by Rob Latham;
In short, Bellmer’s dolls were intended, as I have argued about ASFRian fantasy, as a desublimatory assault on the normative, stable, and cohesive subject, and in particular on the psychic armoring of the fascist body by which fragmentary, fluid, and chaotic drives were repressed, abjected, and mapped onto the Other, represented in the case of the Nazis by women, Jews, homosexuals, and Communists.25 They speak not to the aesthetic gaze, in which the sexual drives are sublimated through the object of beauty, but to the curious gaze of Pandora, who opens the box and experiences the uncanny vertigo of her own true nature (as does Nathanael when the automaton is revealed as not just a mechanical object, but an extension of the mechanical compulsions of his own psyche).
Eye robot
Although Bellmer’s dolls were a product of their time, created in dialogue with the cult of the perfect body within fascism and the hysterical body within Surrealism, they anticipate the dislocation of bodies and identities, as well as the interpretive dilemmas they inspire, within current technocultural contexts. Indeed, the extent to which cyberbodies are, like those of hysteria, displaced (by prosthetic implants and extensions), dispersed (by communications technologies), and subject to forces outside the will of their owners, is one of the central concerns of cyberpunk. Moreover, like Bellmer’s disarticulated dolls, when such technologically decentered figures cross into gendered and eroticized zones, they often inspire conflicting claims of containment and liberation, as well as conflicting reactions between alienation and desire. The connection between the two is made explicit in Mamoru Oshii’s Ghost in the Shell 2: Innocence, which uses Bellmer’s dolls as a central visual thematic to underscore the uncanniness of embodiment and desire when mediated through technology.
The sequel to Ghost in the Shell (Kôkaku kidôtai, 1995) and similarly adapted from the manga by Masamune Shirow, Ghost in the Shell 2 explores at greater length not only the kinds of questions posed by the first film about mind, matter, and spirit in the information age, but also the aesthetic and ethical concerns raised by the encounter with artificial bodies. The second film, in fact, seems to address directly the critical reception of the first film, particularly concerning the central protagonist, Major Motoko Kusanagi, a female cyborg. The first film takes place in the year 2029; Kusanagi is a special agent whose bodily parts are entirely artificial—except for her original brain tissue, which is encased within a titanium skull—and owned by her employer, the government security force Section Nine. Both cybernetically enhanced and contained, posthuman and sexualized, Kusanagi has inspired comparisons to Haraway’s cyborg in her embodiment of both “the liberating and the dehumanizing power of technology,” as well as critiques of her seemingly ambivalent suspension “between a progressive and a reactionary politics of technology or gender” (Bolton 730). Carl Silvio, for example, casts her as a posthuman update of Pandora, suggesting a deceptive split between her appearance and reality; while she poses as a radical cyborg, she ultimately reinforces traditional gender roles, exhibiting a duplicity of which a great deal of cyborg culture is guilty: “There is thus what might be called an element of seduction at work, whereby information technology often presents itself to us as potentially liberating when in fact our actual interactions with it often reinforce conventional social structures of domination” (55).
If Silvio’s critique seems to echo those of Bellmer’s dolls in its concern for the status of the culturally and physically situated female body in the face of its disarticulated counterparts, other critics have, in a manner similar to Bellmer’s defenders, pointed out that such a critique ignores the distinction between representations of embodiment and actual bodies. Christopher Bolton, for example, draws from the long tradition of puppetry in Japan in order to illustrate the ways that artificial bodies (especially when represented within anime) are marked by an added layer of performativity that inflects their meaning, and he suggests that to ignore the performative aspect of these bodies is to miss their critical potential:
Concerned with linking anime to a real-world context in which flesh-and-blood bodies are threatened with genuine objectification and violence, this approach treats fictional cyborgs on more or less the same plane as living human subjects. But treating Kusanagi as a living subject clearly misses the ways in which her body will always fall inside quotation marks; she is a virtual or performed subject that is both unreal and more than real from the start. (737)
Bolton’s reminder of the puppet’s artifice is a leitmotif in Ghost in the Shell 2, a film that explores the confusion between the human and the artificial, as well as the real and the virtual, in relation to a central question posed, significantly, by a character named and fashioned after Donna Haraway: “Why are humans so obsessed with recreating themselves?” In the film, the character Haraway is a coroner, who appears in a lab surrounded by android parts, including bodies hanging from hooks and a vat of artificial eyes. Although one might be tempted to compare her appearance to that of the real Donna Haraway, when the scene comes to a close, she lifts up a faceplate to reveal, behind her eyes, a technological viewing apparatus implanted in her skull. Echoing the scene in “The Sandman” when Olympia’s eyes are removed, it is one of many reminders in the film that we are in the zone of the visual uncanny, where the boundary between human and machine, as well as between perception and reality, is unstable. Indeed, Oshii has explicitly stated the importance of the concept of the uncanny for understanding his film, which displays, in particular, an “obsession with the uncanniness of ningyõ (literally, ‘human-shaped figures’) in the form of dolls, puppets, automata, androids, and cyborgs” (Brown 222). Moreover, Ghost in the Shell 2, like Hoffmann’s story, interrogates the uncanny in the realm of desire through the figure of the female android.
The film opens three years after the close of the story in the first Ghost in the Shell. A prototype of a new gynoid model, the Hadaly 2052 (manufactured by the corporation Locus Solus specifically for sexual purposes), is running amok, and there have been a string of incidents in which Hadaly gynoids kill their owners and, shortly thereafter, self-destruct.26 The film follows the investigation by Section Nine, this time led by the cyborg Batou, the second in command under Kusanagi in the first film, and his “mostly-human” partner, Togusa. It opens with a chase scene in which Batou follows the trail of a Hadaly who has just killed her owner and two police officers. Batou corners the wayward gynoid in an alleyway and, just as he is about to shoot her, she cries “Help Me!” and proceeds to self-destruct, ripping open her own chest to expose her inner mechanism and then ejecting her metallic skull like a jack-in-the box (see Figure 6). This scene is more than suggestive of the kind of unmasking central to both ASFRian and Bellmerian fantasy; as Steven Brown notes, it was inspired by an illustration by Bellmer entitled Rose ouverte la nuit (Rose opened at night, 1934), in which a young girl rips open her own skin to reveal her inner organs (see Brown 239). The reference to Bellmer becomes more explicit in the opening credits that follow, in which title cards are interspersed with a series of sequences depicting the manufacture of a Hadaly-type gynoid, whose ball-jointed limbs move into place within a watery environment, accompanied by the sound of clicking gears. As a pair of legs without a torso floats into view, it seems to undergo binary fission, doubling before our eyes into a nearly exact replica of Hans Bellmer’s ball-jointed doll, before separating into two identical gynoids (see Figure 7). The reference to Bellmer is reinforced yet again later in the film, when Batou finds a significant clue to the mystery of the malfunctioning gynoids in a representation of the Japanese reissue of Bellmer’s book “The Doll” (published in 1995) that he discovers at a crime scene. An inspector has been murdered by yakuza affiliated with Locus Solus. The clue, slipped into the pages of Bellmer’s book, is a holographic image of a young girl, whom Batou will eventually discover to be only one of many children abducted by the yakuza for Locus Solus for the purpose of “ghost dubbing”—extracting their ghosts to ensoul the Hadaly gynoids and make them more desirable
. The children’s bodies are kept suspended within mechanical cocoons. This turn of events not only echoes the narrative of Villiers’s Tomorrow’s Eve, in which the android Hadaly (the citational source for the Hadaly androids in the film) is ensouled by a living woman in a catatonic trance, it offers an inversion of the climactic scene within Hoffmann’s tale: instead of the woman being revealed as an android, the android is unmasked to reveal the human beneath the mechanical exterior of the doll.
Figure 6 Exploding Gynoid in Ghost in the Shell 2.
Figure 7 Ball-jointed doll in GIS2 opening.
Although the resolution to the murder mystery is conventional in the sense that the villains are stopped and the young girls are saved, the ethical landscape of the film is less clear. When the little girl from the holographic portrait is freed, she explains to Batou that the inspector was killed for overwriting the ethics code programmed into the gynoids, which prevented them from harming humans or themselves, the rationale being that if they malfunctioned violently, it would help to draw attention to the children’s plight. “But what about the victims?” Batou asks, to which she reacts with a tearful outburst, “I didn’t want to become a doll!” Kusanagi, whose ghost appears in the film only in this final scene in the form of one of the Bellmerian Hadaly bodies (which she has temporarily inhabited to help Batou), responds to the little girl’s lament: “If the dolls could speak, no doubt they’d scream: I didn’t want to become human.” There is, so the film seems to suggest, an equal injustice committed in forcing humanity upon the doll (whether visually, narratively, or critically), as there is in making human girls doll-like. Indeed, the film not only “remediates Bellmer’s dolls” (Brown 223), it also rearticulates his aesthetic conviction (quoted above) that “the object identical to itself remains devoid of reality.” This point is emphasized through a variety of strategies, both narrative and formal, for defamiliarizing or rendering uncanny the characters in the drama in ways that remind us repeatedly that they are not who or what they appear to be.
One strategy of defamiliarization, which Oshii borrows from French New Wave director Jean-Luc Godard, is citationality. The dialogue throughout the film is interspersed with quotations of other authors—meditations on the relationship between mind and body, as well as between dolls and humans, drawn from such diverse sources as René Descartes, Heinrich von Kleist, Villiers de l’Isle Adam, Julien Offray de La Mettrie, John Milton, and the Old Testament. This technique adds a critical dimension to the film by drawing attention to the history of ideas that informs its inquiry into the human and the artificial. In this way, the characters speak both for themselves and within a network of cultural and historical relations of which the viewer is made continually aware. Characters are also disrupted from within the diegesis through e-brain hacking. There are numerous instances in the film in which, as Brown puts it, “one character literally or metaphorically ‘pulls the strings’ of another” (224) by hacking into his cyberbrin (implanted cybernetic components that allow the brain to interface directly with information networks) in order to exert control or to implant false realities or memories.27 A third strategy, of particular relevance to the discussion in this essay, is the kind of unmasking so critical to ASFRian desire and fantasy. The Bellmerian disassembly of the gynoid at the beginning of the film, which sets the murder investigation in motion, will return in an uncanny sequence towards the end, when Batou and Togusa enter the mansion of the hacker Kim. In the middle of interrogating a grotesque, puppet-like figure whose e-brain Kim controls (raising the question of whether or not Kim is still alive), Togusa’s e-brain is hacked in such a way that the scene that we have just watched is repeated three times. In the first repetition, the puppet figure that Kim inhabits is now a doppelgänger of Togusa, who quotes freely from Jentsch’s essay on the uncanny relationship between dolls and humans. The (seemingly) real Togusa watches in confusion when, suddenly, Batou turns mechanically towards him and his interior metallic skull ejects out of his head in the exact same manner as that of the wayward gynoid at the beginning of the film. This sequence of events sets off the second repetition: this time, the puppet that Kim inhabits resembles Batou, and the sequence ends with Togusa’s chest exploding to reveal an artificial interior. It is only in the third repetition that Togusa is freed from the virtual feedback loop in which he has been caught.
This sequence, which employs all three techniques mentioned above, renders uncanny not only the characters in the film—both partly human and nonhuman, the investigators and the investigated—but also the film itself. Its kaleidoscopic repetition of events is confusing (particularly since Togusa is visible, so that the cinematic viewer is outside his point of view while being inside his hallucination) until we realize, along with Togusa, that an intervening force is at work. It also, as Brown notes, points to larger questions about “what it means to be human in a posthuman world and how we are to relate to all the ningyõ (dolls, puppets, automata, and androids) that inhabit the world with us” (234). To the extent that answers are offered in the film, they are partially inscribed by its representational approach to the bodies that populate its posthuman landscape.
In conjunction with the above strategies of defamiliarization, Oshii adopts a style of animation that combines 2D and 3D rendering, which reinforces the critical distance between representation and reality, human and humanoid. The more lifelike the character, the more simply rendered and the more likely that it is a hand-drawn cell animation, whereas less animate objects are often computer-rendered in 3D. The contrast is made particularly apparent in a scene of an outdoor puppet festival, based on the annual Dajia Matsu Festival in Taiwan (see Oshii and Yamada 195), in which the larger-than-life, mechanical puppets are virtually realistic, while the living humans who move within and around them are both simply animated and partially obscured. The animation style of the film conforms to the aesthetic tenets of bukimi no tani or “The Uncanny Valley,” a theory first articulated in a 1970 paper on robot design by Masahiro Mori, the man considered the father of Japanese industrial robots, but increasingly applied to computer-generated effects within animation and gaming.28 Inspired by Jentsch’s essay on the uncanny, Mori suggests that as anthropomorphic creations become increasingly human, they create greater expectations of human movement, behavior, and appearance, and when those expectations are not met, they produce a creepy or uncanny feeling. His ideas are illustrated in a graph, which charts the degree of realism or humanness achieved (both in terms of motion and appearance) and the resultant sensation evoked (see Figure 8). At one end of the graph are toys and puppets, while at the other end is perfect verisimilitude, both ends of which, according to Mori, inspire various degrees of pleasure. The graph dips dramatically into the unpleasurable uncanny valley between these two points, where one finds prosthetics and, at the lowest point on the graph, the moving corpse or zombie (see Mori).
Like Jentsch’s theory of the uncanny, Mori’s “Uncanny Valley” assumes the emotional investment and pleasure of humans in the nonhuman, offering an aesthetic program for enhancing that pleasure through a dedicated artificiality rather than a simulation of humanness. In so doing, Mori’s views echo traditional Japanese aesthetics as influenced by Buddhism (Mori discusses the relationship between Buddhism and robotics in his 1974 book The Buddha in the Robot)—which tend to emphasize evocation over description, achieved via the interplay of opposite states, such as light and shadow or sound and silence. Thus they help to forge a bridge between current forms of technological embodiment and Japan’s historical legacy of dolls and puppets.29 Indeed, the current prototypes for Japanese humanoid companions, which tend to have a distinctly anti-realistic toy or puppet-like appearance, are often viewed as descendents of karakuri ningyõ, autonomous mechanical or clockwork dolls that were popularized during the Edo period in Japan, which could perform a variety of tasks and entertainments, and whose goal “was not realism but charm” (Hornyak 25).
One of the most popular karakuri was a childlike chahakobi
ningyõ or “tea-serving” doll, which is recreated in the uncanny mansion sequence of Ghost in the Shell 2. A small puppet-like figure, the tea-serving doll would travel across the room with a teacup on a small serving tray and, after the cup was taken, it would wait for its return to the tray, after which it would turn and travel in the opposite direction (see figure 9). The doll’s ability to “perform the quintessential act of Japanese hospitality—serving green tea” (Hornyak 21) reflects Japanese cultural attitudes towards mechanical humans, including robotic humanoid companions, which are, as Hornyak notes, “social machines” designed primarily for “communication with human beings” (21).30 The subtle and abstract motion achieved by this “robot from the Edo period” has informed everything from bunraku puppet theater to the cyberbodies of anime, all of which use an economy of expression to achieve a maximum emotional impact, reinforcing the idea, expressed by Mori’s theory, that the deepest chords of humanity are better struck through an intended artifice rather than through realism.
Figure 8 The Uncanny Valley.
Ghost in the Shell 2 recreates the legacy of Japanese ningyõ, including dolls, puppets, and automated mechanisms (karakuri), and reminds us of their relevance to the cyberbodies of the current technosphere. To the extent that we either desire or revile such bodies, it is, so the film suggests, because we have compromised their “innocence” with our own projections of humanity. In repeatedly shattering our illusions through its various strategies of unmasking, the film illustrates the affective power of nonhuman bodies recognized for what they are—both “real and unreal, simultaneously more and less than human” (Bolton 745)—and it demonstrates the possibilities for recuperating an increasingly lost humanity in relation to such bodies. For example, while the character Batou is a visually static, largely artificial being, his relationships with his basset hound, Gabriel, and Motoko Kusanagi form the emotional center of the film.31 Like Haraway, whose Companion Species Manifesto on dog-human relationships echoes her “Cyborg Manifesto” in its concern for “an ethics and politics committed to the flourishing of significant otherness” (Companion 3), the film uses the relational interplay among a cyborg, a dog, and a networked ghost to gesture towards an ethical and emotional reciprocity outside of anthrocentrism. It is in the service of the relationship between Batou and Kusanagi, in particular, that, according to Oshii and Misaki Yamada (who wrote the novelized prequel to the film), the narrative structure of the “murder mystery” is used:
