In the movie “Face/Off” when John Travolta and Nicolas Cage stared at each other following face-switching surgery, we were meant to understand that both men had a very uncanny feeling. Seeing a copy of oneself that isn’t quite oneself elicits an emotional response robotics experts call the “uncanny valley,” a phenomenon psychologists have been trying to explain in human terms for decades, beginning with Freud himself. Seed’s Joe Kloc explores the history and science of this strange phenomenon in “Into the Uncanny Valley.”
Seminal child psychologist Jean Piaget saw infants as unthinking bowls of Jell-O. Touched by the world around them, they sense their own jiggling, which causes nascent perception—leading to experimental jiggling, more complex perception, and so on—until repeated and recursive interaction with the self and the environment transforms them from a bowl of Jell-O into a thinking adult capable of, well, having strong opinions about Jell-O, among other things.
C.S. Peirce saw the human mind as a habit-making machine. Novel things (situations, actions, objects) elicit an emotional reaction, often discomfort. As the novel thing becomes old and the perceiver inured, the thing becomes part of the background, the idea internalized, the physicality automated. Symbols become mere icons, startling meaning becomes mere expectation, overtly conscious action becomes subconscious reaction.
In either model that which does not belong is attended to or focused on, and there is a relationship between “belonging” (to one’s view of the world) and the attributes we observe. But when something is supposed to fit expectations perfectly but in fact does not fit in some unexpected way, discomfort happens. For instance: Your father has a mustache, your mother does not. Then one day, your mother has the stash and your father is clean-shaven. It is only one tiny detail among many details. But it would be enough. It would be…uncanny.
Freud attributed the uncanny to the outcome of a human attempt to deny death by making copies of oneself which, when viewed by the original, seem eerie. But recent research suggests that monkeys regard nearly perfect monkey manikins as strange, just as humans see modern very human-like robots as strange. If monkeys are creeped-out by slightly off monkey-doppelgangers, than Freud’s explanation of the uncanny is unlikely. But monkeys probably also start out life as bowls of jiggling Jell-O, and certainly they attend to the novel and ignore that to which they are inured like Peirce would expect. So the uncanny could be the result of a sudden and unexpected wrong turn in the normal process of perception and learning, by people or monkeys. Kloc’s article discusses a wide range of other possible explanations, including more straightforward adaptive models. Perhaps, it is suggested, the uncanny valley is an unsafe place to be so we are shaped by evolution to recognize it.
The uncanny, not-quite-perfect duplicate of ourselves (or our loved ones) is the stuff of so much fiction and myth that it must be powerful and may even be important. The near-duplicate human is a tool of shamans, the immortal spirit of religions, the face in portraiture, the death mask. And it is also the bane of those who wish to create robots with artificial intelligence that can stand in for individual humans even if they are not “real.” We seem innately to not trust our doppelgangers or our Frankensteins. We feel strange when we see human faces on the backs of crabs or human heads growing out of places they should not be. When pressed for a rational reason for this distrust or discomfort, it seems sufficient to say little more than, “Well, the whole thing totally creeps me out.”
And when we find ourselves able to satisfy our questions about a phenomenon with such a meaningless and irrational statement, that means we are probably on to something very interesting.
Originally published November 16, 2009