The history of colour cinematography is entangled with representation of racial difference. As Kirsty Sinclair Dootson notes, ‘Even the briefest survey of landmark films made by American market-leader Technicolor evidences that […] people of colour were routinely exploited as part of the system’s chromatic appeals: from the Orientalist fantasy used to debut its two-colour system Toll of the Sea (1922), and the “Mexican” musical-short that launched its three-strip process La Cucaracha (1934), to the notoriously racist feature that secured the firm’s market dominance in classical Hollywood (that has come under renewed scrutiny of late) – Gone with the Wind (1939)’.1 What Dootson calls the ‘intimate connections between the politics of colour-as-hue and the politics of colour-as-race’ have not been lost on film and media scholars, but the connections when schematized can often be automatic and straightforward. The deeply rooted conflation of race and skin tone promotes the assumption that the development of colour cinematography leads to better representations of racial differences qua different hues of the skin. In her discussion of the 1962 Technicolor film My Geisha (Jack Cardiff), for example, Gina Marchetti turns to the ‘tragic interracial romance’, The Toll of the Sea (Chester M. Franklin, 1922), as an early example in which ‘color cinematography accentuates the difference in skin tones between the Anna May Wong character and her Caucasian husband, his wife and her Caucasian-looking son’.2 Likewise, John Belton notes that
Technicolor’s initial efforts at both two-colour and three-colour motion pictures displayed its ability to capture a variety of skin tones, ranging from the Chinese-American colouring of Anna May Wong in The Toll of the Sea (1922) to the Latin look of Stefi Duna [sic] in La Cucaracha (1934).3
These repeated claims regarding Anna May Wong’s skin tone in her 1922 vehicle are both intriguing and suspect. Tracing the technological history of early two-colour processes and the reception history of The Toll of the Sea, the first feature film using Technicolor’s two-colour subtractive process, I found that the alleged difference in skin tone is by no means accentuated, and there is no such thing as the so-called ‘Chinese-American colouring of Anna May Wong’, at least not in terms of the colour of her skin. In fact one contemporary reviewer was particularly disturbed by the lack of such a difference. Unmoved by the fad of colour cinematography, the founder of Variety, Sime Silverman, insists in a review of The Toll of the Sea that ‘Nothing in a moving picture story can rise superior to the story. Coloring never will, never has, and doesn’t here.’ He continues:
The coloring runs without streaks, the camera catching the natural colors aparently [sic], although what seemed something of a freak in this process is that the pallid color given to the complexion of the Chinese extended to the faces of the Americans as well. Perhaps white cannot be taken by this camera with its pallid shade enveloping all faces, white being open to question as a color or for coloring in specific connection. But it was a noticeable defect in the coloring scheme.4
What exactly is the difference between pallid and white? And how are we to understand the ‘coloring scheme’ of two-colour Technicolor and the roles race and racialization play in the development of natural colour cinematography?
This essay interrogates the relationship between colour-as-hue and colour-as-race by focusing on representations of East Asian people – the so-called ‘yellow’ race – in early natural colour cinema. As I shall discuss, the ‘natural’ in natural colour cinema is a highly mediated and ideologically loaded construct. Challenging the epidermal determinism that dominates the discussions of race in colour cinematography, I show that the racial difference in this variant on the Madame Butterfly story is located not in the characters’ different skin tones but in the bright yet ‘naturalistic’ colours of the exotic mise-en-scene, including the ornamental gardens, foliage and costumes in which racialized bodies are enclosed. Put bluntly, the skin tone of Asian peoples is not literally yellow. And the colour yellow is largely absent in this early Technicolor film because it cannot be reproduced adequately by the two-colour subtractive methods used in the multitude of orientalist productions in the 1920s and early 1930s, despite the fact that bright yellow was used in trade advertisements to promote films like The Toll of the Sea (figures 1 and 2).5
Motion Picture News, 1923. Media History Digital Library.
Lobby card for Toll of the Sea.
Scholarship on the entanglement between race and colour cinematography generally follows two sets of discourse. The first exposes the privileging of the white skin tones in the development of representational colour technologies: that is, how the ‘reproduction of pleasing flesh tone’ is the single most important goal in the history of natural colour processes and colour cinematography.6 The second analyses the phenomenon of chromophobia, defined by David Batchelor as the ‘loathing of colour’, ‘a fear of contamination and corruption by something that is unknown or appears unknowable’.7 Chromophobia necessitates the ‘purging of colour’ by relegating it to the realm of the ‘superficial, the supplementary, the inessential or the cosmetic’ or by attaching it to something alien and dangerous, the ‘feminine, the oriental, the primitive, the infantile, the vulgar, the queer or the pathological’.8 As such, colour belongs to the category of the ornament that Anne Anlin Cheng powerfully argues as being central to the long history of aesthetic-philosophic debates about the ‘differences between excrescence and essence, surface and interiority, the peripheral and the central, femininity and masculinity’.9 Just as the discourse of racialized femininity is enmeshed in this western philosophic history, the boundary between self and other is drawn in early natural colour films through similar binarisms: the pitting of the natural against unnatural reproduction; the tension between displaying and containing colour; the indeterminacy between realism and spectacle; and the oscillation between the illusion of stereoscope and the tactility of flatness.10
The theory of ornamentalism provocatively proposed by Cheng urges us to go beyond colour as the biological index of race and to see it as the artifice and technology through which a human figure such as Lotus Flower, played by Wong in The Toll of the Sea, is racialized. A ‘theory of being’, ornamentalism reveals the ontology of Asiatic femininity – the yellow woman – as that of perihumanity: ‘a peculiar state of being produced out of the fusion between “thingliness” and “personness”’, a state of racial difference not marked by the colour of the skin or the ‘organic flesh’, but by the various ways in which the figure is aestheticized and reified.11 As Cheng states, we cannot ‘talk about yellow female flesh without also engaging a history of material-aesthetic productions’, since the ‘yellow woman’s history is entwined with the production and fates of silk, ceramics, celluloid, machinery, and other forms of animated objectness’.12 In this essay I examine the entanglement of the ornamentalist yellow woman and the material-aesthetic history of early natural colour cinematography through a case study of The Toll of the Sea.
Although this landmark film appears in almost every account of early Technicolor, the film’s flagrant interplay between the then-new two-colour subtractive method and the logic of orientalist display has never been fully addressed. I therefore limit my focus to Technicolor processes before the advent of the company’s trademark dye-transfer or imbibition technology, which formed, in combination with the full-colour prism and printing systems, the technological base of Technicolor’s dominance since the 1930s. Given the proliferation of scholarship in recent years on the history of colour cinema in general and the technological development of Technicolor in particular, my account of the technological history up to the release of The Toll of the Sea focuses on two aspects of natural colour process that are most relevant to understanding the racial ideology implicated in the aesthetics of the film: the problem of fringing and the limited range of spectrum that could be reproduced by a two-colour process.
According to Paul Read, the term ‘natural colour’ was coined by film technologists and innovators such as William van Doren Kelley in the early 20th century to describe films ‘photographed so that the colors are selected entirely by optical and mechanical means and reproduced again in a like manner’ and to distinguish these from ‘films arbitrarily colored with dyes’.13 Nothing in the so-called ‘natural colour process’ is completely natural, direct and indexical, as Brian Winston famously argues. Instead of directly recording all the colour information (the impractical technology of the now-obsolete ‘direct methods’ that capture colours with ‘all the information of the chromaticities and intensities’), both additive and subtractive natural colour processes are predicated upon ‘choice and decision’ made during the development of these technologies.14 A case in point would be the selection of colour filters used for ‘satisfactory colour separation’ in Technicolor No. II. Complementary green and red-orange were chosen by Technicolor scientists in order to transmit the widest possible spectrum, but the rendition of specific hues had to be carefully manipulated, with some colours prioritized over others. Physicist and psychologist Leonard Troland, who was elected president of the Optical Society of America and chaired the committee on colorimetry from 1922 to 1923,15 reported to Daniel Frost Comstock at Technicolor in 1923, in the form of patent notes, that ‘the vital requisites regarding colour saturation […] are the following’:
A – Flesh-tint must be rendered sufficiently red-orange.
B – Foliage must be shown sufficiently green.
C – Skies should not be too green; probably the more nearly neutral they can be made the better they will appear.16
Eastman A. Weaver and Troland’s spectro-photometric study suggests that the spectra of flesh-tint and foliage are
of such a complementary character (each being weak approximately where the other is strong) that it looks as if a pair of filters could be made which would not only distinguish these two tints strongly from each other, but also distinguish each one separately from white, thus satisfying conditions A&B above.17
They continue by specifying the exact dyes – Acid Rhodamine and Filter Yellow K for the red and Erio Green G (Schultz #564) Uranine and Eastman Yellow for the green – that would ‘make a fair approach to the ideal’, despite the fact that ‘with only the imperfect dyes available commercially, sharply defined transmissions such as the above can only be obtained approximately’.18 As James Layton and David Pierce conclude, ‘In making such decisions, flesh tones and foliage were prioritized over the blues of sky and sea’.19 Comstock also recalled that in the development of two-colour processes, ‘Other less important colors were sufficiently well rendered as to be generally acceptable until the color sense of the public had had time to become far more critical’.20 However, the colour sense of the public was trained precisely by those who commercialized colour, often intentionally, as will become clear in my discussion of colour consciousness.21
It is thus important to analyse the two-colour aesthetics in its own right, rather than treating it as an initial failure in the progress towards the telos of three-colour subtractive and imbibition dye transfer methods that constitute the ‘glorious Technicolor’ aesthetics we know today. Significantly, two-colour films rely heavily on the convention of orientalist display that allows and justifies the seemingly paradoxical coexistence of naturalness and ornamentality. In fact ornamentalism constitutes and reinforces the two-colour aesthetics’ claim to naturalness. Initially a compromise of William Friese-Greene’s three-colour systems developed at the turn of the 20th century, two-colour systems were based on George Albert Smith’s discovery that using only two primary colours could provide adequate results while significantly reducing the projection speed and amount of film stock required for three-colour systems. In 1902, Charles Urban acquired the copyright of the system developed in Frederic Eugene Ives’s London laboratory by Frederick Marshall Lee and Edward Raymond Turner, and exploited it commercially by founding the Natural Colour Kinematograph Company and branding the system as Kinemacolor when it was released to the public in 1909.22 ‘Up to the First World War’, Luke McKernan shows, ‘it was the only natural motion picture colour process most audiences could see’.23
A crucial precedent of Technicolor No. II, the history of Kinemacolor is closely associated with exotic spectacles and ornamentalism, especially with the filming and exhibition of the 1911 Delhi Durbar that amounts to the singular natural colour reproduction of one of the most extravagant events in early 20th-century world history.24 Having made a reputation with natural colour royal events such as the funeral of King Edward VII and the coronation of King George V, Kinemacolor’s triumph with the Durbar films, despite the technical imperfections that would lead to the company’s sudden demise, is hardly surprising. After all, colour was deemed necessary in order to embalm and recreate these once-in-a-lifetime, awe-inspiring events. In the 1902 Durbar, when filming was conducted on an ad hoc basis and by amateur cameramen due to the novelty of cinematography, people were already doubting the medium’s capacity to record such glorious reality in black-and-white.25 Small wonder, then, that when Kinemacolor exhibited its natural colour Durbar films in February 1912, with music specifically composed for a full orchestra, with bagpipes, fifes, drum corps and chorus, the audience who had seen the Durbar news films in the previous month still flocked to the Scala theatre to experience the full programme, titled With Our King and Queen Through India. Even the Scala stage was transformed into a scale model of the Taj Mahal.26 Unlike other Durbar films that were marketed, produced and exhibited as news, for which timeliness is key, the two-and-a-half-hour With Our King and Queen Through India was programmed and staged as a feature attraction, at a time when even the longest films only ran for 60 minutes. While colour cinematography and the prestigious programming certainly enhanced the spectacle of the Durbar and the aura of royalty, the ornamentalism of the exotic spectacles and the presence of royal family members likewise distracted the audience from noticing the rather limited colour spectrum reproduced by the two-colour process, as well as the flickering and fringing that resulted from the additive method’s time parallax, with small differences between alternating red and green records.27
Facing similar challenges deemed insurmountable by filmmakers, technicians and investors after years of experimentations and exhibitions, Technicolor benefited significantly from a feasible and flexible financial and organizational model that kept bringing in resources – both the money and the talents – that sustained the decades-long process of problem solving.28 Unlike the many natural colour processes developed and exploited commercially from 1911 to 1928 that gave rise to what Paolo Cherchi Usai calls ‘a period of aggressive competition’, when each process ‘in turn emerged, enjoyed an ephemeral public life, and then sank into oblivion’,29 Comstock devised a strategy of ‘progressive step development’. The essence of this strategy, in his words, ‘was to plan it, as far as could be seen ahead, in a series of steps, each step not requiring too much money or time and, also, at the end of each step to show convincing pictures on screen’.30
The Toll of the Sea was the second of these ‘convincing pictures’ shown on screen. For several years after the completion of The Gulf Between in 1917, a film shot in Florida in order to test out the additive two-colour process Technicolor No. I, scientists and engineers of Technicolor were devoted to creating a subtractive method that sufficiently avoided the problem of fringing and did not require a special projector for screening. Achieving these two goals demanded the maintenance of ‘accurate registration throughout every stage of the process without manual intervention’, so that the Technicolor process could ‘become a practical option for commercial film production’ where ‘time-consuming human adjustments had to be minimized’.31 The resulting Technicolor No. II process contained a new subtractive camera and a printing method that could print two colour records onto one single filmstrip. Engineer and cameraman Joseph Arthur Ball was tasked with overseeing the design of the new camera (figures 3–4). His design renovated the additive camera’s beam-splitting prism system with a semi-mirrored surface that split light equally, separating colours to be registered symmetrically on a single negative, with the two vertically inverted colour frames placed one above the other (figures 5–6). The camera was to be built from scratch: the new prism system required extra space in the camera and the mechanism meant that two frames, instead of one, were to be moved between each exposure.32 This new subtractive camera was therefore bulky and hard to crank.
Technicolor No. II camera. Courtesy of the George Eastman Museum. Technicolor No. II beam-splitting prism, filters and film gate, removed from camera. Courtesy of the George Eastman Museum.
Technicolor negatives of The Toll of the Sea. Credit: UCLA Film & Television Archive. Photographs by Barbara Flueckiger. Timeline of Historical Film Colors, <https://filmcolors. org/>.
Comstock devised ‘a new duplex positive film’ in 1920 that enabled the engineers to put the printed image at the bottom of a single-coated film (‘through the back’) in order to avoid the ‘violent alternations of red and green, as well as of brightness and darkness’ that resulted from the uneven thickness of the emulsion on the Eastman double-coated positive film. Two separate films with component colour records would then be glued together, ‘celluloid to celluloid’.33 A cementing machine was constructed in order to apply cement or adhesive evenly and quickly – ‘preferably within approximately one-tenth of a second’ – before the films were pressed together.34 The ingenious cementing process, however, would produce serious problems for Technicolor over the next few years. When The Toll of the Sea was released, the Technicolor laboratory in Boston was slow to catch up with the demand of prints by enthusiastic exhibitors given the complexity of the process and the limited capacity of the factory. The prints also cost approximately eight times more than a black-and-white feature, which made it too high a risk for major Hollywood studios to adopt the colour system for their productions, notwithstanding the commercial success of The Toll of the Sea (the film grossed more than $250,000 and Technicolor received around $165,000).35 Of more importance were the technical problems that still plagued the process. ‘This double-coated film’, Herbert Kalmus recalled, ‘is considerably thicker than ordinary black-and-white film, with emulsion on both sides which tends to make it cup more readily and scratch more noticeably than black-and-white film’. As with the additive method, Kalmus was quick to conclude that ‘this double-coated process was at best but a temporary method, and the work of developing a true imbibition process was being pressed in our research department’.36The Toll of the Sea nevertheless remained a landmark film not only because it represented the technological milestone of a natural colour process that was adaptable to standard projector, but also because it ‘convinced the investors and thousands of other viewers that colour pictures had promise’ – the promise of a pleasing aesthetic and of commercial success.37 Even with the advent of the imbibition printing process in the late 1920s, the double-coated method was still occasionally used, especially for the opening night, for its sharper registration of colours and clearer rendering of human faces.38
Comstock credited the pleasing aesthetic of Technicolor No. II and The Toll of the Sea to ‘the precise split’ of the spectrum, which ‘was so arranged that mixing the two [colours] in the right proportions showed a very pleasing flesh tint’.39 The perception of the ‘flesh tint’ was based on a specific notion of whiteness. For example, Ray Rennahan, the young cameraman hired temporarily in 1921 to run tests for Technicolor in Hollywood and who would be the most celebrated colour cinematographer in the 40 years to come, showed the engineers inexperienced in photography the capabilities of the process by hiring ‘a beautiful redhead model’ as the subject for photographic tests. Using ‘a smooth flesh color make up with eye shadowing and soft lip and cheek rouge’ on the model, against the background of ‘a soft green foliage’ in the garden, Rennahan proved with the correctly lit and diffused close-up shots how pleasing a white female face could be in two-colour cinematography.40 Indeed, ‘feminine beauty’, as Troland asserts in his 1927 essay on the psychological aspects of natural colour cinematography, is ‘Undoubtedly the greatest “kick” of color’. ‘All pretty girls in black and white are pale and consumptive’, Troland proclaims, including ‘Miss Clara Bow [… who] loses entirely her famous auburn colored locks when delineated by the black and white camera’. It is thus reasonable ‘to advocate the cause of colored motion pictures on the ground that color adds to “sex appeal”’, even though he is not sure about ‘to what extent it is moral’.41 In light of this privileging of white femininity, it should be obvious that the ‘sufficiently red-orange’ flesh tints prioritized in the Technicolor colour scheme is designed to ‘best reproduc[e] Caucasian skin tones’, supporting Winston’s claim about the ‘white technology’ of colour cinema.42
The lack of difference in skin tone in The Toll of the Sea problematizes the period’s preoccupation with an idealized representation of white femininity in colour cinematography. If the ‘pallid color given to the complexion of the Chinese [is] extended to the faces of the Americans as well’, to quote from Silverman’s review, then how does the film register racial difference, especially the difference between Wong’s character – the yellow woman par excellence – and the white woman to whom she loses her lover, child and life? Departing from the well-rehearsed argument concerning the representation of skin tones, I shall shift my attention to the other prioritized colour component – the ‘sufficiently green foliage’ – that has been largely overlooked in discussions of racial representation in early natural colour cinema. To understand the role of nature in the history of early natural colour processes and its relationship to orientalist fantasies, I now turn to the film itself.
As screenwriter Frances Marion declared, the story of The Toll of the Sea is ‘practically the step-daughter of Madame Butterfly’.43 It opens as a cautionary tale, with the colourfully illustrated title cards introducing the ‘legend of far-away China which tells of the beauty and the treachery of the siren Sea – whose favors are a mortgage upon the soul – who gives gifts of joy and love, but demands double payment in disappointment and loneliness’. The gift, in this case, is an American, Allen Carver, rescued by Lotus Flower (Wong) from the sea in Hong Kong, a ‘great gift’ that she prayed the sea to give and for which she is willing to be asked of ‘anything in return’. The American marries her in the ‘Chinese fashion’, and then returns to the USA alone after receiving an urgent cable. He subsequently forgets her ‘in his awakened love for the sweet American girl he had known since childhood’, while Lotus Flower pines for his return season after season, giving birth to and raising his child in Hong Kong. When Carver does come back, his American wife urges him to tell Lotus Flower the truth. Painfully realizing that he is indeed not her ‘honorable husband’, Lotus Flower gives the child away to the American couple, telling the loving boy that ‘Lotus Flower only little Chinese nurse for you till sweet mother comes from America [sic]’. Now that the child is ‘safe for always’, ‘in [the] nest of kindly arms’ of Carver’s Caucasian wife, Lotus Flower comes to the turbulent sea for the last time, to pay the ‘great debt’ with her emptied life.
As Sarah Street and Joshua Yumibe have shown, Technicolor exploited orientalist fantasies of western audiences to showcase the new two-colour process, which was best at reproducing reds and greens. Aligning the ‘era of NATURAL color’ with the ‘unquestioning depiction of the East in Orientalist terms’, The Toll of the Sea as a natural colour film valorizes the fantastical vision of the East as ‘realistic’ while using the orientalist ideology to corroborate its claim to ‘realism’. For Street and Yumibe, the film is a ‘double winner’: ‘a “realistic” depiction of China in keeping with prevalent Western imagery of the East and an acceptable narrative about the fate of a Chinese woman who has a sexual relationship with an American man, gives up their child and kills herself’.44 As an early Technicolor feature, the film’s claim to realism is important in terms of David Bordwell, Janet Staiger and Kristin Thompson’s canonical argument that technological and formal innovations ultimately serve narrative continuity, and that the sign of an innovation’s maturity and success is its integration into the narration without drawing excessive attention to itself. In short, The Toll of the Sea could be seen as ‘realist’ because it conforms to the convention of storytelling that depicts the exotic as colourful, the Orient as ornamental.
In his overview of the history of Technicolor, Bordwell unsurprisingly foregrounds the tension between the novelties of colour and the so-called ‘classical paradigm’. He traces Technicolor’s effort, over decades, to fulfil Hollywood’s demands – that is, to make colour ‘only something which is added to the story’.45 One of the principles of ‘classical’ narration is the centrality of characters, which translates, in terms of film style, into the primacy of the representation of human faces. In his chapter on Technicolor, Bordwell quotes the following response from Edgar Selwyn, a Broadway stage producer and the co-founder of Goldwyn Pictures, to Comstock’s question of ‘what the public wants’ around 1920:
The drama is the center of human interest, – not flower gardens and dresses, – the human being is the center of the drama, the face is the center of the human being and the eyes are the center of the face. If a process is not sharp enough to show clearly the whites of a person’s eyes at a reasonable distance, it isn’t any good no matter what it is.46
Bordwell immediately concludes that ‘The plausible rendering of complexion and expression became the chief goal of Technicolor’s research’.47 While the statement is true for a progressive account of the history of Technicolor, for my purposes it is worth noting that more than one reviewer in 1922 identified the subject matter of The Toll of the Sea as particularly suitable for colour cinematography precisely because it offers ‘flower gardens and dresses’ galore.
As Street points out in her study of British early natural colour processes, ‘nature and natural beauty persisted as the “litmus test” of successful colour film exhibition well into the Technicolor period as part of “the cinema of attractions” in non-fiction film well after 1907’.48 Despite his damning verdict of the ‘noticeable defect in the coloring scheme’, Silverman had to admit that ‘the natural colors or the coloring in this Technicolor product is attractive, as it brings out the foliage or strikes the colorful dress of the Chinese’.49 A review published by The New York Times on 27 November also claims that, ‘The producers of the picture have been wise in selecting as their subject a story which lends itself to colored treatment’. It continues,
It is based on an old Chinese legend, it is said, and most of its scenes are in gay gardens and by the sea. Here are settings just waiting for reproduction in colors, and in them are Chinese people whose costumes of elaborate and finely embroidered silks and severely plain cotton permit richness and the effective variety of contrast.50
The ‘gay gardens’, ‘the sea’ and the ‘costumes of elaborate and finely embroidered silks’ are, as both reviews make clear, distinctly ‘Chinese’. And it is these ‘Chinese’ elements that are ‘just waiting for reproduction in colors’. Motion Picture News makes the same point even more explicitly, recommending the film to exhibitors:
‘The Toll of the Sea’ is an excellent subject for color treatment. What land is richer in colors than China? Here we have a Chinese background for a story of an ancient legend. Here are the flowers, the sea, the soft skies in a harmonious arrangement of exquisite shades. Here is the land of the love boat and romance.51
The review continues to emphasize how the film ‘offers in the bargain a story admirably suited for the Technicolor process’ because it makes one forget the ‘environment and [be] transported into a rich make-believe land’. ‘Where is the person’, the author asks rhetorically, ‘who does not love a beautiful flower or a glorious sunset?’52 The critical success of The Toll of the Sea is thus inseparable from the ‘flower gardens and dresses’ that demonstrate most powerfully the novelty of the natural colour process while functioning as a metonym of the exotic and fantastic landscape.
As Craig Clunas argues in his essay on western descriptions of the Chinese garden, there is ‘no nonideological uses of nature’ and there can be ‘no meaningful opposition of nature to ideology at any level’.53 In stark contrast to the 19th-century accounts that view the Chinese gardens as perverting nature and as an expression of the Chinese penchant for grotesque and monstrous artificiality (‘The Chinese deal with rocks, trees, streams, etc. as they do with the feet of their women’), the early 20th century witnessed the revalidation of the Chinese garden ‘as one of the great artifacts of that civilization, precisely on account of its closeness to nature’.54 Rather than equating the ‘unnatural, irrational, and overcomplicated’ garden with the stagnant despotism of the imperial Chinese government (as opposed to the enlightened, liberal governments of Britain and France), the post-imperial China revived ‘an interest in the antiquity of the Chinese garden and its essential invariance over time’.55 From the 1920s onward a modernizing China was met with an orientalist discourse that claims the proximity to nature embodied in the presumably homogeneous art of the Chinese garden as a cultural and racial essence, a ‘continuous thread [that] cannot be provided by anything like conscious choice on the part of individual Chinese actors in specific historic circumstances’.56 Moreover, through claiming the continuity of Chinese culture and rejecting the coexistence of ‘modern’ and ‘China’ in the same proposition, this discourse promotes a notion of authenticity mired in an inherent (that is, ancient, timeless) closeness to and empathy for ‘nature’. In short, ‘That which was authentic had to be ancient’ – and therefore, naturalistic.57
This brief genealogy of western accounts of the Chinese garden in the early 20th century is instrumental to my discussion of the reception of The Toll of the Sea. None of the reviews cited above mistakes the landscape created in Santa Monica and Hollywood for being shot on-location in China. What they recognize is rather an orientalist landscape composed of chinoiserie objects and the apparent naturalness associated with the Chinese garden in a western imagination. In this sense, The Toll of the Sea is not unlike the myriad of Hollywood films from the era that make eclectic use of exotic architectural motifs.58 What is unique in this case is that the coupling of naturalness and authenticity legitimizes – indeed, naturalizes and authenticates – the novelty and ornamentality of colour cinematography. If, at the early stage of natural colour processes, the rendering of a plausible Caucasian complexion signaled the possibility of colour cinema’s integration into the classical paradigm, the attraction of the technological novelty was maintained and showcased through the spectacularization of the exotic, especially by foregrounding the ornaments: the flower gardens and dresses and all that is ancillary to ‘drama’.
In other words, colour, ornaments and the (naturalistic) Orient are aligned in a way that troubles the common association between colour and race as referring primarily, if not exclusively, to the colour of the skin. Instead of the skin tone – ‘the pallid shade enveloping all faces’ – it is the colorful decor, costumes and the colour cinematography itself that are exotic and racially different in The Toll of the Sea.59 Specifically it is the relationship between the Chinese heroine and her environment that marks her as racially distinct from her Caucasian lover and female counterpart. Rather than being foregrounded as a figure that is separated clearly from the background according to the norm of (western) perspectival composition, the aptly named Lotus Flower is part of the natural environment. She is almost always framed beside the flowers and foliage, with the colour of her dresses and her ornamental headpieces blending in with the colour of the environment (figure 7).60 The association between ornamental dresses, natural environment, and Chineseness is most explicitly demonstrated when the previously colourfully dressed bi-racial child (played by a Caucasian girl, Priscilla Moran) is being bathed, her colourful costume literally washed away and replaced with a white sailor-suit: a symbolic transition from Chineseness to whiteness being prerequisite to the child’s acceptance into the white nuclear family.61
Lotus Flower blends with her environment, in Toll of the Sea.
Technicolor’s later insistence on the control of mise-en-scene, with strategies such as colour separation – enshrined in Natalie Kalmus’s seminal essay ‘Color consciousness’ – could apply retrospectively to The Toll of the Sea. ‘One very important phase of making color pictures’, colour separation stipulates that ‘when one color is placed in front of or beside another color, there must be enough difference in their hues to separate one from the other photographically’.62 Kalmus provides an example:
there must be enough difference in the colors of an actor’s face or costume and the walls of the set to make him [sic] stand out from the colors back of him; otherwise, he will blend into the background and become indistinguishable, as does a polar bear in the snow.63
The principle is in line with the tradition of perspectival composition that centres the human subject by creating an illusion of three-dimensional space;64 indeed, such is the aim of ‘good’ colour design. ‘If the colors are properly handled’, Kalmus continues, ‘it is possible to make it appear as though the actors were actually standing there in person, thus creating the illusion of the third dimension’.65
The colour control department led by Natalie Kalmus and the official rhetoric of ‘colour consciousness’ were not yet in place at the early stage of two-colour film productions. Nevertheless, an analysis of the mise-en-scene of The Toll of the Sea will show that although the strong clash between complementary colours such as red and green would give way to low-contrast colour harmony that produces a more restrained colour palette,66 some of the principles of colour consciousness, such as colour separation, were already being practised. For one thing, the Americans in The Toll of the Sea are surely not polar bears in the snow; their beige or grey suits provide obvious contrast to the verdant background, marking them as sovereign (colonial) agents exploring the world. Even when the American wife, Mrs Carver, first comes to Lotus Flower’s garden wearing a light green, gauzy dress,67 she is seen touring around in a landscape with an extensive depth of field that is further emphasized by the random bushes blocking the bottom-right foreground (figure 8). Her behaviour – feeding white ducks and petting an exotic peacock – could also be interpreted as explorative or even intrusive, especially when crosscut with the two-shot of Allen Carver and Lotus Flower, in which the latter painfully realizes that Carver is not her ‘honorable husband’.
Mrs Carver tours the landscape, in Toll of the Sea.
Studying the earliest hand-colouring practice, Yumibe shows how femininity and colour were intertwined in the discourses around coloured dance films, and how the sensuous and erotic quality of the dance can in part be contributed to the three-dimensionality created by applied colour process. ‘Color’, he writes, ‘adds a projective dimensionality to the film image: it creates the sensational impression of stereoscopy, proceeding from the screen toward the viewer’.68 It would be wrong, however, to assume that colour necessarily adds depth to the two-dimensional image. At least one of the two other factors of mise-en-scene needs to work in tandem with colour to create the stereoscopic effect. One is movement, as in the case of the early dance film; the other, more fundamental factor is the aforementioned colour separation, or the distinction between figure and ground. The British period drama The Glorious Adventure (J. Stuart Blackton, 1922), for example, was said to achieve ‘stereoscopic values’ through Prizmacolor mainly because of the mise-en-scene that emphasizes depth of field, for instance ‘when the flames of fire are seen through doorways and windows, appearing to come towards the viewer as the fire gets out of control’.69
In contrast, without the proper colour separation allotted to the white subjects, scenes from The Toll of the Sea that feature Lotus Flower are flattened by colour. She does not navigate a three-dimensional space with the same agency. Rather, as dictated by the mise-en-scene, she is so enclosed in the colourful garden and costumes that she becomes an ornament herself: a ‘body ornament’ that Cheng theorizes as the yellow woman’s state of being, and a fate inscribed in the career of Wong in Hollywood as she was routinely cast in roles designed mainly to add ‘authentic’ oriental flavour to the mise-en-scene.70 Scott Higgins notes the stunning scene in which ‘strong red flowers in the fore- and background complement the green of [Lotus Flower’s] dress in center frame’, and claims that this iconic image is ‘strongly stylized as romantically glamorous’, the composition of which ‘lends it a sense of completeness [that] militates against the feeling that something is missing [in the two-colour cinematography]’.71 (figure 9) The completeness arguably derives from the red and green coverage of the whole screen, compensating for the absence of part of the colour spectrum. The clashing of the two complementary primary colours that cover the screen, moreover, flattens the image into a single plane – a colour palette in which the contour that distinguishes the figure from the ground gets blurred. As such the mise-en-scene ‘equate[s] her with nature, exoticism, and the East’72 by tasking Lotus Flower/Wong – or more specifically, her face – with a function similar to that of the nearby red flowers and the green clothes that define her figure: to act simultaneously as an index and an icon of the authentic Orient.
A flattened colour palette of red and green, in Toll of the Sea.
In this case the Orient is the flattened and the colourful. It is also the static and the captive. The bulkiness of the two-colour camera resulted in stationary shots, which turned out to comply with the notion of the Orient as timeless and motionless. The stillness is further accented with the constant and repetitive framing of Lotus Flower as sitting still, waiting and emoting. She is captive in both senses of the word, and so too is the colour. Chained to the timeless and motionless Orient, the unruly colour is contained, either as the constant background of the three-dimensional space, or as constituting the two-dimensional palette enclosing the demure Oriental face.
There is one mystery around the reception of the film that remains unsolved. To recap Silverman’s comments on the film’s colour scheme: it is ‘something of a freak’ that ‘the pallid color given to the complexion of the Chinese [is] extended to the faces of the Americans as well’. Michael Keevak’s Becoming Yellow: A Short History of Racial Thinking, in which he investigates ‘how East Asians became yellow in the Western imagination’, provides important insight for us to make sense of the insisted difference between ‘pallid’ and ‘white’.
As Keevak states right at the outset of his study, ‘in nearly all the earliest accounts of the region [East Asia], beginning with the narratives of Marco Polo and the missionary friars of the thirteenth century, if the skin color of the inhabitants was mentioned at all it was specifically referred to as white’.73 The idea that Chinese people must be something other than white comes from the history of racialized thinking in the 19th century. In short, East Asians only became yellow when they were ‘lumped together as a yellow race’, that is, when the colour yellow became a racial designation in the 19th century.74 However, as Keevak delineates, even before East Asians became yellow, Europeans in the 17th and 18th centuries already qualified the visible whiteness of East Asians as ‘lacking a lively colour’ (zonder levendige verve), or even as having a deathly pallor (eine Todten-Bleiche). He describes how Francis Bacon, for instance, remarked that the Chinese were ‘of an ill complexion (being olivaster)’. Carl Linnaeus importantly used the term luridus (sallow, pale, ghastly) – the colour of disease in botany and medicine – in place of fuscus, an extremely broad and vague term that could be translated as ‘dark’, ‘brown’ or ‘swarthy’, to classify Homo asiaticus in the definitive tenth edition of his Systema naturae. Instead of other more common or neutral terms, Linnaeus chose luridus, one with such pejorative connotations as horror, ugliness and pallor, similar to the English word ‘lurid’.75 Keevak asserts that this choice was ‘almost certainly linked to a new eighteenth-century sinophobia that saw the Chinese no longer as white, civilized, morally superior, and capable of Christian conversion, but instead as pale yellow, despotic, stagnant, and forever mired in pagan superstition’.76 Silverman’s seemingly casual use of the adjective ‘pallid’ to describe the skin tone of the Chinese in The Toll of the Sea can thus be read in the context of a long lineage of scientific racist discourse, in which theories of colour were entangled with natural science, humoural theory (yellow’s association with jaundice, for instance) and colonialism. In other words, the ‘pallid color’ is not so much an objective hue categorically ‘given’ to the complexion of a pre-existing race of the ‘Chinese’ or ‘Asians’; rather, it functions as a symbolic and even moral attribution mobilized to maintain the historically constructed colour line that does not exist epidermically as much as psychologically.77
This particular anxiety over marking racial difference in the colour of the skin becomes even more intriguing if we consider Silverman’s Jewish identity. Occupying a ‘position of structural instability’ that constitutes the ‘limit case of whiteness’ throughout western history, Jewish people were among the ethnic groups assimilated into the American melting pot during the Progressive Era.78 As Michael Rogin argues in Blackface, White Noise: Jewish Immigrants in the Hollywood Melting Pot, the Americanization of white ethnic immigrants such as the Jewish, Irish, Germans and Italians was predicated upon the exclusion of racial groups such as blacks, Asians and Native Americans.79 In the history of early Technicolor, for example, famous Jewish actor Eddie Cantor performs a multitude of ethnic and racial masquerades in Whoopee! (Thornton Freeland, 1930; shot in Technicolor No. III, a two-colour subtractive, dye-transfer process), impersonating a Greek cook, a blackface minstrel and a Native American while interjecting Jewish banter that continually ruptures the masquerades. The fluidity of racial and ethnic identities embodied by Cantor ultimately serves the surprise ‘happy ending’ of this comedic musical, in which the Native American protagonist Wanenis (Paul Gregory) is revealed to be white, and therefore legitimate to marry the white heroine Sally Morgan (Eleanor Hunt).80 As a result of the reinvention and consolidation of whiteness as a category in the interwar USA, the colour line could be porous for the assimilated whites yet must remain impervious to contamination by the non-whites. The logic is most vividly illustrated in the spectacular two-colour Technicolor revue, King of Jazz (John Murray Anderson, 1930), in which only white, European ethnic groups and cultures – British, Spanish, Russian, French, Austrian and Scottish – and the music they perform are accepted into the literal, gigantic melting pot (figure 10).81
The melting pot, in King of Jazz (1930). Library of Congress.
In Richard Dyer’s words, ‘the notion of whiteness [is] a coalition’; white is not a literal colour when it is used to categorize a group of people, although the representations of white people are importantly recognizable by way of colour.82 The designation of the colour ‘white’ as one’s racial identity depends on the interaction of a set of elements, of which skin tone is just one. Physiognomic features other than flesh tone – such as the shape of nose, eyes and lips, the type and colour of hair, and the build of the body – could all be factors in the historical perception of one’s ‘colour’.83 The valorization of colour as a shorthand for the complex historical process of racialization is crucial to our understanding of the history of colour cinema, for this abstraction of colour as the sole index of race has proved especially inadequate and defective in the development of this representational technology. The fact that white people are not literally white – that is, they cannot be reduced to and directly represented by ‘the colour of snow or bleached linen’84 – meant that tremendous resources were leveraged to research the ‘plausible rendering of [white] complexion and expression’.
The endeavour to achieve such ‘plausibility’ in colour cinematography is even trickier if we consider Dyer’s observation that, in western visual culture, the colour of white as a racial identity depends in fact on the absence of colour. He writes:
Whites must be seen to be white, yet whiteness as race resides in invisible properties and whiteness as power is maintained by being unseen. To be seen as white is to have one’s corporeality registered, yet true whiteness resides in the non-corporeal. The paradox and dynamic of this are expressed in the very choice of white to characterise us. White is both a colour and, at once, not a colour and the sign that which is colourless because it cannot be seen: the soul, the mind, and also emptiness, non-existence and death, all of which is the sign that makes white people visible as white, while simultaneously signifying the true character of white people, which is invisible.85
This revelation of white as a paradoxical colourless colour, the supreme power of which lies in its simultaneous visibility and invisibility, sheds new light on Silverman’s comment on the predicament of colour cinema. In trying to account for that ‘something of a freak’, the ‘defect in the coloring scheme’, he suggests that ‘Perhaps white cannot be taken by this camera with its pallid shade enveloping all faces, white being open to question as a color or for coloring in specific connection’. Not only is white as a hue an indeterminate category: white does not exist on the spectrum of colour but ‘is made up of all colours fused together’.86 More importantly, its intangible, non-corporeal and spiritual dimension cannot be directly captured by colour camera. What is wrong in the representation of whiteness in The Toll of the Sea, according to this reading of the review, is then not merely the absence of racial difference based on skin tone, but the absence of those intangible virtues – beauty, purity, transparency and spirituality – associated with whiteness: an absence that arguably degrades what is supposed to be white to become pallid.
This is, of course, more than a technological defect. If we take into consideration The Toll of the Sea’s story, an aspect most valued by a professional critic like Silverman, it becomes clear that the absence is moreover a misplacement, for the symbolic values associated with whiteness are embodied by Wong’s non-white character, Lotus Flower. Compared to the loving, faithful, sacrificial – in a word, virtuous – Chinese heroine, the American Allen Carver is indeed treacherous, and his Caucasian wife colludes in this treachery. By abandoning Lotus Flower and adopting (or less euphemistically, stealing) her son, they effectively kill her off, while the story excuses their horrendous acts as being the toll of the sea. With or without Marion’s critical intention, the addition of the legend of the treacherous sea as a narrative framing device imposes a layer of self-awareness to the Madame Butterfly master narrative. By placing the blame on a fabricated curse and the implausible villainy of the sea, the plot ironically foregrounds the cruelty of this romance and magnifies the moral ambivalence and social criticism central to melodrama. The opening legend could be read, quite uncannily, as an allegory for colonialism, in which the sea stands for the foreign colonial power that asks for ‘double payment’ in exchange for whatever gifts it promises. In other words, the fatal mistake that Lotus Flower commits is not that she receives the gift from the treacherous sea, but that she mistakes whiteness, and the power it possesses, as a sign of virtue and honour. She is ultimately punished for this blind faith in the white man as well as for her failure to recognize an unsurpassable racial difference. Proudly claiming that she converses ‘only in American language’ and having a traveling costume tailored, as the intertitle card reads, from ‘a most chic American fashion book inherited from her grandmother’ ready for the journey to ‘those United States’, Lotus Flower is under the illusion that interracial romance guarantees cultural and racial assimilation. Only when she meets the legitimate Mrs Carver, the gentle white woman wearing a monochromatic, flowing, elegant dress, does she realize that her Chineseness, foregrounded by the carefully prepared ceremonious, ornamental and colour-clashing bridal robes, automatically delegitimizes her status as the honourable wife. Only then comes the revelation that her veneration for the white man entails the inherent inferiority and subjugation of her self – that she is, indeed, only the ‘little Chinese nurse’, unassimilable and totally dispensable to the white nuclear family.
As a quintessential family melodrama that capitalizes on the taboo against interracial marriage and concludes with a compulsory tragic ending, The Toll of the Sea could be viewed as illustrating Thomas Elsaesser’s classic framework of that genre: the world is enclosed, the characters are acted upon, and the decor and objects literally give ‘expressive colour and chromatic contrast to the story-line’, giving ‘melos’ to ‘drama’.87 In this light the Orient offers a perfect setting – idealized, exotic, enclosed – for the family melodrama, where contradictory emotions and morals can be worked out. While Marchetti poignantly summarizes the ideological appeal of the Madame Butterfly story to the western audience as the legitimization of the West’s abuses and authority through the heroine’s self-sacrifice, this view cancels the moral – and relatedly racial – ambivalence that melodrama supposedly works through by channeling emotional excesses through the equally excessive mise-en-scene.88 Moreover, some scholars have recently challenged the received wisdom of film melodrama as an archaic precedent or anomaly to the so-called ‘classical cinema’. Linda Williams, for instance, argues that melodrama in its ‘very orchestration of feelings and sensations’ was in fact the new norm in the early and mid 20th centuries, ‘the true global matrix for the experience of modernity’.89 What defines melodrama as a mode, according to Williams, is not its inherent excessiveness (a definition which shows ‘the tendency to view melodrama as old and regressive’) but its persistent work of ‘seeking a better justice’.90 She corrects Bordwell’s misreading of Frank Borzage’s statement from 1922: ‘Today in the pictures we have the old melodramatic situations fitted out decently with true characters’. Rather than condemning melodrama as intrinsically unrealistic or excessive and that ‘old melodrama is negated by new classicism’, as Bordwell suggests, Williams reads Borzage as acknowledging and insisting on melodrama’s capacity to grow and adapt ‘in tension with new norms of what seems, or feels, true’.91
Echoing Borzage’s investment in melodrama’s agility, a contemporary reviewer of The Toll of the Sea embraces the Madame Butterfly story as one of the ‘fundamentally true, or truly sentimental’ stories that ‘can stand repeated retelling and become persuasive anew whenever they are well told’.92 What feels true, as this review makes clear, is not (only) the taboo against interracial romance, but also Lotus Flower’s suffering as performed by Wong, who is ‘naturally Chinese and exactly, easily natural, even in her most tormented scenes’. Despite its essentialist emphasis on Wong’s apparent Chineseness, the review expands on the naturalism of Wong’s performance. The writer describes Wong’s character as ‘the trusting child of a carelessly considered race’ and proclaims that ‘Miss Wong stirs in the spectator all the sympathy her part calls for, and she never repels one by an excess of theatrical “feeling”’. Instead of being archaically theatrical and ‘melodramatic’, so to speak, Wong’s performance is cinematic and modern: ‘Completely unconscious of the camera, with a fine sense of proportion and remarkable pantomime accuracy’. And it is for this reason that the review advocates for something that is denied in this film – an opportunity for inclusion – that ‘She should be seen again and often on the screen’.93
Similarly, as I have discussed, the ornate oriental setting of this melodrama was not received as excessive but rather as ‘natural’ and real. Reviewers raved about the ‘uncanny excellency’ of the colour cinematography’s reproduction of ‘actual persons and places’, as if ‘just a little bit of life lifted out of the Orient, framed and sent here to delight the eye and sadden the heart’.94 In addition to the ‘veritable flesh tints of hands and faces’ that make the ‘Figures in the drama […] utterly human’,95 the natural environment, the ‘natural’ (non-yellowface) Chinese and the naturalistic performance, all work to naturalize Technicolor’s new colour process.
In the history of colour cinema, The Toll of the Sea is only one of the first ‘demonstration films’96 that use exoticism and orientalist display to negotiate the long and winding path towards technologically reliable, commercially viable, aesthetically desirable and ideologically acceptable natural colour processes. Even a cursory look at Technicolor’s filmography between 1915 and 1935, carefully compiled and annotated by Layton and Pierce, yields a long list of films set in ‘exotic’ locations such as Egypt, Mexico, the South Seas, the Middle East and even Spain.97 Titles that allude to fantasies of the Far East alone include Japanese Carnival (1928), Manchu Love (Elmer Clifton, 1929, ‘the first American-made picture that boasts an all-Chinese cast’98), In a Chinese Temple Garden (1929), Light of India (Elmer Clifton, 1929), The Japanese Bowl (Jack Haskell, 1930), A Chinese Flower Boat (1931) and The Chinese Nightingale (Rudolf Ising, 1935). In addition, one could point to illustrious scenes that marry orientalism and Technicolor, such as the opening sequence in The King of Kings (Cecil B. DeMille, 1927), in which the Japanese actor Sojin Kamiyama, notorious for his role as the villainous Prince of the Mongols in Raoul Walsh’s The Thief of Bagdad (1924), is costumed in ornamental reds and greens (figure 11). A similar colour palette and ornamentalist spectacle also dominate lesser-known sequences such as the Berkeleyesque musical number ‘A Chinese Fantasy’, featuring Nick Lucas and Myrna Loy and introduced by the canine star Rin-Tin-Tin in the Warner Bros. revue Show of Shows (John G. Adolfi, 1929) (figure 12).
Kamiyama costumed in red and green in King of Kings (1927).
‘A Chinese Fantasy’ in Show of Shows (1929).
Curiously, this colourful fascination with the Far East seems to dissipate by the mid to late 1930s when Technicolor was finally able to reproduce the colour yellow. This may be the result of Technicolor’s transition from a demonstration mode to the restrained design that makes colour just another stylistic choice integrated into the so-called ‘classical’ narration. In other words, the ‘Technicolor look’, defined by Higgins as the ‘quality of design that helped make images feel perfectly organized, correct, and inevitable’, is ultimately at odds with the orientalist display marked by excess, contrast and ornament.99 Furthermore, the aesthetic and technological development of natural colour processes coincided with, and was compounded by, the geopolitical situation of the 1930s, when conflicts in Asia such as the Second Sino–Japanese War created a demand for more ‘realistic’ instead of outlandish portrayals of the Far East, especially that of Republican China.100 In the 1930s, as Karen Leong writes, ‘Americans began to imagine China differently, no longer as an alien and distant culture and land, but as a demonstration of the promise held by American democracy and culture to transform other nations’.101 The convergence between the glorious Technicolor and the colourful Orient would return in the 1950s in the form of what Christina Klein calls a ‘Cold War Orientalism’, which created ‘a global imaginary of integration’.102 Films such as Sayonara (Joshua Logan, 1957), The World of Suzie Wong (Richard Quine, 1960) and My Geisha (Jack Cardiff, 1962), as well as Richard Rogers and Oscar Hammerstein’s oriental musicals South Pacific (Joshua Logan, 1958) and Flower Drum Song (Henry Koster, 1961), continued the practice of early natural colour cinema to exploit and romanticize oriental landscapes and peoples – but with a twist. In the Cold War era, these colourful, ‘oriental’ films buttressed the colonialist and expansionist ideology of fostering ‘understanding’ of and creating coalition with – again through the metonymy of interracial intimacies – the non-communist (or the ‘good’) part of Asia.
While this essay is preoccupied with the relationship between natural colour cinematography and a historical orientalism that conceives Asia as timeless, motionless and close to nature, the era of digital colour seems to reverse this framework as we witness the rise of a techno-orientalism that imagines Asia and Asians ‘in hypo- or hypertechnological terms in cultural productions and political discourse’.103 Nevertheless, the use of ‘oriental’ and racialized bodies in general to naturalize, authenticate and mystify technological transitions still holds true in the advent of digital cinematography, as Alice Maurice shows with the short promotional film Tattoo (Bill Paxton, 2011), made to showcase the new digital (5K) Red Epic camera.104 As the western imagination of the ‘Orient’ continues to change, and exotic settings are reconfigured in relation to technological novelties in different historical conjunctures, the story of the evolving intimacy between colour-as-hue and colour-as-race has remaining chapters to be told.
Footnotes
1 Kirsty Sinclair Dootson, ‘The politics of colour’, Frames Cinema Journal, no. 17 (2020).
2 Gina Marchetti, Romance and the ‘Yellow Peril’: Race, Sex and Discursive Strategies in Hollywood Fiction (Berkeley, CA: University of California Press, 1993), p. 182.
3 John Belton, ‘“Taking the color out of color”: two-colour Technicolor, The Black Pirate, and blackened dyes’, in Giovanna Fossati, Victoria Jackson, Bregt Lameris, Elif Rogen-Kaynakçi, Sarah Street and Joshua Yumibe (eds), The Colour Fantastic: Chromatic Worlds of Silent Cinema (Amsterdam: Amsterdam University Press, 2018), p. 102.
4 Sime Silverman, ‘The Toll of the Sea’, Variety, 1 December 1922, p. 35.
5 James Layton and David Pierce, The Dawn of Technicolor 1915–1935 (Rochester, NY: George Eastman House, 2015), pp. 94–95.
6 Richard Dyer, White, 20th anniversary edn (London: Routledge, 2017), p. 93. See also Brian Winston, ‘A whole technology of dyeing: a note on ideology and the apparatus of the chromatic moving image’, Daedalus, vol. 114, no. 4 (1985), pp. 105–23; Lorna Roth, ‘Looking at Shirley, the ultimate norm: colour balance, image technologies, and cognitive equity’, Canadian Journal of Communication, vol. 34 (2009), pp. 111–36; Genevieve Yue, ‘China Girls in the film laboratory’, Girl Head: Feminism and Film Materiality (New York, NY: Fordham University Press, 2021).
7 David Batchelor, Chromophobia (London: Reaktion Books, 2000), p. 22. See also Michael Taussig, What Color is the Sacred? (Chicago, IL: University of Chicago Press, 2009); Sabine Doran, The Culture of Yellow: or, The Visual Politics of Late Modernity (New York, NY: Bloomsbury, 2013). Examples of scholarship on race and representational colour technologies that builds on the concept of chromophobia include Dyer, White; Kirsty Sinclair Dootson, ‘“The Hollywood powder puff war”: Technicolor cosmetics in the 1930s’, Film History, vol. 28, no. 1 (2016), pp. 107–31; Sarah Street and Joshua Yumibe, Chromatic Modernity: Color, Cinema and Media of the 1920s (New York, NY: Columbia University Press, 2019); Susan Murray, Bright Signals: A History of Color Television (Durham, NC: Duke University Press, 2018).
8 Batchelor, Chromophobia, pp. 22–23.
9 Anne Anlin Cheng, Ornamentalism (New York, NY: Oxford University Press, 2019), p. 15.
10 This dynamic is also explored in Alice Maurice, The Cinema and Its Shadow: Race and Technology in Early Cinema (Minneapolis, MN: University of Minnesota Press, 2013), in which she traces the ‘turn to racialized bodies and the rhetoric of race difference to prop up or “reintroduce” the cinema, especially in times of transition; the effort to authenticate the image through appeals to a racialized essentialism; the use of racial boundaries to figure cinematic boundaries; and the exploitation of race to “embody” the apparatus and to suggest a communal, transformational model of ideal spectatorship’ (pp. 198–99).
11 Cheng, Ornamentalism, pp. 4, 18.
12 Ibid., pp. xi–xii.
13 Paul Read, ‘“Unnatural colours”: an introduction to colouring techniques in silent era movies’, Film History, vol. 21, no. 1 (2009), pp. 9–11.
14 Winston, ‘A whole technology of dyeing’.
15 Sean F. Johnston, ‘The construction of colorimetry by committee’, Science in Context, vol. 9, no. 4 (1996), p. 405.
16 Eastman A. Weaver and Leonard Troland, ‘Patent Notes – Improved Taking Filters’, Subtractive: Camera, Etch Process 1 and 2, Negative, Projection Pr., Varnish, 30 January 1923, pp. 1–2, Technicolor Notebooks Collection, George Eastman House. These requisites echoed with Weaver’s notes on the Prizma’s demonstration of its subtractive system using both single- and double-emulsion film stocks in 1922. Weaver wrote, ‘The single-emulsion reel showed very poor colors indeed. Flesh tints were absolutely white and greens were lacking except for a very pale blue. Reds were fair but unsaturated and loss of detail was excessive.’ The dye-toning double-coated reel was ‘very much better but it had noticeable defects, namely, very considerable loss of detail, entire absence of yellow, blues and grays instead of greens, especially on foliage, and general lack of saturation’. See Weaver, ‘Notes on Prizma Exhibition at Engineers Club’, 27 February 1922, Technicolor Notebooks Collection, George Eastman House.
17 Ibid.
18 Ibid.
19 Layton and Pierce, The Dawn of Technicolor, p. 82.
20 Daniel Frost Comstock, An Outline of the History of the Beginning of the Technicolor Development in Boston (Cambridge, MA: Comstock & Wescott, Engineers, 1961), p. 8, Technicolor Corporate Archive, George Eastman House.
21 See also Chapter 1, ‘The colors of modernity’, in Joshua Yumibe, Moving Color: Early Film, Mass Culture, Modernism (New Brunswick, NJ: Rutgers University Press, 2012).
22 Paolo Cherchi Usai, Silent Cinema: An Introduction (London: British Film Institute, 2000), pp. 28–29.
23 Luke McKernan, ‘“The modern elixir of life”: Kinemacolor, royalty and the Delhi Durbar’, Film History: An International Journal, vol. 21, no. 2 (2009), p. 122.
24 A Mughal word of Persian origin, ‘Durbar’ means a ‘reception, a court, or body of officials at such a court’, and was appropriated by the British raj to describe the 1877 ceremonies that proclaimed Queen Victoria as Empress of India. See McKernan, ‘“The modern elixir of life”’, p. 128. On ornamentalism and the Delhi Durbar, see David Cannadine, Ornamentalism: How the British Saw Their Empire (London: Penguin, 2001). As Cheng mentions, ‘Cannadine argues that class, more than race, provided the lens through which Britain imagined its empire hierarchically’ and hence misses the opportunity to interrogate ‘the intertwining of the Orient in the history of the ornament’. Cheng, Ornamentalism, p. 18. On the long associations of India and chromatic exoticism filtered through the economics of colonialism, see Natasha Eaton, Colour, Art and Empire: Visual Culture and the Nomadism of Representation (London: IB Tauris, 2013).
25 See Stephen Bottomore, ‘“An amazing quarter mile of moving gold, gems and genealogy”: filming India’s 1902/03 Delhi Durbar’, Historical Journal of Film, Radio and Television, vol. 15, no. 4 (1995), p. 499.
26 McKernan, ‘“The modern elixir of life”’, p. 131.
27 Sarah Street, Colour Films in Britain: The Negotiation of Innovation 1900–55 (London: Palgrave Macmillan, 2012), p. 14.
28 In this study I use the term ‘Technicolor’ mostly to refer to the group of people, including engineers, scientists, technicians, investors and staff, who contributed to the technological development, commercial success, aesthetics and ideology of colour cinematography under the brand name of Technicolor. As such, the label of ‘Technicolor’ both overlaps with and is distinct from the company that is called the Technicolor Motion Picture Corporation. It is important to note that Technicolor was established in late 1914 in order to focus on the research and development of natural colour processes that were part of the agenda of the industrial research firm, Kalmus, Comstock & Wescott, Inc., under which most of the research activities pertaining to Technicolor were conducted before 1925. See Layton and Pierce, The Dawn of Technicolor; Street and Yumibe, Chromatic Modernity, pp. 53–61.
29 Cherchi Usai, Silent Cinema, p. 33.
30 Comstock, An Outline, p. 6 (underline in original).
31 Layton and Pierce, The Dawn of Technicolor, p. 70.
32 Ibid., pp. 69–71.
33 Comstock, ‘A new duplex positive film’, Subtractive, Book 15C: Misc. Notes, Notes on Trips, 6 April 1920, p. 82, Technicolor Notebooks Collection, George Eastman House; Weaver, ‘Comments on the technical work of the 1914–1925 period’, in Comstock, An Outline (no page no. in this source).
34 Qtd in Layton and Pierce, The Dawn of Technicolor, p. 75.
35 Ibid., p. 96; Herbert T. Kalmus, ‘Technicolor adventures in cinemaland’, Journal of the Society of Motion Picture Engineers, vol. 31, no. 6 (1938), p. 567.
36 Kalmus, ‘Technicolor adventures in cinemaland’, pp. 570–71.
37 Comstock, An Outline, p. 8.
38 See Layton and Pierce, The Dawn of Technicolor, pp. 226, 230.
39 Comstock, An Outline, p. 8.
40 Layton and Pierce, The Dawn of Technicolor, p. 91. On the history of Technicolor cosmetics, see Dootson, ‘“The Hollywood powder puff war”’.
41 Troland, ‘Some psychological aspects of natural color motion pictures’, Transactions of the Society of Motion Picture Engineers, vol. 11, no. 32 (1927), pp. 687–88.
42 Winston, ‘A whole technology of dyeing’, p. 108. On how the ideal of white femininity is fundamental to filmic materiality, especially to film stock and the development and practice of quality-control measures in film laboratories, see also Genevieve Yue, ‘The China Girl on the margins of film’, October, no. 153 (2015), pp. 96–116.
43 Cari Beauchamp, Without Lying Down: Frances Marion and the Powerful Women of Early Hollywood (Berkeley, CA: University of California Press, 1997), p. 143.
44 Street and Yumibe, Chromatic Modernity, p. 95.
45 David Bordwell, Janet Staiger and Kristin Thompson, The Classical Hollywood Cinema: Film Style and Mode of Production to 1960 (London: Routledge, 1985), p. 595.
46 Comstock, An Outline, p. 7; qtd in Bordwell, Staiger and Thompson, The Classical Hollywood Cinema, p. 596.
47 Bordwell, Staiger and Thompson, The Classical Hollywood Cinema, p. 596.
48 Street, Colour Films in Britain, p. 23
49 Silverman, ‘The Toll of the Sea’, p. 35.
50 ‘The screen’, The New York Times, 27 November 1922, p. 22.
51 ‘Pictures and people’, Motion Picture News, 9 December 1922, p. 2900.
52 For more similar comments, see also ‘Decidedly the best thing that has been accomplished in colors’, Film Daily, 3 December 1922, p. 7. The paragraph reads: ‘The producers were also keen to recognize in this sympathetic Chinese story by Frances Marion, a decidedly appropriate vehicle for their experiment. The atmosphere lends itself especially well to coloring and provides a fine variety of combinations. The scenes in the garden, those by the sea and those in the Chinese girl’s home are all mighty attractive. Pictorially the feature has almost unlimited appeal.’
53 Craig Clunas, ‘Nature and ideology in western descriptions of the Chinese Garden’, in Joachim Wolschke-Bulmahn (ed.), Nature and Ideology: Natural Garden Design in the Twentieth Century (Washington, DC: Dumbarton Oaks Research Library and Collection, 1997), p. 33.
54 Ibid., pp. 27, 29.
55 Ibid., pp. 25, 30.
56 Ibid., p. 31.
57 Ibid., p. 30.
58 See Juan Antonio Ramírez, Architecture for the Screen: A Critical Study of Set Design in Hollywood’s Golden Age, trans. John F. Moffitt (Jefferson, NC: McFarland, 2004).
59 Nicholas Gaskill makes a similar argument in Chromographia: American Literature and the Modernization of Color (Minneapolis, MN: University of Minnesota Press, 2018), with regard to Harlem Renaissance in the 1920s and 1930s, in which ‘calls to linger over the surface of dark skin’ instead ‘forge a chromatic lexicon for describing skin that routes the visibility of race through the materials of modern color and so disrupts the perceptual habits that make bodily surfaces markers of racial identity’ (p. 215). In this case, ‘Skin tone becomes yet another shining product of the color revolution’ (p. 217).
60 Indeed, as Street and Yumibe suggest, Lotus Flower’s clothes are of the colours of nature, in contrast to the browns and neutral colours worn by the American characters, a colour code made explicit and reinforced when it is transgressed, as Lotus Flower changes to a dark outfit anticipating a journey with her lover to the USA only to be disappointed and ridiculed. Street and Yumibe, Chromatic Modernity, p. 96.
61 I want to thank Screen’s anonymous reviewer for suggesting this instance and for helping me flesh out the argument here.
62 Natalie M. Kalmus, ‘Color consciousness’, Journal of the Society of Motion Picture Engineers, August 1935, p. 146.
63 Ibid.
64 For canonical accounts of the relationship between cinematic representation and the codes of perspective, see André Bazin, ‘The ontology of the photographic image’, in What is Cinema? Volume I (Berkeley, CA: University of California Press, 2005), pp. 9–16, in which he famously claims, ‘Perspective was the original sin of Western painting’ (p. 12). See also Stephen Heath, ‘Narrative space’, Screen, vol. 17, no. 3 (1976), pp. 86–112; and, for a historicization of the cinema’s ‘aspiration to three-dimensionality [… as] satisfied by the blossoming of the Institutional Mode of Representation from around 1910’, Noël Burch, Life to Those Shadows (Berkeley, CA: University of California Press, 1990), pp. 6–7.
65 Kalmus, ‘Color consciousness’, p. 146.
66 Scott Higgins, Harnessing the Technicolor Rainbow: Color Design in the 1930s (Austin, TX: University of Texas Press, 2007), p. 17.
67 Street and Yumibe, in Chromatic Modernity, pp. 96–97, argue that the light green dress ‘still connects her to the natural setting of the garden; [with] lightly hued flowers on her hat continu[ing] this theme, suggesting that even across racial difference, there is a bond of femininity between the two women’.
68 Yumibe, Moving Color, p. 56.
69 Street, Colour Films in Britain, p. 31. On colour imaging processes and stereoscopy, see also Kim Timby, ‘Colour photography and stereoscopy: parallel histories’, History of Photography, vol. 29, no. 2 (2005), pp. 183–96; Birk Weiberg, ‘Functional colors: the varied applications of complementary hues’, Film History, vol. 29, no. 2 (2017), pp. 91–107.
70 Cheng, Ornamentalism, p. 2. On the relationship between orientalism and ornaments in American silent cinema, see Ruth Meyer, ‘The glittering machine of modernity: the Chinatown in American silent film’, Modernism/modernity, vol. 16, no. 4 (2009), pp. 661–84.
71 Higgins, Harnessing the Technicolor Rainbow, p. 4.
72 Street and Yumibe, Chromatic Modernity, p. 96.
73 Michael Keevak, Becoming Yellow: A Short History of Racial Thinking (Princeton, NJ: Princeton University Press, 2011), p. 1.
74 Ibid., p. 2.
75 Ibid., pp. 50–52.
76 Ibid., p. 36.
77 On how the history of skin colour became the index of racial difference, see also Anne Lafont, ‘How skin color became a racial marker: art historical perspectives on race’, Eighteenth-Century Studies, vol. 51, no. 1 (2017), pp. 89–113; Nancy Shoemaker, ‘How Indians got to be red’, The Amercian Historical Review, vol. 102, no. 3 (1997), pp. 625–44; James Delbourgo, ‘The Newtonian slave body: racial enlightenment in the Atlantic world’, Atlantic Studies: Literary, Cultural and Historical Perspectives, vol. 9, no. 2 (2012), pp. 185–207.
78 Dyer, White, p. 53.
79 Michael Rogin, Blackface, White Noise: Jewish Immigrants in the Hollywood Melting Pot (Berkeley, CA: University of California Press, 1996), pp. 12–13.
80 See Henry Jenkins, ‘“Shall we make it for New York or for distribution?”: Eddie Cantor, “Whoopee”, and regional resistance to the talkies’, Cinema Journal, vol. 29, no. 3 (1990), pp. 32–52. Similar endings of revelation of the protagonist’s identity as white instead of non-white could be seen in other, mostly comedic, films capitalizing on the taboo of interracial romance from the period. An important example is East is West (Sidney Franklin, 1922), starring Constance Talmadge in yellowface and adapted for the screen by Frances Marion. The Toll of the Sea was in fact a derivative project of East is West, released at the same time, with Chester Franklin, the less well-known brother of Sidney Franklin, as the director. See Layton and Pierce, The Dawn of Technicolor, pp. 92–93.
81 Street and Yumibe, Chromatic Modernity, pp. 244–48.
82 Dyer, White, p. 51.
83 Ibid., p. 42.
84 Ibid., p. 45. See also Dootson, ‘“The Hollywood powder puff war”’, pp. 107–31.
85 Ibid.
86 Ibid., p. 47.
87 Thomas Elsaesser, ‘Tales of sound and fury: observations on the family melodrama’, in Christine Gledhill (ed.), Home is Where the Heart is: Studies in Melodrama and the Woman’s Film (London: British Film Institute, 1987), pp. 50, 55.
88 Marchetti, Romance and the ‘Yellow Peril’, p. 79.
89 Linda Williams, ‘“Tales of sound and fury…” or, the elephant of melodrama’, in Christine Gledhill and Linda Williams (eds), Melodrama Unbound: Across History, Media and National Cultures (New York, NY: Columbia University Press, 2018), pp. 211–12.
90 Ibid., p. 214.
91 Ibid., p. 213.
92 ‘The screen’, The New York Times, p. 22.
93 Ibid.
94 ‘Critics run out of adjectives in describing “THE TOLL OF THE SEA”’, The New York Times, 29 November 1922, p. 24.
95 Ibid.
96 On ‘demonstration films’ or the ‘demonstration mode of color design’, see Higgins, ‘Forging a new aesthetic: from opera to color consciousness’ and ‘A feature-length demonstration: Becky Sharp’, in Harnessing the Technicolor Rainbow.
97 Layton and Pierce, The Dawn of Technicolor, pp. 300–414.
98 ‘A little from “lots”’, The Film Daily, 29 January 1929, p. 8. One of the stars in Manchu Love is actually the Japanese actor Sojin Kamiyama.
99 Higgins, Harnessing the Technicolor Rainbow, p. 2.
101 Karen Leong, The China Mystique: Pearl S. Buck, Anna May Wong, Mayling Soong and the Transformation of American Orientalism (Berkeley, CA: University of California Press, 2005), p. 1. Dorothy B. Jones discusses the change from outlandish to ‘realistic’ portrayals of China in the earliest book-length study on the representations of Asians in US cinema, The Portrayal of China and India on the American Screen, 1896–1955 (Cambridge, MA: MIT Press, 1955). The most representative of this realistic trend is probably the Oscar winning The Good Earth (Sidney Franklin, 1937). It nevertheless features Caucasian actors in yellowface as the main characters. See Jeanette Roan, ‘Knowing China: accuracy, authenticity and The Good Earth’, in Envisioning Asia: On Location, Travel and the Cinematic Geography of US Orientalism (Ann Arbor, MI: University of Michigan, 2010).
102 Christina Klein, Cold War Orientalism: Asia in the Middlebrow Imagination, 1945–1961 (Berkeley, CA: University of California Press, 2003), p. 23.
103 David S. Roh, Betsy Huang and Greta A. Niu, ‘Technologizing orientalism’, in Roh, Huang and Niu (eds), Techno-Orientalism: Imagining Asia in Speculative Fiction, History and Media (New Brunswick, NJ: Rutgers University Press, 2015), p. 2.
104 Maurice, ‘Red, white and blue: digital cinema, race and Avatar’, in The Cinema and Its Shadow, pp. 187–223.
100 Republican Chinese government also played an important role in the history of Hollywood censorship through protesting orientalist – or the so-called ‘China-humiliating’ (ruhua) – productions from Hollywood. See, for example, Eric Smoodin, ‘Going Hollywood sooner or later: Chinese censorship and The Bitter Tea of General Yen’, in Jon Lewis and Eric Smoodin (eds), Looking Past the Screen: Case Studies in American Film History and Method (Durham, NC: Duke University Press, 2007), pp. 169–200; Yiman Wang, ‘The crisscrossed stare: protest and propaganda in China’s not-so-silent era’, in Jennifer M. Bean, Anupama Kapse and Laura Horak (eds), Silent Cinema and the Politics of Space (Bloomington, IN: Indiana University Press, 2014), pp. 186–209.
For their comments, guidance and help at various stages, I would like to thank Jennifer M. Bean, Eric Ames, James Tweedie, Leilani Nishime, Mal Ahern, Kirsten Moana Thompson, Kirsty Sinclair Dootson, Yiman Wang, Jingsi Shen and Huanyu Yue. Special thanks to Sophia Lorent at the George Eastman Museum and Zoran Sinobad at the Library of Congress for their remote assistance during the pandemic, and to Barbara Flueckiger for allowing me to use the images from the invaluable database of Timeline of Historical Film Colors (filmcolors.org). I have had an author’s greatest blessing to receive the incisive, detailed and supportive feedback and suggestions from Screen’s anonymous reviewers; I hold my deepest gratitude for their engagement with my work.
