Happy holidays from The Popular Uncanny.
Happy holidays from The Popular Uncanny.
I like to think I’m good at keeping up with research on the Uncanny, but somehow I missed an important event this June: IEEE Spectrum published the first complete English translation of Masahiro Mori’s highly influential article on “The Uncanny Valley” (originally published in what they call “an obscure Japanese journal called Energy in 1970,” and circulating in the robotics community and popular culture in only partial form. This current translation, by robotics experts Karl F. MacDorman and Norri Kageki, has been authorized and reviewed by Mori himself.
Here are my initial notes as I read Mori’s article in full:
+ Movement has a more significant role in the theory than I think people who work with this theory really recognize. In the introduction to the essay, he makes the stunningly simple point that “many people struggle through life by persistently pushing without understanding the effectiveness of pulling back. That is why people usually are puzzled when faced with some phenomenon that this function (an algebraic equation for “monotonically increasing” or accelerating forward movement) cannot represent.” In other words, the “uncanny” is referring to the “puzzling” phenomenology of feeling “pulled back” from a situation where we expect forward motion. I like this, as it gives me another way of thinking of the “double-take” that I associate often with das Unheimliche.
+ Death is conceived as the end of movement. In the end of Mori’s article, he associates this “pulling back” with death itself: “into the still valley of the corpse and not the valley animated by the living dead.” He even goes so far as to theorize that the repulsion of the uncanny is “an integral part of our instinct for self preservation…that protects us from proximal, rather than distal, sources of danger.”
+ One shouldn’t forget that the “Valley” is always symbolic, a metaphor for a sensation. I am struck by Mori’s reliance on metaphors throughout the article…and reminded that the very idea of the “valley” is really a geographic analogy for a dip in his infamous graph. The “sinking feeling” one associates with a dip in the road or the sudden plunge of a roller coaster might be just the right sensation he is after in this. He directly compare the “uncanny valley” to an “approach” of a hiker climbing a mountain who must sometimes traverse “intervening hills and valleys”. Mori thesis statement encapsulates this in a nutshell: “I have noticed that in climbing toward the goal of making robots appear like a human, our affinity for them increases until we come to a valley which I call the ‘uncanny valley’” (emphasis added).
+ Aesthetics and childhood factor into this theory as much as Freud’s. Mori notes that the trend for designing robots that look human really started to pick up in toy robots, rather than in the (perhaps more frightening) factory robots that replace human workers. Obviously, the aesthetics mean more than the instrumental functions of these toys, which are like Freud’s puppets or uncanny dolls. Interestingly, Mori writes that “Children seem to feel deeply connected to these toy robots” and puts them on the top of the first “hill” before the chart dips down toward the deadly “uncanny” valley. What I would note here is that Freud conceives of the uncanny as a return of a repressed or infantile belief that such objects as toys and dolls have life all their own. After the dip of the “uncanny valley” Mori returns to toys by citing the “Bunraku Puppet” as an example. So perhaps Mori’s chart actually follows Freud’s logic to the letter — the dip or valley is the return of the repressed, insofar as it follows the same chronological structure of childhood belief, followed by its later return in adulthood.
+ The focus on hands, rather than faces or heads, is intriguing. Mori focuses on the robotic hand as his primary example: “we could be startled during a handshake by its limp boneless grip together with its texture and coldness…the hand becomes uncanny.” Later Mori develops this by describing a robotic hand as prosthetic limb: “…if someone wearing the hand in a dark place shook a woman’s hand with it, the woman would assuredly shriek.” Note that Sigmund Freud’s original article on the Uncanny (1919) features examples of dismembered limbs and hands that “move of their own accord” as well. The hand is a particularly loaded body part: it is a way we communicate by sign, it is one of the ways that “human” is separated from other members of the “animal” kingdom (by opposable thumb), and it is something we look to as a signifier of intention. Language is rooted in the hand. And it is a metaphor for control (i.e. having everything “in hand”). The Uncanny, as I think of it, is often a phenomena that disorients us and — often as if by an “unseen hand” — reminds us of our lack of mastery in a situation where we normally would presume we had it.
+ As Mori progresses to unpack his theory, the more his descriptions of prosthetic devices become akin to horror fiction. Indeed, he makes the comparison himself: “Imagine a craftsman being awakened suddenly in the dead of night. He searches downstairs for something among a crowd of mannequins in his workshop. If the mannequins started to move, it would be like a horror story.” Indeed, the surprising movement where one expected stillness from the inorganic objects would be startling and felt as uncanny. It would not merely be identity confusion. It would feel as if the robots were attacking. The boundaries between “story” and reality would become blurry. Science would become science fiction “made real.”
I’m sure I’ll return to this article again in the future. For now, of equal interest is Kageki’s contemporary interview with Mori himself also published in the 2012 issue of Spectrum, asking him to look back on the theory. My favorite moment is the when Kageki asks him, “Do you think there are robots that have crossed the uncanny valley?” Mori suggests that the HRP-4C robot is one of them. But then he doubles back and says “on second thought, it may still have a bit of eeriness in it.”
I am so thankful to Mori and the translators for re-releasing this version of the article in English for scholarly review. It has been relatively frustrating to see the essay referred to so often in both the design and gaming community — as well as in scholarly circles — without having access to the complete source, and now that it is available I hope others will continue to test and explore its legitimacy as a way of thinking about horror aesthetics and anthropological design.
I’ve written often about the Uncanny Valley elsewhere on this blog.
“These adorable pets offer a real pet ownership experience without the hassles and expense. Say goodbye to feedings and vet bills. Say hello to lots of love and cuddles. Perfect Petzzz – the ultimate pet.” — Perfect Petzzz website
“It is not a toy,” [VP of Marketing] Clarkson says, “but this is the closest you can get to real pet ownership without the hassles or responsibilities of owning a real pet.” — journalgazette.net
“In 2005, Perfect Petzzz® generated more than $20 million in retail sales in its first full year of operation. In fact, the Perfect Petzzz cart program was named the most successful new product concept in 2005. With the overwhelming demand for these lifelike puppies and kittens, we’ve seen other companies try to produce imitations.” — CD3 Press Release to PP Mall Dealers
Perfect Petzzz are stuffed animals that breathe. The autonomous movement of their fur — controlled by a battery-powered engine you don’t expect to be there — is enough to trick the eye into presuming that the puppy or kitten curled up on the floor is actually a living, breathing, pet. Cute, and perhaps attractive to your hand’s caress, until you touch it and realize it’s not real. Then you are startled and the toy enters the already doll-crowded realm of the popular uncanny.
Of course, the Perfect Petzzz (the ”zzz’s” are for snoring) are plastic. And therefore the animal it represents is literally as dead as it looks, with its eyes closed and body stiffened into a disturbing fetal curl. It should not move, but it does, and it is this representation of death-stirred-to-life — of the presumed inanimate object surprising us with its animation — that gets our reaction. The tricky switcheroo of statuses between familiar and unfamiliar spin the roulette wheel of certainty: the domesticated animal is rendered un-familiar (stuffed, inanimate) then restored to a heimish (cozy) status of sleeping and napping..
It is surely cute, and there is little difference between a breathing stuffed animal and a toy doll that burps or blinks. Of course, even the cutest of dolls are inherently uncanny in the way they are semblances, pale imitations of life…but the creepy thing in this case is not so much its status as automaton, as the fact that this “sleeper” never wakes up. These are comatose pets…and that, perhaps, is what makes them so “perfect.” Like the commodities these organic creatures have become, our domesticated pets are “perfect” when they are behaved, controlled, and easily replaceable after they expire. Even more, these plastic pals are simulacratic forms of taxidermy (and surely a savvy taxidermist has already borrowed the motor or at least the concept for an experiment or two). Another form of death, fantastically alive through the magic show of animism, nostalgia and fantasy. Living, breathing, death.
While theories of the “uncanny valley” are debatable (see Hanson’s “Upending the Uncanny Valley” (.pdf)), the quest for human-like androids and automatons continue to compel their designers. At Carnegie-Mellon University’s anthropomorphism.org, I found an interesting early study of robot head design that shows how these designers sometimes make choices about when to make robots anthropomorphic (human-like), and when to avoid such resemblance.
In “All Robots Are Not Created Equal,” Carl F. DiSalvo (et. al, 2002) analyzes the human perception of the humanoid robot head in alarming detail, from the length between the top of the head and the brow-line, to the diameter of the eyeball, to the distance between pupils. The researchers want to know: how human should a robot head be, and is this contingent upon the context in which they are employed? Their study suggests that eyes, mouth, ears and nose — in that order — seem to be the most important traits for us to perceive the “humanness” in a machine. But the most interesting conclusion they draw, in my view, is that the more servile and industrial the robot, the less we want to perceive its resemblance to us. Thus, not all robots are created equal: “consumer” robots often are purposely more “robotic-looking” (mechanical) in design, since they often perform servitude and routine functions that would crush the spirit of any real human, while others — especially “fictional” — robots are often the most human-like of all, reflecting our projected fantasies for them as “characters.” DiSalvo and crew propose that the following elements of robot design would create the ideal “human-like” robot:
1. wide head, wide eyes
2. features that dominate the face
3. complexity and detail in the eyes
4. four or more features
6. humanistic form language
To what degree is our notion of the “double” located on the head, the face and its various features? Freud’s classic itinerary of uncanny traits include doll’s eyes and language, and I would suggest that the more the traits listed above appear in a doppelganger, the more uncanny that double might be. Moreover, the role of the uncanny valley is at work here, and while this theory is not directly addressed in DiSalvo’s article, it’s worth considering the degree to which the factor of increasing “likeness” in robot head design follows the x-axis of the classic uncanny valley:
It is useful to consider not only the “uncanny” in this chart, but the way that that assumptions about use value and instrumentality lie behind its structure. There is a politics of self/othering at work in this schema that is rarely discussed. One of the fundamental principles of the Uncanny as it is classically understood in aesthetics is that, symbolically, the “double” is a harbinger of death for the subject that perceives it. This is a complicated notion, but on one level what this means is that when the self perceives itself as disembodied and located in another entity — through its mirror image — we unconsciously recognize how “replaceable” we are and this is felt as uncanny. We do not only respond, typically, with fear: we also feel compelled to separate the Self from the Other as a form of protection against the threat that the Other presents. A power relationship transpires: the psyche construes a hierarchical separation that institutes the Self in a higher subject position than the Other, in order to retain its sense of mastery over identity. The Other is subjugated into a lower position, often one that is loathed or considered repulsive. While such Othering is “harmless” in fiction, this is also a dream that reproduces the politics of everyday life.
There is a generalized fear of robots and other forms of artificial intelligence “replacing” mankind; we see it everywhere in science fiction, but it is also a very real threat to the labor force. Robot design participates in a self/othering dynamic that domesticates these anxieties. Could the uncanny valley be a symptom of class conflict as much as some organic reaction formation? I think so.
I thank my colleagues at Seton Hill University, Laura Patterson and Maureen Vissat, for recently passing along a YouTube link to “Doll Face” by Andrew Huang. It’s a brilliant treatment of the relationship between media technology and gender identity, using uncanny structures like automatism and the compulsion to repeat to deliver its message.
The video sent me to Huang’s website, which features many stunningly uncanny animations worth sharing, analyzing, and potentially using in a college classroom. Huang’s art is more than “pop” but it appeals to the popular imagination through iconic treatements of domesticity-made-strange. His excellent short film, The Gloaming features deja vu in a disturbingly ominous way, reminiscent of the work of Jan Svankmajer or the Brothers Quay. Even his advertisements for Moo Studios use fantastic transformations of ordinary furniture and objects, giving them an unexpected life all their own. But his music video for Eric Avery’s “All Remote and No Control” is perhaps the most horrifying and uncanny of them all, as it represents the boundaries between the urban and the domestic under transgression by an almost Lovecraftian representation of nature — with chilling results. Here’s the version from YouTube, but a higher quality version is on Andrew Huang’s excellent website itself.
For Halloween, the readers of Oobject voted for their Top 12 Videos of Creepy Automata. A great theme, from cats in a milk churn to maniacally laughing dolls. One of my favorites is this clip of a Decaying 1880s Automaton Harpist by Vichy:
I won’t belabor how uncanny the signifiers are here, from the doll’s movement on its own accord to the way the eyes seem to cast around and occassionally return one’s gaze. The decaying apparatus is like one of Hans Bellmer’s dolls stirred into life by an electrical current. But it’s the fluid movement of the dead hands and arms that get me — human in their plucking of the strings of an absent (ghost?) harp, as the doll plays along with a creepy tune. Unheimlich!
If you go to Oobject, be careful. You might find yourself spending hours on end in their wonderful “weird” category. Or their list could inspire a day- or week-long browsing expedition in youtube for “automata.”
[See my related discussion of medical mannikins on Oobject in a previous blog entry.]
Wade Marynowsky’s weblog, Autonomous Mutations, highlights current uncanny art projects and other manifestations of das Unheimliche and is full of fantastic and unique examples of the aesthetic (like Karakuri ningyo), links to Machine art, and also references to uncanny theory.
I say he features the “aesthetic” of the uncanny because his blog is an offshoot of his own excellent art inquiries, featured in “Autonomous Improvisation” one of several multimedia exhibits that Marynowsky has created that integrates popular music with concept art to inquire into the uncanny nature of art and computer programming. Marynowsky describes his initial project (pictured above) and artistic intentions this way:
a prepared pianola is linked to a network of computers and is programmed to orchestrate [a palette of videotaped musicians from diverse genres performing for the camera] creating an ever-changing composition. This is presented via three-channels of audio–visual projection. Through non-determinist re-composition, the work questions if it is possible for improvisation to be programmed, or if this is simply a paradoxical endeavor. More significantly, ‘Autonomous Improvisation v1′ asks us to consider what is imposed on human autonomy in an increasingly computer-controlled society.
The Autonomous Improv exhibit employs stock icons of the uncanny in random patterns (e.g. images flicker on the wall like ghosts, projecting masked singers who look like dolls, mix-mastering clowns working turntables, etc, while the old-fashioned pianola works like a player piano with candles aglow in its heart), but the real effect of the uncanny is felt when all of this “clicks” into an accidental surprise where it seems like the object is actually making music and working in some kind of uncanny harmony.
The following video from YouTube doesn’t do the sound justice, but it is a good example of the exhibit in motion. Those interested in this project should seek out Marynowsky’s DVD from his own Demux label.
In “Alien Jukebox,” a review of the exhibit for ArtSpace, (available in a a pdf file), Sean Lowry raises an interesting question about the role of the artist/author in constructing an experience of the uncanny.
Artists have long created systems or parameters in which events might unfold. Whilst the exact outcome might be unknown, a kind of outcome is expected…whenever an artist has waited for something to rust, for torn paper to fall, for mediums to react with one another, for randomly cut up audiotape to be spliced together, or for an audience to interact with a performance, repeated activities are demonstrated to produce self-similar outcomes rather than specific outcomes. Since Marynowsky has devised the program, he has also designed the kind of outcome that it will produce. But in establishing the parameters within which the work will operate, the fact that its final configuration is automated does not necessarily imply that the technology is acting “on its own accord.”
Here Lowry is arguing against Wade Marynowsky’s assertion that the work is a form of uncanny “automata” because it does not act “on its own accord,” but is rather programmed for a particular outcome to transpire. The so-called “ghost in the machine” was put there by the artist. In other words, the uncanny experience is like a firecracker: it may stun us with a burst of angst in response to its automatic semblance, but the artist has still metaphorically lit the fuse.
But all art has an artist standing in the curtains behind it in some way (even long after his or her death!), and this is not necessarily an issue. It doesn’t reduce the uncanny effect of the artwork on the spectator. Another way of thinking about what I would call the “player piano” effect: the exhibit aims at randomness that syncs up music and image in an uncanny way, but it is nevertheless programmed in much the same ways as a player piano relies upon a script to play its tune. However, in this case, the script roll has been programmed to randomize its notation and express that randomization through imagery of the uncanny that invites the observer to reflect on the random nature of the moment in juxtaposition to its orchestrated planning and programming. It is this irreconcilable conflict, perhaps, which is felt as uncanny.
It is the planned accident that anticipates uncanny synchronicity (a la Karl Jung)…perhaps this is one of the appeals of all improv.
I like how Lowry muses over Marynowsky’s work by recalling and contrasting it to the work of the surrealists, like Duchamp, who employed the uncanny but in a highly conceptual (and less experiential) way. Later in his review, he brilliantly writes that artwork like this is inherently more intimate, a symptom that the role of authorship [has shifted] toward that of artist as facilitator of experience.
Lowry considers the way technology works in relationship to this cultural shift, and muses over its implications:
The viewer now enters a relationship with a machine that extends the art action across time. Just as cinema once provided a new and strange way of experiencing life, the computer is now seemingly and endlessly extending our experience of the uncanny. “Notions of originality,” as French critic Nicolas Bourriaud put it in 2002, “…are slowly blurred in this new cultural landscape”…The challenge facing emerging and hybrid art forms is the need to generate and maintain a public.
Given Marynowsky’s integration of music and programming, and his blog’s features on musical bots, I would imagine that Marynowsky would be interested in David Byrne’s ‘singing robot’ project, and vice-versa.