The Uncanny Valley and Avatars

The “Uncanny Valley” and Avatars

The word Avatar comes from Sanskrit avatāra a passing down, equivalent to ava down + -tāra a passing over. In Hindu mythology this is the deliberate descent by a god into the land of mortals, usually for the purpose of destroying evil or leading the righteous down the right path. Most often associated with the Hindu god Vishnu, though the concept has been applied to other deities. Similar concepts are found in other religions – one thinks of the Angel Gabriel for example. The word then came to be used for a computer representation of a user and this dates back to at least 1985 when it was used for a character in the Ultima series of computer games. To begin with the goal was to become ‘the avatar” but later games assumed you were an avatar. There are many types of ‘avatar” varying from a small square-shaped area close to the user’s forum post where the avatar is placed to a much larger representation of a person/creature which can be customized. These avatars can be realistic representations of who the person is and what they are doing (helped by motion-capture technologies that mirror every movement) or they can be strategically altered to show only what you want them to.

I first came across the concept of the “Uncanny Valley” when looking at the work of Alan Warburton, following one of my tutor’s recommendations.   In his video work Goodbye Uncanny Valley (2017) , Warburton looks at the background and current state of  of photoreal CGI. In his introductory statement he writes – “computer graphics have conquered the Uncanny Valley, that strange place where things are almost real. .….but not quite”

Subsequently, whilst searching for links to the history of dolls, avatars and robots to gain more background context for my assignment, I came across this concept again and discovered there is more to it. Information was gained to begin with from a comprehensive a Wikipedia entry and I then did further research. This concept was proposed by Masahiro Mori, a robotics professor and translated from Japanese to “uncanny valley” in the book Robots; Fact, Fiction, and Prediction, written by Jasia Reichardt (1978). This is linked with the concept of the ‘uncanny’-  Ernst Jentsch and Sigmund Freud (see an earlier post here). The hypothesis is that observers’ emotional response to a robot becomes increasingly positive as the appearance of a robot is made more human until it reaches a point beyond which the response becomes revulsion. This reaction reverses as the robot’s appearance continues to become less distinguishable from a human being.  The ‘uncanny valley” is thus that in between stage. Presumably it varies between individuals. Technologies have advanced greatly since the late 1970s, with more sophisticated robots and lifelike Avatars but the “Uncanny Valley” still needs to be considered and overcome. This YouTube video from Mashable explains further

Charles Darwin noticed a similar effect when looking at a particular snake, referring to the relative shapes and positions of the pupil of the eye, jaws and nose . the writer Jamais Cascio also noted a similar effect that occurs with body modifications beyond what would normally be possible. Examples are also given from several films using computer-generated imagery (CGI). This made me think as well of the film Avatar.  It took me quite a while to get used to the faces of the Na’vi beings. It just occurred to me though that, maybe, it’s that when a robot’s face most resembles a snake’s head that this revulsion occurs and is due to a primal recognition of danger in a snake shape – maybe even the story of the serpent in the Garden of Eden connects here.

Mori’s theory was examined more closely in a paper by E. V. Pujals and N. Buffington, Secrets of The Cabbage Patch: Pediophobia and The Fear of The Inanimate (2007 . their view is that Mori’s hypothesis can relate to all inanimate objects that, in one form or another, become or seem to be animate. This is linked to a perennial human desire for playthings to become truly alive. Examples are given of the myth of Pygmalion who fell in love with a statue he sculpted and the goddess Aphrodite brought it to life in response to his prayers. This fantasy can also be found in Hoffman’s the Nutcracker and the Mouse King (1816); Pixar’s Toy Story.

This fantasy of toys becoming alive is rooted in the child’s emotional attachment to playthings and the way in which their ways of relating thus ‘prove particular to their respective individual needs’, which I would say are also influenced by social/cultural reinforcement, given that the authors provide the examples of baby dolls in the early 1900’s promoting nurturing and companionship tendencies connected with ‘mothering’  whilst in the early 1960’s Barbie dolls were seen as a symbol of freedom from future responsibilities of motherhood. Also, ‘Star Wars figurines allowed boys to play God in orchestrating miniature worlds of high tech adventures’. Pujals and Buffington go on to contend that, despite this strong attachment to playthings, literature and films ‘perverted the toy and created disquieting versions of the same lovable characters that inspire fantasy’, even in Toy Story with its reanimation of dismembered toys in one particular scene – the threshold of horror for Sid (the sadistic boy) being when Woody speaks to him using facial expressions.

However, Pujals and Buffington perceive limitations in Mori’s hypothesis in that adults and children’s fears are not necessarily linked; they have different perspectives on life. The fear that Mori’s hypothesis explains, could suggest that it is due to the great awareness of adulthood, that adults are more afraid, afraid of the unknown because their knowledge does not extend to the areas that concern the respective unknown or afraid of the already known because they understand the full gravity of its implications.  Adults have a need for order and search out boundaries. They provide evidence for this, e.g. Andy, the child, is unfazed by Chuckie coming to life.

Design guidelines have been written including attention to the proportion of facial features skin texture detail and changing features to a more cartoon-like style also eliminated the uncanny (Tinwell et al  2010:05). As more and more characters now appear in animation and video games factors such as form of motion, sound features, timing and facial animation (particularly in the upper part of the face) also contribute to an uncanny valley effect. Tinwell et all reported that  “The uncanny may be related to the importance of being able to swiftly and accurately detect the emotion being expressed by another as it helps us to predict their likely behaviour……” the results of the current study also provide fairly compelling evidence that perception of the uncanny in virtual characters displaying inadequate facial animation is greatly influenced by the type of emotion the character is portraying ( 2010 :25) they concluded, “Overall, the results indicate that attempts to embed truly authentic and convincing human-like affective signals in video game characters still has some way to go”.

The study referred to above focussed on video games. Below is a video looking at the way in which film-makers work to find ways around the uncanny valley effect.



Pujals, E.V. and Buffington (2007) PWR 2; Rhetoric of the Monstrous accessed at
Tinwell, A et al (2011) Facial Expression of Emotion and Perception of the Uncanny Valley in Virtual Characters, Accessed through )