With bright red lips, rosy cheeks, and painted smiles, clowns are creepy. They make us balloons shaped like puppies, dance, and tell jokes — they even come out of small clown cars, all to make us laugh. So why are these (supposedly) beloved clowns so unnerving?
It’s because of the “uncanny valley” — the term describes the revulsion humans have for things that seem like people, but aren’t quite what we are used to, a la clowns. They look like us, but their fake smiles and teardrop eyes don’t quite fit what we expect most people to look like.
One theory about why the uncanny valley exists is “pathogen avoidance”: we avoid things that don’t look “right” because it could mean they’re carrying some sort of disease that we could contract.
But the uncanny valley doesn’t exist for every human-like being; there’s a point at which things become unnerving to us:
Puppets and prostheses don’t disconcert us, since they’re obviously not human. But once we start getting into the realm of zombies, that’s when we begin to experience the uncanny valley. Humanoid robots creep us out, because they sorta look and move like real humans but not quite. Industrial robots don’t unnerve us because they’re obviously not human or humanoid.
Animation has also lent itself to the uncanny valley. “The Polar Express” features some cute kids that made people uncomfortable. People were unnerved by these faces and bodies that were supposed to represent real humans, but didn’t quite showcase real-enough faces.
Regardless of the theories, why the uncanny valley exists is a real problem researchers are trying to figure out. There is a general consensus that anything trying to mimic human behavior causes the repulsion in people. If an android has jerky movements or doesn’t make good eye contact, the uncanny valley will be in effect. This is likely because the expectation between how something humanoid should look and act and how the humanoid actually looks and acts causes a disconnect in the humans’ brains.
A lot of this disconnect has to do with motion and perception. Because of the brain’s “action perception system,” it is alert to human motions and appearances. Anything that doesn’t fit what the brain is used to (real humans, most likely) makes the brain uncomfortable: it can’t process the information well.
A team of researchers, led by Ayse Pinar Saygin at the University of California, San Diego, ran a bunch of fMRI studies of subjects, aged 20 to 36, that had no prior experience with robots and were shown videos of three different conditions:
- “Actroid” (an android actress)
- Human
- Android stripped of skin and only showing mechanical parts
Each of these situations had the “actor” doing normal, everyday things, like picking something up from a table. Using brain imaging, the researchers determined that there were differing brain responses in each condition, especially in the parietal and visual cortices. That is, the brain areas associated with movement perception and vision. When the actroid was shown, the brain couldn’t process the information well. When the mechanical robot or the human was shown, the brain did just fine.
This makes some sense, since there’s no evolutionary need for a human brain to care about bio-like appearances or biological motion. What the brain is looking for, however, is a match between appearance and motion. Dogs should walk like dogs, jaguars should run like jaguars, humans should walk like humans, and androids, well, they just move weirdly. They look like people, but don’t move like us. So our brains can’t process that, and we feel repulsed.
In the future, as androids become more human-like to the point where the brain can barely perceive differences between a humanoid and a real human, the uncanny valley probably won’t exist for the people interacting with these advanced androids. So long as the android sports human congruent motion and appearances, the human brain will be just fine.
This of course begs the question: what happens with society begins to include androids in everyday interactions, from your office receptionist to your family physician? Would we be able to connect with them — even empathize with them? Could our society grow up with these androids, develop intimate relationships with them, or even have lives centered on them?
