Week 7 of the “Image” section of the U Chicago Humanities Core course “Media Aesthetics” is devoted to Freud’s essay on the uncanny. It’s a tough week: students have to work through the usual trickiness of disentangling Freud’s ideas from the ideas he cites from Ernst Jentsch, and, on top of this, the week has to act as a general introduction to psychoanalysis, which will return in the form of Laura Mulvey’s “Visual Pleasure and Narrative Cinema.” After a tough lecture in the first class meeting of this week, I decided to give my students something that was both more fun, and more hands-on. For the week’s second meeting, I gave them small group work based around Mori Masahiro‘s concept of the uncanny valley.
My reading for this week is simply Mori’s original article “The Uncanny Valley.”[i] This proved to be unusually light reading for this particular course (it is only three pages!), so rather than explicate the material with any sort of lecture, I just asked students some questions to gauge comprehension. I asked what the two axes of Mori’s famous graph were, expecting students to be able to answer affinity and human likeness.
I asked students what central claim does Mori made about line along these axes represents. The answer is that it is not a monotonically increasing function—rather, it has a fairly notable “valley.” I asked them to identify this valley, and why Mori thinks it occurs. Mori thinks that the uncanny valley is a result of a certain approximation of humanlike form, without a full approximation of humanlike movement or behavior. This reminds us of corpses (among other things).
Here, I shifted from basic competence questions to more theoretical questions, connecting Mori into the Freudian themes of the week. In what ways, I asked, does Mori’s notion of the uncanny valley connect (or fail to connect) with Freud’s writings on the uncanny? Here, I hoped that students would point out that both figures are interested in corpses as bearers of uncanniness. Both also address confusion in the line between the animate and inanimate (although Freud, in the end, largely dismisses this as a site of uncanniness, as part of his break with Jentsch). I also asked in what ways Mori’s conclusions show the ways in which thinking about the mind had changed between 1919, when Freud was writing, and 1970, when Mori was. Here, I hope to call attention to psychoanalysis’ trademark interest in the unconsciousness and repressed obsessions/anxieties, especially prominent when contrasted with Mori’s approach, which hews closer to a sort of proto-evolutionary psychology.
Finally I asked: Do we consider robots media? Why or why not? How do they intersect with definitions of media we’ve entertained so far? I didn’t have anywhere in particular I wanted to go with this question, but given the various definitions of media we had taken a tour through in the course, I thought it would prompt interesting discussion.
After this free-form discussion, I broke the students down into three groups. I assigned each group four YouTube videos, based on the categories of “robots,” “videogames,” and “CGI movies.” I told each group that they were going to present on these videos, and passed out a worksheet with several questions they should address when they discussed these videos in front of the class. The questions were as follows:
- Where do you think each of the clips in your category falls on Mori’s scale of human likeness/affinity? Do they fall before the uncanny valley, or directly into it? Do any of these attempts at portraying humans reach Mori’s peak on the other side of the uncanny valley?
- What are some of the contributing factors to your placement of these attempts at representing humans on the scale? What factors make it more likely for something to fall into the valley? What make factors make it more likely for something to avoid it? Talk about the factor of movement specifically, as that’s something that Mori himself singles out.
- To what effect does the uncanny valley rely upon the intended aesthetic effect of a scene or demonstration? For instance: are attempts at sentimentality or cuteness that go awry particularly ripe for uncanniness? Why, or why not?
And, with that, students began watching their own cluster of videos. Here’s what I assigned:
Group 1: Robots
The Hall of Presidents (Walt Disney World, 1971 – ):
Repliee R1 (Osaka University Robotics Department, 2008)
Actroid-F (Osaka University Robotics Department, 2011)
Actroid (Kokoro Compand, Ltd., 2013)
Group 2: Videogames
D, ending (WARP, 1995)
Super Mario 64, main menu screen (Nintendo EAD, 1996)
The Elder Scrolls IV: Oblivion, dialogue sequence (Bethesday Game Studios, 2006)
Nvidia Face Works Tech Demo (Nvidia Corporation, 2013)
Group 3: CGI Movies
Final Fantasy: The Spirits Within, dialogue scene (Hironobu Sakaguchi, 2001)
The Polar Express, musical number (Robert Zemeckis, 2004)
The Incredibles, clip (Brad Bird, 2004)
TRON: Legacy, old Jeff Bridges vs. “young” Jeff Bridges (Joseph Kosinski, 2010)
I drew a reproduction of Mori’s graph on the board, and, as the groups presented, encouraged students to discuss where they would place the subject of each video. The group presenting had the final say as to where to put things. Here are their results:
Some of the videos ended up precisely where I thought they would. The cartoony style of Super Mario 64 and The Incredibles helps to insulate its characters against uncanniness, and, as I expected, students placed them on the right side of the graph, before the valley. The fact that Polar Express‘s children and Repliee R-1 were positioned at the absolute lowest point of the valley wasn’t a surprise, either. (This provoked a good conversation as to why representations of children gone wrong seem to be especially unsettling.) I was surprised that students placed Actroid-F and Disney’s presidential automatons on the “upward swing” of the valley—I wouldn’t have predicted that!
[i]. Mori Masahiro. “The Uncanny Valley.” Translated by Karl F. MacDorman and Norri Kageki. IEEE Robotics & Automoton Magazine. June 2012.