Originally developed for use in robotics, the “Uncanny Valley” hypothesis has increasingly been applied to film as computer animation and make-up effects continue to advance. Defined in 1970 by Japanese roboticist Masashio Mori, the term attempts to explain how people perceive and identify with robots in terms of how much the robot resembles a human. As the robot appears increasingly human, our emotional response to the creature increases in tandem and we identify and empathize more and more with the robot. This holds true until the artificial creation appears so human-like that our response quickly reverses to one of revulsion. Mori argued that artificial creatures which appear extremely, but not completely, human provoke unsettling responses as the viewer finds the robot strange and “uncanny.” These robots come close to reality but are clearly artificial. The actual name for the theory stems from the sudden dip or “valley” in the chart (shown below) plotting our “familiarity” and emotional response to a robot, against an increasing amount “human likeness.” Maybe it’s best if Frank and Tracy from “30 Rock” explain in this home-made video (complete with laugh track):
Frank: As artificial representations of humans become more and more realistic, they reach a point where stop being endearing and become creepy.
Tracy: Tell it to me in Star Wars.
Frank: Alright. We like R2-D2, and C-3PO. (Frank points to the left of the valley)
Tracy: They’re nice.
Frank: And up here (to the right of the valley) we have a real person like Han Solo.
Tracy: He acts like he doesn’t care, but he does.
Frank: But down here we have a CGI stormtrooper or Tom Hanks in the Polar Express.
Tracy: I’m scared! Get me out of there!
The cinematic appropriation of the term seems increasingly apropos as studios continue to develop stories centered around non-human characters. The Uncanny Valley helps explain why at first glance we like Wall-E and The Iron Giant (modestly human-like, very endearing), but are uncomfortable with the apes in the original Planet of the Apes (1968) (almost human, unsettling appearance). This is intentionally used as an effect in zombie and monster movies (almost human, very uncomfortable) obviously along with other techniques (gore/violence/etc).
In many ways, the theory formally defines what many viewers already understand. The characters in some films are flat-out disturbing, even as their artists attempt to create a more human-like image. The Polar Express (2004), Beowulf (2007), and Tron: Legacy (2010) provide us with notable failed attempts. Even the almighty Pixar found themselves guilty of violating the uncanny rule in their early short Tin Toy (1988). Some argue that Sid from their beloved Toy Story (1995) teeters precariously close to the valley’s edge.
Interestingly, while films like The Polar Express and Beowulf made earnest attempts to advance the field and create a more perfect digital human, other studios found alternative means to circumvent the problem. Instead of chasing down more accurate depictions, they create a more endearing image by the application of Neoteny. This biological term refers to the physical juvenilization of the adults in a species which tends to make the adult appear more “likable.” When applied to design, the use of neotenic traits means featuring large eyes, a large head, a flattened face, and other associated changes. These neotenic traits have been commonly employed in traditional animation for some time.
Many film critics and historians theorize that modern visual effects may have reached the point that computer generated characters appear so human-like that we are exciting the “uncanny valley.” Our familiarly and identification with the characters is increasing, higher even than with Wall-E. They cite the newly released Rise of the Planet of the Apes as evidence for our near-exit from the valley. These critics are also part of the group arguing that Andy Serkis deserves at least an Oscar nomination for his work. As a leading pioneer in the field of motion capture CGI, he might deserve more.
Thanks for reading,