[Update: This post is the first in an ongoing series about the Uncanny Valley. Click here to read them all.]
A couple of weeks ago when I was teaching Descartes’ Meditations, one of my students made reference to something called the “uncanny valley,” which I had never heard of before but which sounded really fascinating. So, I went home and did a little investigating to find out exactly what it was. It turns out that the uncanny valley is a theory by Japanese roboticist Masahiro Mori. Mori borrowed the term “uncanny” (Das Umheimliche, literally, the “un-home-ly”) from Freudian psychoanalysis, of course, where Freud theorized that we experience a profound cognitive dissonance when presented with instances of things that are both familiar and strange. That cognitive dissonance is expressed in the paradoxical nature of being attracted to, but at the same time repulsed by, an object. Here’s what the theory of the uncanny valley hypothesizes (from Wikipedia):
Mori’s hypothesis states that as a robot is made more humanlike in its appearance and motion, the emotional response from a human being to the robot will become increasingly positive and empathic, until a point is reached beyond which the response quickly becomes that of strong repulsion. However, as the appearance and motion continue to become more distinguishable from a human being, the emotional response becomes positive once more and approaches human-to-human empathy levels.
This area of repulsive response aroused by a robot with appearance and motion between a “barely human” and “fully human” entity is called the uncanny valley. The name captures the idea that a robot which is “almost human” will seem overly “strange” to a human being and thus will fail to evoke the empathic response required for productive human-robot interaction.
And here’s a graph showing the hypothesized emotional response of human subjects plotted against the anthropomorphism of a robot. The uncanny valley is the region of negative emotional response to robots that are “almost human. ” (You can click on the graph for a clearer image.)
My first reaction to this was something like “well, yeah, of course that’s true.” I don’t work with robots and I don’t really play video games, but I have seen the movie The Polar Express (with Tom Hanks, pictured above) and, as Mori hypothesized, I did find the “almost human” CGI animation in that film to be a little creepy. The idea here is that the uncanny valley seems like a strange anomaly– that is, it seems like we should feel greater affection and empathy for non-human simulations as they more closely approximate “real” human appearance and movement, but at some point (the uncanny valley), our empathetic reponse to them drops off dramatically and we become revulsed by the imitation. But why is this the case?
I can imagine several possible explanations, all of which are problematic in their own way, but my first intuition is to say that there is a fundamental “reality/appearance” distinction here that we are deeply invested, cognitively, in preserving. The one thing the chart above is missing, I think, is an indicator of what that line is and where it is crossed. What seems obvious to me is that there should be a “NON-HUMAN/HUMAN” divider, like the one I have added here. (See amended graph.)
This way, we can understand how the uncanny valley has its “upward” curve. Without it, there would simply have to be a radical “break” between the simulation and the real human being, with no gradual upwards approximation back toward the real. The non-human/human divide must intersect at the point of the human corpse, I think, which is the point at which whatever revulsion we feel at the “simulation” of the human– my hypothesis is that this revulsion is motivated by our aversion to “deception”– begins to be replaced by an affectively similar, but categorically different, kind of revulsion, i.e., the kind of revulsion that is motivated by our aversion to morbidity and mortality. So, the most real and life-like human simulation will provoke almost the exact same response as a corpse, and they will both be coincidentally located at the bottom of the valley of the uncanny.
BUT… when I tried floating my hypothesis to a few of my colleagues over dinner last night, one of them pointed out to me that I may not have accomplished as much with this explanation as I had hoped. Inserting the non-human/human (appearance/reality) axis certainly DOES explain how the uncanny valley can be charted as a “valley”– and not simply a precipice– but it still presumes that both the image (or simulation of the real) and “the real” itself exist on some kind of a continuum. Yet, as close as our affective reactions to an “almost real” human simulation and a real human corpse may be, there is nonetheless a radical categorical break between how we think about the two.
Up next: what does this have to do with the metaphysics of “race”?
Before anyone leaves a comment to this effect, let me go ahead and say that, yes, I HAVE seen the "30 Rock" clip that explains the uncanny valley in terms of porn and Star Wars. If you haven't seen it yourself, it's here.
I had seen that 30 Rock clip too, but I had totally forgotten about it until reading your post.
Fascinating, and I think I agree with Professor Grady — I suspect he may have brought Blanchot in here, esp. re: the image and the corpse but if he didn't, someone should.
But here's the question that immediately interests me: lets say for a second that at the very least we can "imagine" a line b/w the human and the non-human. To a certain degree we would even be justified in doing so in, for example, a strictly Platonic account of the human (particularly if we think of the human as a Platonic essence, something Socrates thinks we probably should do in the Parmenides at least and that therefore human beings participate in the form by being like, but also slightly different than that form).
The Platonic ideal of the human, what defines us as human, would be all the way to the right on your chart. We would all diverge a bit from that ideal (I suspect according to the "metal" of our soul and the extent to which our souls have been corrupted), and at a certain point there would be a precipice — not coming from the non-human but from the human.
But here the thing that defines us as human, or where we would experience the "essence of the human" most clearly would actually be the point where we most acutely feel the absence of the human, where we feel like there ought to be a human but there is not (I might make a dig at classic analytic philosophy a la Turing Machine here and suggest that the idea that we'd have to resort to intellectual tests to reach this theory is a sign of the autism of a certain kind of logicism). Hence our morbid fascination with corpses and C3PO (and porn stars, but that's a longer discussion). What defines the human is the line that seperates the absence of the human from the presence of the human, rather than some super-human or quintessential, hyper-present human essence.
And yet, the word uncanny should give us pause: both Freud and Heidegger understood full well that the uncanny was the name for a distinctively human experience. For both of them, it was the act of decisive violence that created the imaginary separation between the human being and the natural world. And yet interestingly Derrida (following Celan following Lenz) will try to invert the uncanny so that it goes in the direction of "monkeys and marionettes…"
So, fascinating stuff. If you and KG are game, we should put together a SPEP panel.
I was totally thinking about the metaphsics of race by the time I came to the last sentence of your post. Are there any good studies that racism increases or decreases the more someone seems to come to race of the racist (aka, are white racists, or white people in general, more or less comfortable dealing with people who are closer, but not quite close enough, to whiteness)? My guess is no. On the other hand, my guess is that violence is increased against those who exist in gender in the uncanny valley. Butch women, effeminate men, etc.
Such violence does seem to play out in terms of the human-animal distinction (but, only sorta). Extreme violence has always been utilized against those people who are seen as being closer to "the animal". Not only can we see this in the work of decolonial scholars, but also if we look at, say, Agamben. Bare life for Agamben occurs when a human is seen to exist in the zone of indifference between bios and zoe. So, in Homo Sacer, when discussing the bandit as werewolf, he argues, “That such a man is defined as a wolf-man and not simply as a wolf (the expression caput lupinum has the form of a juridical statute) is decisive here” (p. 105). This is also true of the analysis of Roberto Esposito in Bios, that the greatest violence by the Nazis were conducted against those seen as both human and animal.
Now, I might counter my argument with the obsession of talking parrots, great apes, dancing elephants and birds, etc. Though my guess would be that for some reason, these traits are not enough for most people to see these other animals as being in an uncanny valley with the human.
Anyway, interesting.