental reaction."⁴⁷

Public rejection of these films was so severe that few remembered them in 1921, when The Carinet of Dr. Caligari first reached the Capitol Theatre. Audiences identified the stylization in this film as "the world seen through a madman's eyes" and promptly forgot the stylistic lesson Tourneur had offered them three years earlier. Now spatial and temporal distortion became signals for lunacy and delirium (note the drug addict's visions in Human Wreckage, 1923).⁴⁸ But even these hallucinatory exceptions were few, and three-dimensional realism ruled unchallenged.

This playground of graphic artists and carpenters lasted until around 1919, when producers began to turn to professional architects to create their settings. Robert M. Haas, who began designing for Famous Players-Lasky in New York in 1919, came to films after eight years as a practicing architect. His new position won two approving articles in the American Architect, which announced the fact that film design was now seen as "structural," not merely decorative, as in stage work. Haas was praised for the solidity of a town he had constructed in Elmhurst, New York, for The Copperhead (1920), where the details were aged to show the passage of time from 1846 to 1904. A rare full-card credit for the art direction on this film reads, "Robert M. Haas, Architecture; Charles O. Seesel, Decorations." His use of a ceiling "built to show" for On with the Dance (1920) was also cited as an example of a new kind of structural realism unlike anything previously seen on the screen. Of course, this new approach to design only underscored the prevailing stylistic mode. "The men who design the 'sets' are constantly striving for the better effect of actuality," observed the American Architect.⁴⁹

What Haas added to the Famous Players-Lasky art department was not just a new sense of creativity but a way of organizing the work flow that recalled the offices of top architectural firms rather than the workshops of fashionable graphic designers. In

the years that followed, teams of draftsmen, sketch artists, and model-makers would set a standard for studio efficiency at Paramount, MGM, and the other great Hollywood studios.

Behind the Camera

Unlike designers, cinematographers were not easily organized into hierarchical departments and in this period often contracted their services much as writers did. Some of the best were under contract to various studios on either a per-week or per-picture basis (for example, Joseph Ruttenberg with Fox from about 1915 to 1926). Some had long-term relationships, not always contractual, with the units of individual directors or stars (most notably G. W. Bitzer with D. W. Griffith from 1908 to 1928). But most drifted in and out of various jobs, hoping, especially before unionization began to offer a modicum of job security, for a decent run of steady employment.⁵⁰ Hal Mohr recalled that free-lance cameramen "used to haunt the studios" in the early 1920s, and groups of job-seekers would meet in the outer offices of production managers with work to offer. Often, the men would informally agree to demand the same salary, say $150 per week, in an effort to prevent a low bidder from depressing the wage scale. Unfortunately, someone always seemed to break the agreement, one reason for the considerable ill will and cliquishness that occasionally afflicted this particular craft.⁵¹

The diaries of Hal Sintzenich, a cameraman who struggled for work in the eastern studios during many production lulls, are filled with painful stories of missed paychecks and lost job opportunities. Erik Barnouw has analyzed the working records contained in these diaries:

Library of Congress [Summer/Fall 1980], p. 310">

Most of these assignments are brief. They involve unlimited hours and intermittent pay. Two or more cameramen are often"cranking" side by side, but sometimes they work simultaneously on different scenes. A cameraman may supervise laboratory work, or edit, or even act; in one film Snitch is handed a robe and becomes a judge. His earnings rise to $175 a week but often fall much lower ("The Sintzenich Diaries," Quarterly Journal of the Library of Congress [Summer/Fall 1980], p. 310).

The American Society of Cinematographers (ASC), a professional organization designed to exchange useful information and promote the qualifications of its members, was chartered in Hollywood in 1919 (although earlier component organizations date from 1913). Most top cameramen quickly joined and as a group resisted the unionization they saw developing in the eastern studios. Dan Clark, ASC president in 1926, explained this anti-union position in terms of their perceived status as artists, not artisans, as well as a resistance to the fixed-wage scales they saw as detrimental to their own members. Or, as Hal Mohr remembered, "I made a pretty good reputation for myself by 1928 and I was pretty much in demand and considered a pretty fine cameraman, getting a high salary. I was getting around $350 a week then. So I figured, what the hell do I want with a union organization?" But Mohr did join, serving during a long career as president not only of the ASC but of the union local as well.⁵²

Largely because of the existence of the ASC and its house organ, the American Cinematographer, artistic and technical problems of cameramen were given relatively sophisticated discussion in a public forum, something that was not often the case for other film workers. During this period, these problems included the use of the close-up, "soft" versus "sharp" photography, the influence of German films and filmmakers, and the use of color.

While not necessarily a photographic issue, the close-up was a matter of some controversy throughout this period and of special concern to the cameramen because one of their prime responsibilities was the lighting of glamorous star portraits. Some historians argue that nickelodeon audiences resisted the introduction of the close-up, and patrons of early features in 1915 had their doubts as well. In reviewing David Harum (1915), the New York Dramatic Mirror commented on one sequence of Harum at dinner, which showed only his hands and the food he was eating. This "caused a spectator behind us to say at once that it was 'a poor picture because you cannot see his face."⁵³

Audiences soon grew sufficiently sophisticated to accept the existence of offscreen space, but various stylistic complaints continued. Extreme close-ups were so rare as to be beyond general notice. The American Cinematographer defined the two forms of the close-up in 1923 as the waist-to-head two-shot and the chest-to-head single figure. Full-face was clearly too rare to consider, despite its dramatic use in Intolerance and other Griffith films. Welford Beaton, an otherwise perspicacious critic, waged a lengthy war against the close-up. In a review entitled "Submerging the Production Under Senseless Close-Ups," he attacked Alexander Korda's The Yellow Lily (1928) as a "close-up orgy" and contended, "In this picture we have elaborate sets which flit across the screen to give place to an endless parade of utterly senseless close-ups. Ordinary business sense would dictate that the sets should be shown for a longer time to justify their cost. " As for von Stroheim's The Wedding March, Beaton told readers, "I would estimate that there are between seven and eight hundred close-ups in the entire picture, proving that von Stroheim treated Griffiths discovery as wildly as he did Pat's bankroll." (The "Pat" in Beaton's review is Pat Powers, producer of The Wedding March.) Similarly, Frank Tuttle, director of a self-described "artistic" film production unit called the Film Guild, wrote in 1922, "The close-up mania is like the drug habit. It grows upon the afflicted company at a constantly accelerating pace until the whole studio is mortally ill of it."⁵⁴

A wide array of diffusion effects suddenly became popular after 1919, when Henrik Sartov and G. W. Bitzer filmed close-ups of Lillian Gish that recalled Photo-Secessionist portraiture. The style soon spread even to war films. "In The Four Horsemen I made all the exteriors I could on dull days in order to use an open lens and get a softer image," explained director Rex Ingram in 1921. "I wanted to get away from the hard, crisp effect of the photograph and get something of the mellow mezzotint of the painting; to get the fidelity of photography, but the softness of the old master; to picture not only the dramatic action, but to give it some of the merit of art."⁵⁵

Some years later, cameraman Henry Sharp, then in New York filming The Crowd (1928), found it useful to study Rembrandt's work in the Metropolitan Museum of Art. "Rembrandt's great strength was his use of one positive light scale. One central 'light perspective' was always used, rather than a multiplicity of attempted effects,

and the results were contrasts in light and shade," he noted. The invocation of Rembrandt was used to cover a variety of pictorial effects, from Alvin Wyckoff's "smash of light from one side or the other" in DeMille's The Warrens of Virginia (1915) to Lee Garmes's celebrated "north light" effect. "Ever since I began, Rem-brandt has been my favorite artist," Garmes told interviewer Charles Higham. "I've always used his technique of north light—of having my main source of light on a set always come from the north…. And of course I've always followed Rembrandt in my fondness for low key."⁵⁶

By the close of the silent period, many of the cameramen most involved in the use of incandescent Mazda lighting, notably Lee Garmes and George Barnes, had eliminated much of the diffusion from their work, but Ernest Palmer, Karl Struss, and Oliver Marsh still continued to make heavy use of the more pictorial style.⁵⁷

The most dramatic outside influence in silent Hollywood was certainly the importation, after 1925, of an entire generation of German filmmakers. Karl Struss, one of the American cameramen on F. W. Murnau's Sunrise (1927), a film with a large number of Germans on the design team, saw the picture as "the forerunner of a new type of picture-play in which thought is expressed pictorially instead of by titles."⁵⁸ Such cameramen as Gilbert Warrenton (The Cat and the Canary, 1927), Ernest

Palmer (The Four Devils, 1928), and Hal Mohr (The Last Warning, 1929) were especially involved in this style which featured complicated moving camera effects, frequent use of trick shots and superimposition, and stylized, low-key lighting schemes. But the death of the most important German technicians (notably Murnau and Leni), photographic problems attendant on the introduction of sound, and a general misuse of the techniques involved muted the impact of this movement after 1929. Looking back at the period, Hal Mohr, who photographed some exceptionally "Germanic" late silent films for Michael Curtiz, Paul Fejos, and Paul Leni, remembered it simply as "the era of the goofy ideas in film":

I'll never forget one thing we did on Broadway [1929; an early talkie that proved to be one of the last great examples of the style]. We had this camera swinging around during one of the musical numbers, just rotating, swinging around the camera, photographing everyone on those sets all at one time like a big merry-go-round type of thing. If you make them dizzy enough they'll think it's a great scene, you know (quoted in Richard Koszarski, "Moving Pictures," Film Comment, September-October 1974, p. 48).

Film Color

Until the introduction of Eastman Color multilayer negative and printing stock in 1949, color in film was largely a matter of craft. During the 1915–1928 period, color systems might be natural or artificial, with the latter by far the most commonly employed. Artificial-color systems date from the birth of cinema and involve a coloring of all or part of the image of each release print. Processes available were tinting, toning, hand-coloring, stenciling, and related variants, all of which were well established before the introduction of feature pictures.⁵⁹

Tinting was defined by the Society of Motion Picture Engineers as "immersing the film in a solution of dye which colors the gelatine, causing the whole picture to have a uniform veil of color on the screen." Far from declining in popularity after the nickelodeon years, by 1920, according to the society's estimate, tinting was used for 80 to 90 percent of all films. Eastman Kodak began supplying pre-tinted release stock by 1921, eliminating the need for individual treatment of each release print. In toning, "a colored image [is] embedded in a layer of colorless gelatine, so that while the highlights are clear, the shadows are colored." This effect was achieved not by dying the prints but by causing a chemical change in the composition of the metallic salts that had created the black-and-white image, converting the original silver deposit into silver ferrocyanide, for example, to create a blue-and-white image. Laboratory costs were higher for toning than for tinting, and of course, no one could produce a release stock that was already toned.⁶⁰

Combined tinting and toning allowed for complex color effects, such as "sunset and moonlight effects over water," but these were still more difficult and costly. King Vidor used a wide range of tinting and toning effects in The Sky Pilot (1921), not only for decorative effect but as a significant element of the film's mise-en-scène:

What I believe I have actually done is to "score" this photoplay for color. I have used a soft violet tint for scenes in which the earlier hours of the morning—the phantom dawn, as the orientals call the period just before sunrise—are represented. For the period after sunrise, I have used a pale yellow tint; for noon, a faint amber; for night, a blue green tone on all objects casting shadows, high lighted with a warm amber.

I have tinted the moonlight scenes with carefully chosen deep blue tone, tinting the moon a faint, almost ethereal, amber, while I have used the conventional amber for interior night scenes. A delicate tint of green is used in all scenes of virgin nature where the day is supposed to be warm, while in the Canadian northwest snow scenes I have used a steel blue tint. To induce certain moods, I have scientifically played upon the varying degrees of happiness and sorrow with varying shades of pink and green. The heights of joy are enhanced with a delicate pink glow, while the depths of grief call for a ghastly gray green tone ("Brought into Focus," New York Times, 27 February 1921).

Another film of the period whose use of color is well documented is von Stroheim's The Devil's Pass Key (1920), in which the various quick cuts of a dance sequence flash green, violet, red, rose, and amber. In The Anatomy of Motion Picture Art (1928) Eric Elliott credited this film with "introducing colour as a psychological influence on the film scene."⁶¹

Highly popular in earlier years, hand-coloring of individual frames was, after the arrival of feature films, usually limited to brief sequences in a few selected prints. Gustav Brock was one of the last to offer this specialty, advertising regularly in the annual Film Daily Yearbook. Brock labored over Foolish Wives, The White Sister (1923), The Navigator (1924), and a number of Marion Davies pictures, including Little Old New York (1923). He did two sequences for this film: one in which Davies, disguised as a boy, overhears a risqué story and blushes; the other, a scene of the stars and stripes being raised on Robert Fulton's Clermont. The eight seconds of film required thirty hours of labor. Brock never colored more than a few prints of any title, so only the prime first-run houses were likely to screen his work.⁶²

The stencil coloring system introduced by Pathé Frères had mechanized this hand-coloring process to a certain degree, using a mimeograph-like system to mass-produce release prints from an original master. Although continuing in France, stencil coloring was rarely seen here after Pathé stopped American production in 1914. Its place was taken by the Handschiegl process, developed by Max Handschiegl, an engraver who had perfected a means of dye-transfer coloring of release prints, otherwise known as imbibition. Using his knowledge of printing inks and engraving technology, Handschiegl prepared as many as three printing matrices to achieve a desired color effect. Beginning in 1917, he worked on such films as Joan the Woman (1917), Greed, The Phantom of the Opera, Sally (1925), and The Big Parade.⁶³

Natural color films, "those photographed so that the colors are selected entirely by optical and mechanical means and reproduced again in a like manner," have a history almost as long as artificially colored films. In the silent-feature era, all those systems which achieved any commercial success initially separated each image into a pair of "color records" (i.e., monochrome images, each recording the presence of a specific primary color), one frame generally carrying the blue-green record, another the red-orange. From these two primaries, shot on panchromatic negative stock, various methods were used to achieve a natural color screen image.⁶⁴

Kinemacolor, the pioneer in this field, was effectively out of business by 1915. An additive system, it had projected its alternate frames through a revolving shutter that reintroduced the original colors. Commercially successful at first, it ultimately failed because of various technical liabilities and managerial problems.⁶⁵

William Van Doren Kelley improved this system by devising a method of coloring the release prints. As a subtractive process, his Prizmacolor could be projected on standard equipment, which Kinemacolor could not. At first, Prizmacolor subtractive release prints had one frame toned red-orange, and the next toned blue-green, with the "natural color" image appearing only when the projected film was viewed onscreen. Kelley later began using duplitized stock (that is, stock with an emulsion on each side), printing the red-orange record on one side and the blue-green perfectly in register behind it. After toning each side the proper shade, he obtained a print in which every frame displayed "natural color" when viewed through transmitted light.

This double-coated system was the progenitor of Trucolor and other two-color processes later marketed by Consolidated Film Industries, which acquired the patent rights.⁶⁶

Because each color record was still recorded sequentially (not simultaneously), there was a problem with color fringes around moving objects, and as a result, Prizma had its greatest impact in relatively static travelogue films. In 1922 Prizma made and released twenty-six shorts, the only all-color regular short-subject package on the market. The process was also used for occasional inserts in such features as Way Down East (1920), The Gilded Lily (1921), and Broadway Rose (1922). In 1921 J. Stuart Blackton directed The Glorious Adventure, an all-Prizmacolor feature filmed in England.⁶⁷

Beyond its problems with fringing, Prizmacolor had two major liabilities: excessive graininess and poor quality control. To solve these problems, Kelley joined forces with Max Handschiegl in 1926 and began to use Handschiegl's imbibition process to produce single-coated release prints under the name Kelley Color. But by this time the color film market was already dominated by the Technicolor Corporation, a situation that would continue for nearly thirty years.⁶⁸

After experimenting with an additive system on The Gulf Between (1917), Herbert T. Kalmus and Daniel Comstock introduced subtractive two-color Technicolor with The Toll of the Sea in 1922. A beam splitter inside the camera recorded the two color records simultaneously, foot-to-foot on the original negative, thus eliminating the fringing problems of earlier systems. Skip-printing every other frame made it possible to separate the blue-green record onto one filmstrip, the red-orange onto another. Prints from these strips were then toned the proper color and cemented back-to-back to form each release print.⁶⁹

Despite the success of Toll of the Sea (produced by Technicolor and distributed by Metro), there was considerable industry resistance to any color system. Rex Ingram agreed to have Technicolor film an exterior sequence for The Prisoner of Zenda (1922), but "we thought the people looked like dark colored oranges, so we took it out of the picture and threw it away," recalled his editor, Grant Whytock, years later.⁷⁰ Successful insert work on The Ten Commandments (1923) and Cytherea (1924) led to the production of such all-Technicolor films as Wanderer of the Wasteland (1924) and The Black Pirate (1926), but this "process two" had inherent difficulties of its own: the cemented film strips buckled under the heat of projection, throwing the picture out of focus.

Finally, Technicolor "process three" substituted a dye-transfer imbibition system for making release prints, one similar to that employed by Kelley and Handschiegl.⁷¹ This produced a smooth, grainless image with far greater uniformity of color than any system based on chemical toning. King of Kings (1927) and The Wedding March (1928) made use of this system, which remained in general use until 1933.

When contracting with the Technicolor Corporation to use their process, producers acquired not only the Technicolor cameras but special cameramen to operate them, at first J. Arthur Ball and later Ray Rennahan. While generally open to new technologies, the ASC at first strongly resisted Technicolor since it meant subordinating their members to a group of color technicians. Phil Rosen, one of the founders of the ASC, felt that color in film was distracting and that "true art does not necessarily mean the exact reproduction of nature." Ray Rennahan, for example, was tolerated by ASC members as a slave to his light meter who always worked by the book, a technocrat rather than an artist. Not until 1938 was he invited to join this club.⁷²

Performance

The studio machinery could deliver scripts, sets, and cinematographers in an impressive and efficient manner, but no studio manager or unit producer could guarantee what might happen after the doors closed and shooting began on the stages. Soviet director Lev Kuleshov had demonstrated that traditional acting was irrelevant in silent pictures, since the editing process imposed its own structure regardless of the intentions of the original performer. This idea would not have surprised the producers of the Our Gang comedies or the many animal films starring Rin-Tin-Tin, Rex the Wonder Horse, or the Dippy Do-Dads (a group of monkeys). Given this situation, actors and actresses needed to work with their directors to create a new style of performance directly for the silent screen, a style that had no real precedent and would change dramatically with the coming of "talkies" after 1927.⁷³

Nothing better illustrates the idiosyncratic nature of silent-film performance than the use of sideline musicians on the set to inspire the cast during filming, a practice that Kevin Brownlow dates to 1913.⁷⁴ As in a Victorian theater piece, this underscoring of drama with music enhanced the essentially melodramatic nature of the performance, but here it was for the benefit of the actors, not the audience. "During the making of a picture music has become essential," wrote one accompanist in 1923. "There was a time when the appearance of a violinist and a pianist on the set in a studio provoked laughter. Their presence elicited ridicule, but not now."⁷⁵

Musicians needed to command a repertoire of sad, dramatic, and joyous arrangements and to be capable of flitting in an instant from the Moonlight Sonata to "Yes, We Have No Bananas." Gounod's "Ave Maria" and the intermezzo from Cavalleria Rusticana were judged most successful for tears, while dramatic action might call for Tchaikovsky's Fourth Symphony, the Massenet Élégie, or a Chopin prelude. According to the accompanist just cited, certain stars or directors would have their own favorite pieces, regardless of the action:

Thomas Meighan likes "Macushla," possibly because he is Irish. Alice Brady prefers selections from Madame Butterfly, and Agnes Ayres has a penchant for "Kiss Me Again." D. W. Griffith sometimes chooses dreamy Hawaiian music ("Real Inspiration," New York Times, 24 June 1923).

One special problem of silent-screen acting was the need to accommodate the over-speeding of projection, which was the inevitable fate of every projected "performance." Milton Sills acknowledged this in the entry on "Motion Picture Acting" he prepared for the fourteenth edition of the Encyclopaedia Britannica. Admitting that performances recorded at 60 feet per minute were typically projected at 90 feet per minute, Sills felt it "necessary for the actor to adopt a more deliberate tempo than that of the stage or real life. He must learn to time his actions in accordance with the requirements of the camera, making it neither too fast nor too slow."⁷⁶

Controlling this tempo was ultimately the job of the director, who would elicit a performance in one of two very different ways. As Gloria Swanson put it:

There's the director who allows the actor to give his own interpretation of a part, and becomes a conductor of an orchestra, moving it down or bringing it up. Then there's the other kind where the director is a thwarted actor, a ham as we call him, who wants to show the actor how to do it (quoted in Rui Nogueira, "I Am Not Going to Write My Memoirs," Sight and Sound, Spring 1969, p. 59).

Here, Swanson was praising her longtime collaborator Cecil B. DeMille, who would respond to actors' questions with an abrupt "I'm not running an acting school!" This approach was very much the opposite of another of Swanson's directors, Erich von Stroheim, who was notorious for indicating exactly how he wanted every gesture delivered. Von Stroheim would keep filming retakes until he saw the performance he had conceived in his mind's eye, a technique Swanson accuses Chaplin of employing as well (and with some justification, given the evidence provided in the Kevin Brownlow-David Gill Thames television series The Unknown Chaplin). As for D. W. Griffith, Swanson claimed:

You could always tell when an actor had been working with Griffith, because they all had the same gestures. All of them. They'd cower like mice when they were frightened, they'd shut not their fingers but their fists, they'd turn down their mouths…. Lillian Gish, Dorothy Gish, all of them. Even the men had a stamp on them (Nogueira, p. 59).

The onscreen evidence indicates that von Stroheim, Chaplin, and Griffith were able to use this Svengali approach to good effect. Unfortunately, far less talented directors often employed equally intrusive techniques:

Well do I remember watching J. Searle Dawley direct Pearl White in an intensely dramatic scene, in which he played all the parts before the rehearsal was over…. Mr. Dawley did everything. He reclined on the floor, as Miss White was to do, and leaned back in the villain's arms. He played the villain, and snatched her to him despite her struggles. And then, as the big red blooded hero, he burst into the room and hurled the villain back against the wall so forcibly that it shook (Inez and Helen Klumph, Screen Acting: Its Requirements and Rewards [New York: Falk, 1922], pp. 185–186).

By the late 1920s this technique had generally fallen out of favor, although some, including Ernst Lubitsch, continued to employ it successfully. Raoul Walsh, another of Swanson's directors, summed up the case for the DeMille approach:

I should like to know how the silent drama is to develop great talent if the practice of curbing the players is adhered to. Granted the director must play an important part, he must supervise the players and see that they are getting the right stuff into the scene. There should, however, be a happy medium, as overdirection causes players to be self-conscious, mechanical and as colorless as dolls or marionettes on a string ("Spontaneity in Acting," New York Times, 27 April 1924).

Editing

As soon as filming had started, the director and editor would screen the footage processed each day ("the dailies"), make initial decisions on the best takes, and begin assembling the work print. This would be relatively easy for scenes where action was limited and only a few characters appeared, but complicated sequences presented far greater problems. Because editing was acknowledged as crucial by everyone from Griffith to Sennett, it was considered important for the director to establish an intimate working relationship with his cutter. Griffith worked for many years with Jimmy Smith, and Rex Ingram had his most successful pictures cut by Grant Whytock.

Ingram would film spectacular scenes for The Four Horsemen of the Apocalypse with twelve or fourteen cameras grinding simultaneously. It was impossible for him to analyze so much footage each night, so he depended on Whytock to make an initial selection—something he did by ignoring all but two or three of the master cameras! The editor also assembled another original negative for foreign markets. Shots were recorded twice, either by adjacent cameras or through an additional take. Kodak duplicating stock was not available until 1926, and previous attempts at producing foreign negatives by duping the original resulted in such poor print quality that overseas audiences complained. For The Four Horsemen of the Apocalypse, a third "original" negative was assembled by Whytock after heavy print demand wore out the primary domestic negative. Produced from third-best takes, this version lacked a few key shots and sequences, and Whytock considered himself lucky that "it all made sense."⁷⁷ (In 1971 Foolish Wives was reconstructed by Arthur Lennig for the American Film Institute by wholesale interweaving of material from such foreign and domestic versions. Predictable problems with matching action resulted, especially given the film's long history of recutting and censorship.)⁷⁸

When a film was completed, it was previewed at a local theater to gauge public reaction. Harold Lloyd, one of the most aggressive users of the preview system, remembered:

Even back in the one-reel days, I would take a picture out to a theater when I knew the picture wasn't right. And the manager used to always have to come out and explain what was going on. When we were doing two-reelers, he came out in white tie and tails to do it, and it was quite an event for him and the audience would listen attentively ("The Serious Business of Being Funny," Film Comment, Fall 1969, p. 47).

By the time he moved to features, Lloyd would take "scientific" readings of audience laughter and plot them on large graphs, using the information to refine the way his comedies played.

Lloyd's rival Buster Keaton filmed a scene for The Navigator in which, as an underwater traffic cop, he directed the movement of swarming schools of fish. Despite the fact that he was quite proud of this expensive gag, Keaton cut it from the picture when preview audiences proved too astonished to laugh.⁷⁹

Occasionally, an already completed film would be drastically reshuffled following previews. Wrote critic Welford Beaton, months before the official premiere of The Wedding March:

"I have seen The Wedding March twice, once at Anaheim and again at Long Beach. The Anaheim version had the weakness of characterizing von Stroheim as an all-good hero, a role that he has neither the appearance nor the personality to play convincingly. Nicki was made so spotless in that version that when the seduction scene was reached it gave the impression that Mitzi, his victim, had been the aggressor. In the final version Nicki is presented as the roue that he was drawn in the original story [sic], and all the scenes which developed that side of him were put back into the picture ("At Last One Fragment of Von's Opus Reaches Screen," Film Spectator, 17 March 1928, p. 7).

After the first preview of The Phantom of the Opera, a decision was made to put some comic relief into the film, and a considerable amount of footage with Chester Conklin was inserted. Later preview screenings led Universal to cut Conklin out of the picture entirely.⁸⁰ For modern audiences, Conklin is back in again, courtesy of a 1930 reissue and some questionable "film restoration."

Usually, of course, the preview process was far less traumatic, just another step in the production and polishing of any film. Henry King's rough cut of Stella Dallas (1925) was 26,000 feet, which he trimmed to 18,000 before previewing it in San Bernardino. Although nobody walked out on this 3^1/2-hour cut, he trimmed another 3,000 feet before running it in Pasadena. At the screening of this fifteen-reel version, some 800 postcards were distributed asking for suggestions of which 420 were returned. It was discovered that Stella's dialogue titles were "resented by the public," so these were all done over. After this version was previewed in another small town, two weak sequences were excised, and the film went into general release at 10,157 feet.⁸¹

The director most taken with the preview idea was probably D. W. Griffith, who considered it an extension of the theatrical out-of-town tryout. Even the opening-night version was just another cut to Griffith, who continued refining his work well after the initial public showings. "As was his habit," Eileen Bowser tell us, "D. W. Griffith accompanied Intolerance on its first runs in the major cities, cutting the prints at the theater, striving to improve it. The result was that the print shown in Boston was not necessarily the same as that shown in New York, and it may be that neither was matched exactly by the original negative."⁸²

Cameraman Hal Sintzenich's diaries show that Griffith not only recut his films after opening night but often continued shooting new footage as well. Notes Erik Barnouw:

On February 10, 1924, Snitch mentions a [preview] showing of America in South Norwalk, Connecticut. On the same day he is shooting scenes for its Valley Forge sequence in nearby Westchester, impelled by the arrival of ideal blizzard weather for "the men in bare feet in the snow trying to pull the big wagon" (2/10/24). The Battle of Princeton is shot the following day, in time to be included four days later in a showing in Danbury, Connecticut, when Snitch is shooting Washington's inauguration. Two days later they are remaking "the stockade scenes." On the day of the New York premiere, February 21, Snitch is still doing close-ups of Carol

Dempster, and during the following week does retakes of "Miss Dempster & Hamilton " and new "stockade scenes with 30 extras" ("The Sintzenich Diaries," p. 326).

Years later, at the Museum of Modern Art, Griffith was still eager to recut his pictures, running the now-antique prints in the Film Library's screening room, never satisfied that he had achieved a definitive version.⁸³

The implications of such behavior for film scholarship are considerable. What is the authentic version of such a work? Griffith's incessant adding and subtracting of footage implies that he saw these films as essentially open texts, capable of showing one face to Boston and another to New York. Yet he copyrighted the montage of The Birth of a Nation and Intolerance by submitting a frame from each shot to the United States Copyright Office. Did he see those versions as definitive? And if so, why the inevitable "improvements" that followed?

The problem goes beyond Griffith. By the late silent period, exhibitors could choose alternate endings for a number of major films. Some audiences, viewing Garbo as Anna Karenina in Clarence Brown's Love (1927), saw Anna throw herself under a train. Other theaters showed Anna happily reunited with Count Vronsky. King Vidor shot seven endings for The Crowd and apparently issued it with two. Griffith's Drums of Love (1928) still exists with a pair of different endings, and when it is screened by archives, audiences sometimes see them one after the other. Why producers suddenly lost confidence in their ability to make such basic decisions is unclear, but if they were moved to think twice about these films, later generations should be at least as careful when asserting that any version of a silent film is definitive.⁸⁴