Short Takes: Mise-en-Scène: Dorothy’s Pigtails, or Integrating Imperfection into Meaning

“George Lucas, engaged in an endless quest for perfect, seamless fantasy, hopes to erase any trace of the tricks necessary to build it. Oz instead understands that the evidence of artifice can itself bear meaning: in this case, that our wildest, most wish-fulfilling fantasies will still reflect our limitations.”

Film buffs are familiar, surely, with what the Internet Movie Database calls “revealing mistakes”: evidence, frequently located in the mise-en-scène, of a fiction film’s artificiality. Typically these mistakes include glimpses of the crew or their equipment and/or inconsistencies or anachronisms in performance, costume, makeup, props, and locations; over time, certain examples have become rather famous — Sonny’s shattered-then-intact windshield in The Godfather (1972), the background extra who covers his ears before a surprise gunshot in North By Northwest (1959) — while others are perhaps known only to the obsessed. Regardless of whether they are noticed or not, the complexity of most filmmaking virtually guarantees that such mistakes will occur and that they will be impractical to correct, like typos chiseled into stone.

Advances in post-production technology make it increasingly feasible for filmmakers to address these mistakes. Moreover, as the controversy over George Lucas’s liberal hand with different editions of Star Wars (1977), The Empire Strikes Back (1980), and Return of the Jedi (1983) demonstrates, this technology highlights the ambiguity of the meaning of “mistake” in this context. The proprietary interest taken by fans debating the ethics of Lucas’s decisions — and the tenor and substance of the discord around whether Greedo or Han Solo shoots first in Star Wars suggests that “ethics” is not too strong a word — is a subject all by itself (for a primer, see Alexandre O. Philippe’s 2010 documentary The People vs. George Lucas). My interest here is simply to note how Lucas’s changes illustrate the difficulty of knowing what counts as a continuity error. For instance, later editions of these films correct some shots in which the original image was flipped horizontally, a device employed during editing to maintain the axis of action, but also a device that produces mirror-image effects such as background figures or military insignia shifting from one side of the screen to another, as in the examples below from the 1980 version of The Empire Strikes Back:

The Empire Strikes Back. 5:04: The gray-haired man and the control panel with vacuum tubes appear on Leia’s right.

The Empire Strikes Back. 519: Slightly later, a shot photographed from the same angle is flipped in order to make it appear that Leia watches Han as he moves through the room. As a consequence, the gray-haired man and the control panel with vacuum tubes now appear on Leia’s left. This shot remains unchanged in the 1997 and 2004 versions.

The Empire Strikes Back. 1983: 1:55:42: Though Empire military insignia are worn on the left, a flip shot briefly causes the insignia to appear on the right side of the uniforms.

The Empire Strikes Back. 2004: 1:57:15: The shot has been changed so that the military insignia now appear on the left side of the uniforms.

I do not think that many voices have been raised in protest over this sort of alteration; in fact, usually the “fan edits” of these films also fix these flip shots. But (to choose a more contentious example) how different, really, is the replacement of Sebastian Shaw with Hayden Christensen in the final scene of Return of the Jedi? Inserting Christensen certainly improves the overall narrative’s continuity, as does the expansion of the climactic celebratory sequence so that it spans the major locations of all six films (the original release shows revelry only in the Ewok village; the 1997 version adds shots of Bespin, Tatooine, and Coruscant; the 2004 version incorporates Naboo as well). And if we grant that these changes are plausibly “corrections” that bolster the coherence of the total fiction, can we say otherwise about Lucas’s overhaul of the musical number in Jabba the Hutt’s palace, which since 1997 features a different song and reconstructs the band with CGI?1 At what point, exactly, should we draw the line between the elements of a film we view as cosmetic mistakes (perhaps to be silently corrected) and the elements we view as fundamental to its meaning?

Return of the Jedi. 1983: 2:06:12: Sebastian Shaw as Anakin Skywalker.

Return of the Jedi. 2004: 2:07:43: Hayden Christensen as Anakin Skywalker.

Return of the Jedi. 1983: 12:22: A modest trio performs for Jabba the Hutt.

Return of the Jedi. 1997/2004: 12:32: An expanded ensemble plays a different song.

Return of the Jedi. 1983: 12:27: the puppet lead singer; the microphone conceals a wire used to control its mouth.

Return of the Jedi. 1997/2004: 13:03: The computer-animated lead singer; no microphone required.

Lucas has every right to revise his work, just as Whitman continually revised Leaves of Grass (he published nine different editions between 1855 and 1892). As those following the Star Wars contretemps know, the bone of contention is Lucas’s disinclination to issue proper editions of the original versions. Clearly, though, for Lucas the most current versions are the proper editions, and the earlier releases are essentially incomplete drafts. Whitman thought of Leaves of Grass much the same way, calling it a project that he had been “working on at great intervals and partially issu[ing]” over several decades preceding the 1892 edition, which, he announced, “absolutely” superseded all previous versions.2 By comparison, the substitution of walkie-talkies for guns in the hands of federal agents in the 2002 version of E.T. the Extra-Terrestrial (1982) drew less controversy than Lucas’s changes precisely because Spielberg packaged the two versions together. Thus we are invited to read the two E.T.s as different artworks (or perhaps to read the 2002 revision as indicative of developments in Spielberg’s artistic judgment), just as we can examine the different versions of Whitman’s poems and understand each iteration as reflecting a discrete point in his career. Lucas, however, strikes a more aggressive stance: imagine if Whitman, in addition to expressing his preference for the 1892 edition, sought to suppress all previous editions of Leaves of Grass (as Lucas has done with the 1978 made-for-television Star Wars “Holiday Special”). While speaking about his changes to the band in Jabba’s palace on the commentary track, Lucas makes his principle clear:

I’m still amused by people that somehow think that when you use cyber technology or digital technology in movies, suddenly it’s fake. But when you look at a scene in here like Jabba’s palace, now there’s some digital characters in there, but they’re no more or less fake than all the other characters that are in here. I mean, is a digital character more fake than a big fat rubber character? [Laughs.] I mean, there’s nothing real here at all, and it’s hard to say that, you know, a rubber character has more integrity than a digital character. What I try to do is just make the characters become believable, so they look realistic enough to where you have a suspension of disbelief and accept them as characters, not as tricks, which is what they all are. (14:30-15:10)

In contrast, one online petition accuses Lucas of demonstrating “disrespect for film history” because, while working on the 1997 edition, the original negatives were permanently altered, making a proper release of the original versions all-but-impossible.3 What Lucas has done might seem like an attempt to rewrite history, though whether we mean by that cinema-production history or the fictional history of the Star Wars universe is not always clear. Indeed, the essence of the disconnection between Lucas and fans upset by his changes lies in a disagreement over the relative values of cinema-production history and fictional history.

The documentary nature of the medium renders the problem of separating these two kinds of history acute when we examine film. As George Wilson puts it, “Our discourse about movies wavers between reference to shots which are of the cast and their performances and reference to shots said to be of the characters and the fictional actions they perform.”4 That is, in both casual conversation and critical analysis, we shift between discussions of a film’s narrative content and of the material evidence of its production. Sometimes we speak in these registers simultaneously: we may adduce the meaning of a number choreographed by Busby Berkeley strictly in terms of character and plot, or we may consider it as an outpouring of sheer technical brilliance, but typically we say that the shot is both part of a constructed narrative and evidence of non-narrative events (a record of real people engaged in real acts). As the Star Wars situation demonstrates, differentiating these two elements can be a vexing problem. On the other hand, I want to explore here an example of how this problem can sometimes improve our understanding of a film: observe, throughout The Wizard of Oz (1939), Dorothy’s pigtails, which repeatedly change length. We might pass this over as no more than a symptom of filmmaking’s exigencies; indeed, Judy Garland’s wig changed several times during the protracted production (when principal photography began, Dorothy was blonde; George Cukor, the third of the film’s five directors, made significant revisions to her appearance).5 Actually, probably because of reshoots, Dorothy’s hair sometimes changes within the same scene, as in the two instances below:

The Wizard of Oz: 36:17

The Wizard of Oz: 36:57

The Wizard of Oz: 1:17:44

The Wizard of Oz: 1:18:01

On other occasions, a more common type of continuity error arises involving the placement of Dorothy’s hair in shots from different angles, as in the scene below, in which the long shots consistently show her hair in front of her shoulders, while the medium shots show her hair brushed back:

The Wizard of Oz: 1:31:25

The Wizard of Oz: 1:31:26

But is there anything that decisively bars us from providing a narrative explanation for Dorothy’s pigtails, especially considering the many other unrealistic events shown (and accounted for) by this film that, after all, chiefly depicts a dream? If we suspect that reshoots created the variability of Dorothy’s hair, perhaps we should also suspect that any apparent continuity errors are not errors at all, since clearly MGM was willing and able to redo scenes repeatedly. (Shutter Island (2010) has several instances of ostensible continuity errors that are clues that portions of the film show a character’s fantasy.) Furthermore, the “horse of a different color” transformation (1:00) — certainly no continuity error — explicitly establishes on the narrative level that, in Oz, hair can change in the blink of an eye.

Worrying about Dorothy’s pigtails might seem nonsensical, but if we accept Oz‘s flying monkeys and talking trees, we have entered a nonsensical world. We may believe that Dorothy’s pigtails are a continuity error, but given the narrative’s hallucinatory, ludic nature, we will struggle to prove it. We should not be surprised that separating Oz from Oz (so to speak) is futile, though, since its basic vocabulary assumes the mutually causal, two-way relationship between fantasy and reality that is a hallmark of the MGM musical. This is clear at the film’s conclusion, which, as Todd S. Gilman observes, omits rather than resolves the problem that propels the film — Miss Gulch’s legal authorization to seize Toto and have him destroyed — leaving us to suppose that she “will, no doubt, soon be back for Toto – waterproofed, armed with a better basket, and possibly accompanied by the sheriff.”6 Though Gilman describes this as a “deferr[al] for the sake of closure” (166), a more apt reading, I think, embraces the convention that Dorothy’s dream-vanquishing the Wicked Witch of the West also reverberates in Kansas (this device reaches its apotheosis in the 1951 An American in Paris, which similarly offers a fantasy sequence that reverses the direction of the real-world plot). Again, objecting to the nonsensical idea that a dream can change reality misses the key point that the film’s subject is a nonsensical world.

By “nonsensical world” I mean not Oz, but rather Kansas as Dorothy experiences it. For example, the film elegantly registers the chaotic absurdity of Dorothy’s Kansas through the plot point that Miss Gulch is permitted to seize Toto herself — an outrageous idea that underscores Dorothy’s nightmarish helplessness. More disturbingly, Aunt Em and Uncle Henry are impotent too: Miss Gulch’s uncheckable and corrupt cruelty, paired with Aunt Em’s inability even to find consolation in calling it by its true name, demonstrate to Dorothy that, regardless of justice or merit, some people remain powerless all their lives. It is little wonder that soon after she decides to run away. As Salman Rushdie writes, “The Wizard of Oz is a film whose driving force is the inadequacy of adults, even of good adults, and how the weakness of grown-ups forces children to take control of their own destinies, and so ironically, to grow up themselves.”7 Indeed, Dorothy’s fantasy revolves around sequences that reinscribe her sense of herself as an irrelevant non-person (she introduces herself as “Dorothy the small and meek” (1:10)) while also (as if against her will, though this reveals only that she fears to claim her true desire) insisting that she possesses power and status. The means by which she fulfills her journey and leaves Oz — a confident, simple declaration of her will was all it ever took — confirms her maturity, her attainment of personhood, which she defines in “Over the Rainbow” as the condition in which “the dreams that you dare to dream / Really do come true” (6:00).

This explains why Rushdie finds the film’s conclusion distressing, since it dismisses Dorothy’s maturation as delusionary and relegates her back to childhood: “How does it come about,” he protests, that “at the close of this radical and enabling film, which teaches us in the least didactic way possible to build on what we have, to make the best of ourselves, that we are given this conservative little homily? Are we to believe that Dorothy has learned no more on her journey than that she didn’t need to make such a journey in the first place? Must we accept that she now accepts the limitations of her home life, and agrees that the things she doesn’t have there are no loss to her?” (56-7). I hasten to remark, however, that Dorothy has a child’s idea of adulthood; one might respond to Rushdie that learning to accept the imperfections of our lives is a better marker of maturity than longing for a place where “troubles melt like lemon drops.” This is what noticing Dorothy’s pigtails can teach us: they signify Oz‘s integration of imperfection into its meaning. I am not suggesting that Dorothy’s pigtails are an intentional error (an oxymoronic term); instead I am arguing that the film operates in a mode in which evidence revealing the material circumstances of its production can function as part of the artwork. Dorothy’s pigtails must be discussed both via the film’s narrative history and its production history because how Oz was made is fundamental to its meaning. Lucas, engaged in an endless quest for perfect, seamless fantasy, hopes to erase any trace of the tricks necessary to build it. Oz instead understands that the evidence of artifice can itself bear meaning: in this case, that our wildest, most wish-fulfilling fantasies will still reflect our limitations. Production history and Oz history become indistinguishable from each other because the film’s fundamental message is not that our dreams come true, but rather that our truths become dreams.