I delivered this paper at the University of St Andrews in 2009. It’s about ‘prosthetic costumes’ – i.e. motion capture and the way that actors ‘wear’ digital imagery. Since I am not going to publish it (and have never pursued publishing it), I’ll put it online here as a blog…
Here we go…
Recently researching an article on Motion Capture as used in Robert Zemeckis’ 2007 adaptation of Beowulf, I was struck by a comment made by actor Bill Nighy with regard to his experience of motion capture not in relation to Zemeckis’ film, but with regard to Gore Verbinski’s Pirates of the Caribbean: At World’s End (2007), in which Nighy plays a rather monstrous version of Davy Jones.
Interviewed about the motion capture technology used to construct his performance, Nighy commented that he considered the motion capture suit he wore to be his “digital pyjamas” (Marshall, 2007: 3). This prompted me to think about the possibility of such a thing as ‘digital costume’—and what precisely this might be or how we might theorise it.
Several ideas, as well as several problems, immediately spring to mind. First the problems: we can understand that digital technology may add or remove details from an image, for example removing or adding tattoos, body piercings, or simply making a character look older than the actor playing him (I am thinking of Brad Pitt in David Fincher’s The Curious Case of Benjamin Button, 2008). In this sense, digital technology is used in the services of, and in many respects is not much different from, traditional make-up. When it is not so much make-up but costume that the digital technology modifies, or rather if, as in the case of Nighy’s Davy Jones, it is both make-up and costume that is modified in order to create a coherent (if somewhat unpleasant) appearance, then the dividing line between make-up and costume is blurred: where does make-up end and costume begin if both are equally immaterial, in that both are the creation of a or a team of computer animators?
The idea of the immaterial nature of a digital costume leads to a second problem: given that the costume is added in post-production, and given that the costume never existed (even if it seeks to replicate the look and feel of conventional textures and fabrics), are we simply talking about a literalisation of the Emperor’s New Clothes scenario? That is, what can we say about a costume that is in fact non-existent, and underneath which somewhere—but where exactly?—is the naked actor, who wears a numerical (i.e. digital) costume that erases or at the very least covers over that nudity?
We can of course argue that all costume functions in the same way: any costume covers up as much as it reveals the human body from which it hangs. However, given the indexical relationship between a photographic image and the material costume that it depicts, a costume that has genuine fabric and texture, we do not so much feel that the Emperor, or in this case Davy Jones, is naked underneath the costume, as that the costume is real, that it did exist in front of the camera/in a pro-filmic manner at the time of shooting. Since we can have no such faith in the reality of a digital costume, since it is made not from fabric but from numbers, perhaps there is something different going on here.
Realism is seemingly the goal of digital animators (certainly this is the case in a blockbuster of At World’s End’s ilk), and a majority of audiences may therefore not notice the entirely false nature of a digital costume. Indeed, the ‘costume’ that Davy Jones wears seems convincingly to mimic historical trends and uniforms (he wears a sea captain’s uniform, worn at the edges and covered in cockles after years of submarine existence). In this sense, it is still a costume that signals to us information about his character in a way that traditional or analogue costume does: corrupted power, masculinity, and a red-grey tint that suggests the threatening colour of crabs or the squid that Davy’s visage is made to represent. However, I would still contend that there is room for theorising the digital costume, or what I term here the prosthetic costume (by way of signalling the above blur between make-up and prosthetics and costume).
And so having addressed some of the problems involved in thinking about digital costume, let us move on to the ideas that come to mind. Firstly, digital technology offers to filmmakers an unparalleled level of control over the image. This means that if a filmmaker is dissatisfied with the way in which a real costume moves or its colour, he can change this detail, or re-clothe his actors in a completely different style. Not only might this happen in such a way that an actor’s clothing and appearance are consistently modified throughout a scene or a film, but this may happen within a scene or even within a single shot. I have not knowingly seen this happen yet in cinema, but there are hints as to how this might work: if in Peter Greenaway’s The Cook, The Thief, His Wife and Her Lover (1989), it is through cutting that Helen Mirren’s Georgina can change dresses from white to red as she moves from one location (restaurant interior) to the next (lavatory) without any apparent lapse of time, this is not the case with the character of Rorschach in the recent film, The Watchmen (Zack Snyder, 2008), whose face mask is constantly morphing within frame and without the need for cuts. Of course, the relationship between Georgina’s location and her costume plays an important part in Greenaway’s film, and the point here is not to suggest any deficiency in Greenaway’s work. However, the shifting costume that Snyder presents to us suggests the possibilities for morphing costumes that may serve a similar function to the morphing appearances that Vivian Sobchack and others have so thoroughly theorised (Sobchack, 2000), and which have perhaps found their real-life counterpart in Generra’s Global Hypercolour line, a series of garments, predominantly t-shirts, that changed colour at different temperatures, and which were popular in the late 1980s and early 1990s. That is, rather than costume becoming a means of expressing an identity that is at any moment in time fixed according to class or gender, the morphing costume has the potential to suggest instability regarding such matters, a certain freedom from the constraints of costume, even if most filmmakers choose not to use digital technology to this end (and even if Rorschach is definitely gendered as male within The Watchmen’s male-dominated narrative).
Stella Bruzzi (1997: 9-10) has correctly argued that the Gaultier-designed costumes worn by Mirren in The Cook, The Thief, His Wife and Her Lover are spectacular, and there is a sense that spectacle plays a key role in understanding the morphing clothes of Rorschach in The Watchmen, as well as of the morph more generally. Furthermore, Sarah Street (2001) has identified how costume can often play a spectacular role in the kind of science fiction or fantasy cinema with which digital technology, in the form of special effects, is generally associated. However, while there is definitely potential for the digital or prosthetic costume to be spectacular (over and above the role of ‘real’ and spectacular costumes in The Matrix (1999) and other films analysed by Street), I would argue that this is not always the case. The practice of changing details of clothing because they distract us from the narrative, i.e. because they are too spectacular, also suggests that the digitisation of costumes can be carried out in the services of narrative, which potentially means that a filmmaker can undermine the work of her or his costume designer if they feel that they are unhappy with it, or that it somehow takes precedence over the story being told. In other words, like traditional costume, digital costume can be, as Jane Gaines and Charlotte Herzog put it, either subservient or spectacular (Gaines & Herzog, 1990).
In his PhD thesis, Richard Dyer explained the relationship between haute couture, luxury, wealth, refinement and power (quoted in Studlar, 2000: 165). While digital costumes do not at present mimic haute couture (and nor, necessarily, does haute couture mimic digital costume), the element of control that digital technology avails to the filmmaker over his actors’ appearance and costume does indirectly reflect a similar relationship between luxury, wealth and power. It is not necessarily that the digital costume itself suggests wealth, refinement or power, although any digital costume that evokes, say, Givenchy’s dresses will bring with it the same connotations of class that Givenchy’s real dresses do, even if the immaterial/digital nature of the ‘fake’/digital ‘Givenchy’ hollows out these associations if we are sharp-eyed enough to see the fakery and/or even bother to stop to think about such matters. However, it is not the costume itself so much as the element of control that the filmmaker has over digital costumes that suggests in a different sense both luxury and wealth (the ability to use such technologies while constructing a film), as well as power: the filmmaker has direct authority over all aspects of the film’s design. Power shifts from the couturier to the filmmaker, from the actor that inhabits the costume to the filmmaker who covers the performance with a numerical skin—and digital effects, including digital costumes, reflect this power, in much the same way that Bill Nighy becomes unrecognisable in At World’s End.
In spite of the above argument, it might be worth noting that to spot a fake Givenchy, be it digital or fabric, is to reveal oneself, i.e. the wearer of the fake costume, as belonging to a certain class: someone who cannot afford the real Givenchy! While Rachel Moseley (2002) has suggested the sense of empowerment that dressing up like a film star can give to filmgoers, I would argue, after Dyer (quoted now in Herzog, 1990: 155), that in some respects there is no substitute for the real thing. While it has been widely recognised that computers have improved the lot of architects, in that now they can more thoroughly model their designs, this is not the case with the fashion industry. Annie Phizacklea (1990: 53-71) has explained that computer-aided design robs the UK fashion industry’s labour force of its skills, and that it is best suited to companies specialising in a large number of short runs. That is, computers are deemed appropriate for clothes designed for mass consumption, but not for haute couture, where there remains an emphasis on manual design and production (without the space to explore here the exploitation of labour forces in the fashion industry as Phizacklea does in her book, Unpacking the Fashion Industry: Gender, Racism and Class in Production). What is true of computers within the fashion industry may remain true of them with regard to cinema and fashion: that is, digital costumes suggest not the elegance and luxury—i.e. the power—of bespoke designs, but, particularly in the case of digital costumes made to mimic pre-existing fashions, an impoverishment of elegance and luxury, a diminished power. In short, knock-off products are considered inferior (by those who can afford the originals?), such that there arises an ambivalence with regard to the digital costume: it suggests wealth (digital effects are expensive), but it also suggests inferiority to the ‘real’ clothes that could be worn. Maybe the Emperor really is naked…
Charles Eckert (1990) has famously established the historical links between the fashion industry and cinema. In the case of digital costumes, the relationship between the two may not be so clear. Without a real-world referent, the digital costume does not serve to sell any real-world clothes, except perhaps party costumes that refer back to the costumes, digital or otherwise, worn by characters in certain films. However, with regard to films like Pirates of the Caribbean, and others such as the Star Wars films (various directors, 1977-2004), we may get a sense of the consumer tie-in through the relationship between cinema and the toy industry. The relationship between cinema and toys remains under-analysed, but if cinematic narratives do play a role in the marketing and sale of toys, perhaps including computer games, then we may be able to identify a sense in which digital costumes, as much as digital special effects in general, are marketed predominantly towards young males, which in turn would suggest that although a morphing costume could in theory undermine fixed notions of gender and class, in that one’s appearance is always shifting, for the main part, e.g. in At World’s End, digital costumes are incorporated into male-dominated narratives in which gender identity is reaffirmed—even if Keira Knightley plays with swords in the films, even if Jack Sparrow (Johnny Depp) is somewhat effeminate in the film, not least through his pronounced use of eyeliner, and even if Will Turner is played by the effeminate Orlando Bloom. In fact, from this perspective, digital costumes as associated with Davy Jones, etc, may come to stand for a threatening masculinity within the film. Meanwhile, the very notion of play that is associated with dressing up, and which has a concomitant potential for reversing or making unclear gendered roles and identities, seems paradoxically to be undermined by the digital costumes within the film: playfulness may be infantile, and digital technology, especially through its use in computer games, may be playful, but the association of the digital with threatening masculinity, and the association of digital costumes with toys targeted at young male viewers, suggests that masculinity underwrites the use of digital technology in cinema, including its application to physical appearance and costume. That the virtual emerges here as being linked to the masculine is perhaps made clear in the virile etymology of virtuality, both coming from vir, the Latin for boy.
Pam Cook (1996: 43) has written that “[c]lothes mark the threshold between the body and the outside world, between the private and the public. They can hide or reveal, but either way they expose our vulnerability.” Digital costumes, particularly when worn in conjunction with digital make-up (the digital prosthesis that is motion capture), blur this distinction: the body disappears under digital pyjamas more fully than it does under real pyjamas, for the very thing that disappears from the digital creature, clothed or otherwise, is precisely its skin, the peau that can be ex-peau-sed in the first place. Real fabrics are removed and instead we have creatures that may well be in the world of the film, but they are neither in nor of the world of the viewer, which is the tantalising role that fashion and costume seems so often to play in analogue cinema: that you could possibly look like that in real life. Anne Jerslev (2006) may have pointed to the possibility for cosmetic make-over in real life through plastic surgery, through the invasion of Trinny and Susannah, through tattoos, piercings, dyes, fashion, make-up, etc. But it remains an elusive and perhaps undesirable goal to look something like Bill Nighy’s Davy Jones, not least because this would involve the removal or effacement of the skin that grounds us in reality. In other words, if Pam Cook (1996) feels that fashion offers us the opportunity to change places out of our everyday roles, digital costumes offer not an escape grounded in the body, but one grounded on virtual air, on nothing.
In a certain sense, the digital costume paradoxically can become even more exclusive than a Givenchy dress: when not imitating existing fashions, or at the very least when allowing characters to wear fashions or uniforms that no longer exist, as per Davy Jones, the digital costume has the potential to suggest an escape that is even more rarefied, perhaps precisely because the digital allows a modification and personalisation of appearance that far exceeds even that of a bespoke dress. No wonder the playful and sometimes disturbing appearances that real life humans give to their digital avatars in virtual environments such as Second Life or World of Warcraft, environments that some users or players feel is the best outlet for their self-expression, in a continuation of the liberating tradition of designing one’s own clothes and/or imitating the fashions of stars that Jackie Stacey (1994) and Rachel Moseley (2002) have identified. If Pam Cook (1996) identifies clothes as a fetish object, then here perhaps the digital costume truly is a fetish: it is a virtual symbol that comes to stand in for the desired appearance that reality cannot offer, the missing appearance that would make us ‘whole’ in real life.
Pam Cook (1996) also explains how nostalgia is a sentiment often associated with costume. And while the freedom enabled by control in the use of digital technology as applied to cinema might indeed allow Davy Jones to wear a sea-drenched costume that is reminiscent of sailors’ uniforms from centuries past, it is interesting to note that this digital costume evokes nostalgia not necessarily for anything real or anything that once we could have known (since no one on Earth was alive to see the real Davy Jones or his locker), but a nostalgia for a past that is non-existent and unknowable or, more precisely, if existent and known, it is only existent in and known through the media. The role of existing costumes no doubt influences the design of virtual clothes or prosthetic costumes in cinema and by users in virtual environments alike, allowing us to indulge our nostalgia for and to fetishise a self that never existed anywhere except in our imagination, a self that typically we characterise as existing in the past (hence, nostalgia), but which may well exist in the future (make-overs giving us real appearances as idiosyncratic and bizarre as those of our Second Life avatars) and even in the present (the virtual present of virtual environments that do now exist alongside the real world in visible and interactive form). No wonder Bill Nighy called them his digital pyjamas: free from the skin, the digital costume does not so much expose/ex-peau-se us, as give us the (at present, but for how long?) illusion of freedom and wholeness of identity that is known only in dreams…
Bruzzi, S. (1997) Undressing Cinema: Clothing and Identity in the Movies, London/New York: Routledge.
Cook, P. (1996) Fashioning the Nation: Costume and Identity in British Cinema, London: British Film Institute.
Eckert, C. (1990) ‘The Carole Lombard in Macy’s Window,’ in Fabrications: Costume and the Female Body (eds. Jane Gaines and Charlotte Herzog), New York/London: Routledge, pp. 100-121.
Gaines, J. and Herzog, C. (1990) Fabrications: Costume and the Female Body, New York/London: Routledge.
Herzog, C. (1990) ‘“Powder Puff” Promotion: The Fashion Show-in-the-Film,’ in Fabrications: Costume and the Female Body (eds. Jane Gaines and Charlotte Herzog), New York/London: Routledge, pp. 134-159.
Jerslev, A. (2006) ‘The Mediated Body: Cosmetic Surgery in Television Drama, Reality Television and Fashion Photography,’ Nordicom Review, 27:2, pp. 133-151.
Marshall, L. (2007) ‘Rise of the Machines,’ Screen International, 1606: 3.
Moseley, R. (2002) Growing Up with Audrey Hepburn, Manchester/New York: Manchester University Press.
Phizacklea, A. (1990) Unpacking the Fashion Industry: Gender, Racism and Class in Production, London/New York: Routledge.
Sobchack, V. (ed.) (2000) Meta-Morphing. Visual Transformation and the Culture of Quick-Change. Minneapolis/London: University of Minnesota Press.
Stacey, J. (1994) Star Gazing: Hollywood Cinema and Female Spectatorship, London/New York: Routledge.
Street, S. (2001) Costume and Cinema: Dress Codes in Popular Film, London: Wallflower.
Studlar, G. (2000) ‘“Chi-Chi Cinderella”: Audrey Hepburn as Couture Countermodel,’ in Hollywood Goes Shopping (eds. David Desser and Garth S Jowett), Minneapolis/London: University of Minnesota Press, pp. 159-178.