(I don’t speak German; the title is what FreeTranslation.com came up with for “The Uncanny and Video Games.”)
Here’s the problem with human character models in modern video games: animators are just good enough to get themselves into trouble. Masahiro Mori, a Japanese roboticist, calls it the uncanny valley: the point at which things aren’t quite human and aren’t quite alien. Anything stuck in the valley tends to freak the hell out of us.
It’s no surprise that a roboticist coined a term describing something in the interstice, or the narrow space between, human and non-human. The more human something seems, the more humans tend to identify and empathize with it — up to a point. That point is Mori’s uncanny valley, and robotics is very interested in getting its machines over Mori’s valley and into ‘real’ humanity.
Wikipedia has a relevant chart, if you’re curious. I don’t agree with some of its data points but it’s a pretty good visual representation of what Mori’s talking about.
Mori’s valley matters to video games because modern animation is beginning to try and depict human beings, not just represent them. The difference is that representation doesn’t usually approach the uncanny valley; it’s only implies the idea of a person. Depiction, on the other hand, is an attempt to reach photographic levels of detail.
How’s depiction going? Take a look at Kaim, the protagonist from Lost Odyssey, a brand-new high-profile RPG from Microsoft.
This is a game with two (and only two) hooks: its graphics and the basic humanity of its characters. In all other respects, Lost Odyssey is utterly traditional and unremarkable.
Does Kaim look human to you? Is he standing like someone holding a weapon should? Most importantly: does his expression communicate anything? To me, he looks bored — and awkward. Microsoft would have done better with an ink-drawn still or a watercolor portrait; at least there, the artists wouldn’t have had to worry about forcing the character’s (very real) humanity through a stiff character model.
(As an aside, I’m playing through Lost Odyssey right now and really enjoying it. Its plot is unremarkable, but its characters are incredibly engaging — and never more so than during the game’s flashbacks, which are nothing but text printed over indistinct backgrounds. If you love novels, check it out; everything good about Lost Odyssey hinges on its text.)
By contrast, check out the cast of Final Fantasy 6, a seminal Squaresoft RPG which came out in 1994. The left-most portraits under each character’s name are the ones I’m specifically interested in. Almost every piece of dialog in the game had a character portrait next to it. They didn’t change based on the character’s emotional state at the time, which was part of the game’s strength. By giving the player one snapshot of how a character looked, then embellishing it with text and other abstract sprite graphics (also on the linked page), Square allowed players a lot of freedom to interpret the characters.
That freedom to interpret, by the way, is why so many players felt particularly connected with Final Fantasy 6′s cast. Players felt personally involved in the characters because they were allowed (or forced) to fill in the blanks themselves.
As a final example, check out modern Mario. Nobody’s going to mistake him for a real person, but nobody’s going to feel that he’s somehow subtly wrong, either. Moreover, as a representation and not a depiction, Mario can convey emotion without having to render it realistically.
I don’t doubt that, some day soon, we’ll be able to realistically animate human models and faces. Until then, however, video games — and all related media — should stick to human representations, not depictions. That way, their characters’ half-baked humanity won’t get in the way of good storytelling — something video games are already struggling with.
Genghis Philip
Gordon Levine
i think that this same principles extends to all human depictions in any artistic form, from charcoal drawings to literature. if something is a caricature representation of an object or form in the real world, we understand that it’s a symbol meant to be representative of that reality, not a perfect reflection of it. if a character in a novel, for example, is depicted in a realistic setting, and with realistic troubles and wishes, but says and acts in ways that are not realistic, but rather idealistic and emblematic of reality, we complain that the character is weak and two-dimensional.
example: my problem with “batman begins” – here we have a director and writer who are trying to pull what has always been a campy character out of the campy setting and into a more realistic world, but unfortunately, the eccentricities inherent in a millionaire who wants to don bat ears to scare criminals make the character of bruce wayne in “batman begins” ill-suited to his new, (semi) realistic environment. as a result, i can’t immerse myself in the movie, because reality has been attempted, but not achieved.
Mori’s valley is definitely a problem for any art that tries to depict human beings. I almost added a paragraph at the end of the original post noting that that’s probably why anime is so popular — it’s highly stylized and representational.
Good point about literature having the same problem, too. I hadn’t consciously made that jump. Of course literature is caught between the same rock and hard place of depiction and representation. I guess I’m too invested in representation to remember that some authors try for depiction.
I liked Batman Begins. I don’t think Christian Bale wasn’t the best choice for the role, though; somebody more emotive — specifically angrier — would’ve worked better.
I wonder if this is part of why people are drawn to sci-fi and fantasy: they know that depiction can never be perfect, and so they move to the opposite end of the spectrum, where people are cardboard cut-outs and the plot is easily recognizable within the first few pages. And the move of recent sf/f toward holding those archetypes within a story that’s a lot grittier and a lot more recognizable to your average person seems to be an attempt to reconcile the two.
The other problem with depiction is that it depends on people recognizing it – take, for example, To The Lighthouse. These are not feelings or people that, in specific, we would recognize as being kin to our American, 21st century perspective. It’s the overall representation that lasts – everyone knows someone like Mrs Ramsay in general.
Representation appeals to our very human need to see patterns, to be able to classify people into groups. Literature that tries to take that away from us… I just can’t see it working. We will, by necessity, start saying “oh, I know someone who has that trait. And that!” and so characters become corollaries to real life.
I’m not quite sure how to relate this back to the art side of things, mainly because I am not an artist who works with visual media. But! Have thoughts.
(I have now commented, Gordon. So there. :P)
Ohyes. Also, a video that pretty much exactly covers Mori’s valley. So. Creepy.
Here.
Very nice post. This Mori’s Valley is an interesting way of looking at some of the difficulties facing games today. It certainly helps explain why people are actually very, very willing to accept technically sub-par graphics. Perhaps a portion of the Wii’s strength in sales is that its unimpressive graphics do not try to blur the line between depiction and representation.
I certainly think that many games, particularly the (as of recent) ill-fated PC game industry, have tried way to hard to work towards depiction. I don’t understand what video games gain from it. They are almost entirely constructed realities. The only things that are directly attributable to reality are probably the voices of characters. Everything else is a construct. Why try to make games emulate our reality? Let them create their own world.
I’m glad to see some people doing theory on video games!
Too bad more people don’t feel the same way about using their imagination to construct characters in games. At least people in the past could argue they were reading while they played RPGs, now almost all of them have voiceovers (which often totally blow). The value of interpretation is why Nintendo has never done voiceovers for Link, let alone any text. Do you think a character should at least have text or does that also impose ‘violence” over the character? Then again, the emotion held within Link’s face in Ocarina of Time, Wind Waker, and Twilight Princess have given the games the emotional depth that make them into such favorites.
Even games like Mother 3 (only in Japanese at this time), which has comparatively poor graphics, has tremendous emotional pull because of the music and character direction. Personally, I think text/voice are less important in (most) video games. I’d prefer a move more toward a cinematic experience in which the player is in control–for instance the feedback of the rumble, the modification of music as you enter battle, the hallucinations as you lose health (as in Eternal Darkness).
I’ve been more interested in the representational values of “violent” games, such as those that immerse people into the battlefields of WWII and Iraq, or in the case of Grand Theft Auto, a made-up city. Do those games present an uncanny experience? Do first person shooters create more of one than a third person shooter. Ho do they construct/represent war, violence, and the “enemies”? Do they sublimate the “death-drive,” carry one through a healing experience (as Kristeva describes horror lit), or merely tools to control the masses minds as Adorno might have it?
A-Dubs
Well, Gordon and Genghis,
I’m wondering – would either of you be interested in being contributing writers for a ‘zine about video game theory?
I went to a wonderful little self-publishing center in Portland, and I had been thinking about some sort of project dealing with ‘theory on video games’.
I probably wouldn’t make very many copies unless it caught on (like 50-100), but I thought you two might be interested.
I’m always happy to, Mattsmith, but you’d need to give me a heads up on time and length, etc.
Consider this a heads-up to start thinking about you would like to write. I’m probably looking at 3-5 pages single-spaced. I don’t know about a deadline yet, but I’d like to hear your ideas first. My e-mail is smithsm1984@gmail.com if it’d be easier to communicate that way.