Even when he cast by character in the shorts, thereby making his animators the equivalents of live actors, Disney was not entirely successful. His most individual characters were always a little generalized compared with real people. The more naked a cartoon’s plot, the more it magnified this shortcoming. A “story cartoon” like Elmer Elephant, with simple characters and simple plot, and music subordinate to both, was unmistakably juvenile, in a way that an intricate miniature operetta like Three Little Pigs was not, even though Pigs was based on a children’s story. Disney’s limited success with casting by character may have persuaded him that he had little to lose by taking a different approach.
The dwarfs in particular demanded a level of complexity that no earlier Disney characters had approached, and by 1935 they still existed mostly as vague story sketches. Audiences had to be able to tell them apart easily—they had to look alike, and yet different, but some elements of their appearance, like their clothes and beards, did not lend themselves to sharp differentiation. The vital task that Disney presented to Fred Moore and Bill Tytla at the beginning of 1936 was to make distinct everything about each dwarf that could be made distinct—eyes, noses, mouths, posture, waistlines. (The idea at first was to differentiate them even further by clothing them in what the color stylist Maurice Noble called “strong, simple colors.”)65
For Snow White herself, Disney had an even more striking answer to the question of how to preserve consistency across the work of several animators. All of Snow White’s scenes were to be photographed in live action first, and the animation would then be based on tracings of the frame blowups. Disney evidently had something like this procedure in mind from early in work on Snow White. A memorandum titled “Routine Procedure on Feature Production”—undated, but written in the fall of 1934—assigned to the writer Harold Helvenston responsibility for “stage settings, sets, props, costumes,” and said that he “will be responsible for the setting of the stages, the production of all props and sets, and will see that the work on the stage progresses with a minimum time load.”66
This was probably Disney’s response to the inadequacies of the character animation he saw in The Goddess of Spring. Why struggle with the animation of the girl, he may have reasoned, when a solution (already used extensively by Disney’s rival cartoon producer Max Fleischer) was close at hand? The dwarfs and not the girl were to be at the center of the film, in any case.
Filming began under Ham Luske’s direction in November 1935, with Marjorie Belcher, the teenage daughter of a dance studio’s owner, performing as Snow White. In the earliest days of the live-action filming, she wore what she later remembered as a sort of helmet, with Snow White’s hair painted on it, in an extreme example of the effort to bring the live action as close as possible to the result desired in the animation. (Snow White’s head was to be larger in proportion to her body than the real girl’s.) The helmet was hot and uncomfortable, she said, “so I’m sure it restricted my movements a lot, and they soon gave that up.” She always wore a Snow White costume during filming, however.67
By the time he wrote his November 25 memorandum, Disney had six story units, ranging in size from one to four men, working on Snow White, and he expected to shuffle those men around as they finished work on particular sequences. In the same spirit of confidence and command evident in so much of what he said in the last months of 1935, he clearly anticipated a steady flow of work from both animators and writers.
In one aspect of the production, he was not disappointed. There was friction between Ham Luske (who adhered to Disney’s wish that Snow White be presented as a sweet child) and Grim Natwick (who wanted the girl to be more mature and knowing), but there was never any question who would prevail, and animation of many of the scenes devoted to Snow White herself proceeded smoothly. In other respects, though, work on Snow White was a steady slog, as Disney and his people struggled to recover from two fundamental errors.
For one thing, they had misjudged the nature of the story. The Grimms’ version of “Snow White” is a serious fable about a girl—and about youth and age, and sexual maturity, and life and death—and not a vehicle for seven funny little men. Even the Marguerite Clark version, as clumsy as it was, assigned the dwarfs a strictly supporting role. But from the beginning, Disney had conceived his version of the story mainly in comic terms, with lots of gags for the dwarfs. It was taken for granted in various synopses and treatments that even the more menacing characters—the queen herself, especially in her disguise as an old hag, and the vultures that were to circle down after her fatal fall—would be treated as figures of fun. The story could not accommodate so radically different a point of view without being changed fundamentally, into something much less serious—not more comic, but trivial—than the Grimms’ original.
(Some changes in the Grimms’ story could be accommodated much more easily. The prince’s kiss that awakens Disney’s Snow White is borrowed from Perrault’s “Sleeping Beauty.” In the Grimms’ original, the girl awakens when a piece of poisoned apple is dislodged from her throat.)
The dwarfs themselves were at the heart of the second mistake.
It was not until late in September 1936 that Fred Moore produced the final model sheets of the dwarfs, each specific enough to be used by the assistant animators who would clean up the animators’ drawings. By then, he and Tytla had been working on the dwarfs—designs and two pilot sequences—for about nine months. Moore, the avatar of charm and cuteness, had so refined the dwarfs that the knobbiness and wiriness of the earliest storyboard drawings had been smoothed away. Beards, jowls, bellies, all were clearly different on each of the seven characters, who had names—Doc, Grumpy, Dopey, Happy, Sneezy, Bashful, and Sleepy—that fit some salient trait. The animators who followed after Moore and Tytla would be working with sturdy, all but foolproof designs.
Or so Disney must have hoped and expected. But even though Moore and Tytla had worked together for so long, there were already visible differences between their versions of some of the dwarfs. Tytla was a powerful draftsman whose work naturally veered away from Moore’s softness and cuteness in favor of a more intense and muscular sort of animation. He had animated the dwarfs as they washed for supper after Snow White sent them outside, and his Grumpy and Doc, in particular, were vigorous physical presences in a way that Moore’s were not.
By late October 1936, other animators had begun to work on scenes with the dwarfs, and as they did, such differences multiplied. For all the work that Moore had put into his model sheets, it was not enough. The animators now trying to master the dwarfs had trouble keeping them distinct on the screen—it was not always clear which dwarf was which—and the same dwarf could look very different in different animators’ hands.
Disney had stepped back from casting by character because of the difficulties it promised. Now reconciling so many different versions of so many different characters was proving to be even more difficult. As he recalled at a meeting almost two years after Snow White was finished: “We’d be in here with Marvin [Woodward] or [James] Culhane or one of those guys and in order to get it over, we used to have to call Fred [Moore] in.”68 Nothing could have seemed more natural than to deal with the dwarfs, and their animators, en masse; and yet, if Disney had assigned the dwarfs individually—one to an animator—he could scarcely have gone through more arduous struggles than those he endured as he tried to pull the other dwarf animators into line with the work of Moore and Tytla, as well as to reconcile the work of his two principal animators.
By the fall of 1936, Disney had abandoned any thought of being Snow White’s director, at least in the usual sense. Although he had planned to leave the shorts directors in place, he brought David Hand onto the film to badger and cajole the dwarf animators. (Hand ultimately became Snow White’s supervising director. The other two shorts directors, Ben Sharpsteen and Wilfred Jackson, also came onto the film to direct parts of it.)
In late 1936 and early 1937, Hand followed Ham Luske’s example and b
egan shooting live action that could guide the dwarf animators, as live action was guiding Luske and the animators working under him on Snow White. The use of live action for the dwarfs actually originated, Frank Thomas and Ollie Johnston have written, “during a discussion of how . . . Dopey should act in a particular situation” (this would have been in early 1936). Someone suggested Eddie Collins, a burlesque comic, as a model. A group went to the burlesque house to see him perform, he was invited to the studio, and “a film was shot of his innovative interpretations of Dopey’s reaction—a completely new concept that began to breathe life into the little cartoon character,” unformed till then. “Freddie Moore had the assignment of doing the experimental animation on Dopey, and he ran the Collins film over and over on his Moviola, searching not so much for specifics as for the overall concept of a character. Then he sat down at his desk and animated a couple of scenes that fairly sparkled with fresh ideas. Walt turned to the men gathered in the sweatbox and said, ‘Why don’t we do more of this?’ ”69
From all appearances, live action did prove genuinely useful on a number of occasions in 1937. The actor Roy Atwell was filmed as he delivered some of Doc’s dialogue, for example, and Frank Thomas, when he animated that scene, successfully caricatured Atwell’s nervous hand movements as he spoke. Thomas so well integrated those movements with the dialogue that the words seemed, as the writer and director Perce Pearce told a studio audience in 1939, to be really “coming out of that character. It wasn’t just some funny dialogue that some dummy was rendering.”70
Live-action filming for the dwarfs petered out quickly, though, until by July 1937 Don Graham was speaking of it as an abandoned effort.71 Too much of what the dwarfs were supposed to do did not lend itself to live-action preparation. There was to be no easy answer to the quandary Disney had created when he chose not to cast by character.
Ultimately, the story of the production of Snow White is not a story of how Disney’s men realized his conception of the film, but of how Disney himself recovered from such potentially fatal mistakes and wound up making a much better film than the one he had set out to make. In the fall of 1935, stenographers began making not just summaries or paraphrases but word-for-word transcripts of what was said in many of the meetings in which Disney took part. There is thus a remarkably comprehensive record of how Snow White was made—a record, in effect, of the ebb and flow of Walt Disney’s thought, since everyone working on the film was responding to his wishes.
As Disney submerged himself in his film, scrutinizing the story over and over again in one meeting after another, he gradually surrendered the idea that he was making a gagged-up film centered on the dwarfs. As late as an October 19, 1936, meeting, he was still thinking of the queen’s transformation into a hag as a semicomic scene (“she could holler for her wart, then as the wart appears she would cackle”), but retrogression of that kind occurred less and less often over the coming months.
Disney had to decide what to do with three whole sequences dominated by the dwarfs and what was supposed to be comic business: the dwarfs would sing as they ate Snow White’s soup at the dinner table; they would hold a “lodge meeting” and decide to build her a bed; another comic sequence would show how they did it.
Disney conceived of his film as a Hollywood product. When he “talked” the general continuity at a meeting with more than two dozen members of his staff on December 22, 1936—it was the first time many of them had heard the story of the film as a whole—he invoked other Hollywood films: “This mirror is draped with curtains, like Dracula. . . . The Queen says a little hocus-pocus, and the mirror appears—sort of a Chandu thing.”72 His audiences expected comedy from him. The thought of abandoning any of those sequences and making the film’s tone more serious could not have been pleasant.
There had already been expressions of discontent, even as the problematic sequences were polished and made ready for animation. The writer Dick Creedon, in a November 15, 1936, memo, made a powerful case for dropping the lodge-meeting and bed-building sequences, arguing that they would divert attention from the critically important encounter between Snow White and the queen (in her guise as an old peddler woman, offering Snow White the poisoned apple) without providing any compensating “entertainment value.” There is no evidence that Disney took Creedon’s objections seriously.73
The soup sequence in particular was redundant, in many ways simply echoing the sequence that preceded it, when the dwarfs washed for dinner. Snow White would try to teach the dwarfs how to eat soup, for one thing, and as Dave Hand remarked at a story meeting in November 1936, the pattern of her speech was “too similar to Doc’s starting the group into washing.”74 Likewise, Dopey—who swallowed a bar of soap in the washing sequence—would swallow a spoon at the table, Grumpy would be surly and come last to the table in the same way he resisted washing, and so on.
The lodge-meeting and bed-building sequences suffered from similar defects, but the bed-building sequence was, besides, atavistic in the compressed, artificial construction envisioned for the bed. When, in a February 23, 1937, story meeting, Ham Luske remarked, “It’s darn near a Santa’s Workshop”— referring to the 1932 Silly Symphony, full of mechanical toys moving in tight synchronization with cheerful music—the writer Otto Englander replied: “That’s what we’re trying to get.”75 The artificiality of the sequence bothered even the layout men who were designing the bed itself: The bedposts were supposed to be four growing trees, but were those trees growing in a perfect rectangle by plan or by accident?76
By the summer of 1937, all three sequences had been wholly or partly animated, but Disney decided to scrap them anyway. By then, he knew that he could make those cuts without any damage to Snow White itself because what was left in the film was so concentrated. There was no need to tell more about the dwarfs through the soup-eating and bed-building sequences, in particular, because so much about them would have already been revealed in earlier sequences.
To achieve that result, Disney had no choice but to scrutinize the work of his animators intently in the sweatbox, ordering changes that were in many cases extraordinarily subtle. Such orders were not nitpicking. It is almost always clear from the sweatbox notes that Disney was asking for changes so that a character better conformed to his conception of that character as it was evolving in his work on the story.
On March 6, 1937, for example, he watched a Dick Lundy scene in which Happy approaches the kettle where Snow White has started soup cooking; Grumpy stops him, warning of poison. “The feeling was not that Happy was going to taste the soup,” the sweatbox notes said, paraphrasing Disney’s comments, “but that he expected Grumpy to interrupt him. This of course is not right. He should keep right on going as though he doesn’t know Grumpy is coming.” Happy was the fattest of the dwarfs; he would not have been easily distracted from the soup.
Moore and Tytla, who had first animated the dwarfs, were not immune from such detailed inspection of their work. On June 11, 1937, Disney sweat-boxed a Tytla scene in which Grumpy takes offense and sticks out his tongue at Snow White. Disney asked for these changes: “Have Grumpy make his reaction . . . a few frames earlier and have him react a little slower. Don’t have [the] reaction so extreme—it would be just sort of a stiffening before he turns around, it is sort of a little take—not violent.”
The animators themselves caught the spirit of what Disney was doing. Lundy recalled animating “a walk that I think was the best walk I ever did; but when I got a test on it, it wasn’t Happy. It was drawn and looked like Happy—but it wasn’t the way Happy walked, so I had to throw it away and redo it, so it would be the way Happy walked. I had everything working . . . twist, and overlap, and all that sort of thing. But it wasn’t Happy, so I just had to toss it. His personality wasn’t there.”77
As work on the film progressed, Disney became ever more absorbed in his characters and the story. Robert Stokes, one of the animators of the girl Snow White, spoke of observing him: “I can remember nights when I w
orked a little bit of overtime, say, and he’d come in and pull up a chair and we’d talk . . . until eleven o’clock, just his views on things. Animation, the character, the type of person this character was—he believed that this character was a live person, and he had a way of instilling that in you. . . . I’d hear him padding around in the various rooms, maybe run a Moviola or flip a few drawings and then go on to the next room.”78
Dick Huemer remembered Disney’s “utter dedication” during work on Snow White: “He used to come on like a madman, hair hanging down, perspiring . . . Christ, he was involved.”79
Wilfred Jackson, who moved over from short subjects to direct part of Snow White in 1937, said: “There is more of Walt Disney himself in that particular picture than in any other picture he made after the very first Mickeys. There wasn’t anything about that picture—any character, any background, any scene, anything in it—that Walt wasn’t right in, right up to the hilt. . . . I mean literally that he had his finger in every detail of that picture, including each line of dialogue, the appearance of each character, the animation that was in each scene . . . nothing was okayed except eventually through his having seen it.”80
It was not just the animation of the dwarfs that caused headaches as production of Snow White spread beyond the small group that had worked on the film through much of 1936. On the shorts, a single layout man and a single background painter typically handled an entire film, assuring a consistency of treatment; but Snow White would, of necessity, be spread among dozens of artists. Here again it fell to Dave Hand to try to fit everyone into a single harness. As the layout artists for different sequences bumped against one another, it was all too easy to miss an opportunity to make what was on the screen seem more real. “There must have been at least fifty or sixty corners in the main room of the Dwarfs’ house,” the layout artist Tom Codrick lamented, “because different units were working on the same room and had basic thoughts about what the room was like or the shape of it.”81
The Animated Man Page 19