The Animator Scientist

Call it a crisis of faith. For the past 20 years, Philippe Vaucher followed the 12 commandments of animation religiously. Not blindly, mind you, but because as an animator, he knew they worked. He used them when he directed Chasse Papillon (The Song Catcher) (2003) with the NFB/ONF and Le puits (The Well) (2013) with Unité centrale. He teaches them today as a professor of animation at Université du Québec en Abitibi-Témiscamingue (UQAT).

But he never really knew why. We spoke at his office in downtown Montreal, where I found myself distracted equally by his books on animation, cognitive science, and his drawings on the wall. Philippe described his doubts, not about the efficacy of the 12 rules, but about the science that made them work. 

“I found that when you teach animation, you always use the 12 rules of animation. They’re simple and they work. Of course, it takes 12 minutes to explain them but 12 years to master. They’re useful, great, and simple but I was always afraid someone would ask me, ‘Why do they work, why do I have to exaggerate a movement, why does that work.’

Even as a student animator, I remember one of my first animation exercises was a walk cycle. And I realized that Muybridge did that, he got it, so I decided to just trace a Muybridge 8-photo walk cycle. But when I looked at my final animation, it was stiff. So I asked myself, why is it stiff? Why does it look like a robot? This made no sense. I traced it as an ‘animation’ of a real person moving but when I watched it, it didn’t seem real. Cut to 10 years later when I’m teaching animation class, I started asking myself why do I have to exaggerate to make it real, to exaggerate a movement not to the point of it looking ridiculous, but amplified. Why? It made no sense to me? And part of me was worried that one day a student would ask me these questions that I really didn't have the answer for.”

Then, motivated by the ongoing curiosity - and the practical perks of having a PhD in order to secure tenure - Philippe wanted to understand why the rules set forth by Walt and his disciples work in terms of the viewers’ perception and thinking.

“I started reading animation articles and books and people would kind of talk about it, but strangely, a lot of the writings on animation don’t talk about movement in animation, at least not that I could use in my classes to explain the 12 principles. So, when I had to think of a PhD topic, I was like, okay, I think this stuff needs answering. So that’s how I decided to get interested in the specific area of research which is cognitive perception and try to find explanations for some of the principles of animation.”

And that’s when Philippe discovered that his questions weren’t really answered by cognitive psychologists. Most research on visual perception in cinema and animation dealt with image properties like color, light and form. There was little in terms of motion that could relate to the 12 principles.  I asked him what gets exaggerated.

“That’s a good question. I guess returning to the story in The Illusion of Life, when Disney was telling the animators to make it more lifelike, more real, he meant more in the spirit of life, more exaggerated.”

So what exactly gets exaggerated (me being a stickler for precision)?

“In animation, I guess it’s everything. Everything’s kind of amplified, especially in the form. Both the comments in The Illusion of Life and my research are starting to show that all of the principles of animation have to do with exaggeration. Appeal is an exaggeration of [personality], anticipation is an exaggeration of a pre-movement, or follow through on overlapping action on exaggeration of a shot. It’s as if all the principles are put on dials and you set real life on ‘5’ and you kind of want to put it on ‘6’ or ‘7’ and in cartoons, or ‘8’ or ‘9’ to really amp it up, so where it is on the dial is a level of exaggeration.

I have a theory or a hypothesis that the level of exaggeration is something that every animator does on a practical level depending on how realistic the style is of the appearance, the shape, but also how that character moves. So in a cartoony character, you’re going to want to amp that movement up. It’s like in order to convince you that the cartoony Bugs Bunny character is angry or cute, you’re going to want to exaggerate the form and movement to convince people of what the character wants to say. Like, ‘Nobody believes I’m real, so I’m just going to talk louder and move my hands more.’

In my personal experience, I remember I wanted to animate this character [the husband in Le Chasse Papillon] in the scene where he’s just snoring. I animated him falling asleep, snoring before a dream sequence. The character was falling asleep and I wanted to show him snore, and I remember animating his mustache move when he breathed…. I guess I over-animated the mustache [exaggerated the motion of the whiskers] and it ended up looking like weird tentacles. It looked so ridiculous because in my visual style, the character was more realistic, so that exaggerated move on the mustache just looked weird. It didn’t work and I had to dial it back a bit.”

So the world that you’re creating with your style, whatever that style is, whatever the logic of the world that you are creating is, that’s going to determine what you mean by exaggeration.

“Absolutely. And I think every animator knows that. And even in a story, the comic relief character will be exaggerated, or if it’s the villain, the villain will be exaggerated. That’s why motion capture doesn’t always work when you think that because you’re using real motion, you can just plug and play it. Every animator knows that motion capture never works straight out of the tube, as a simple plug and play kind of thing.... .I believe that for the actor being filmed for motion capture, the action has to be exaggerated a bit.”   

But Philippe believes that the principles may also involve how the treatment of motion impacts the viewer’s recognition of a character’s emotion. Indeed, it’s anthropomorphism – the ascription of human emotion and, above all, intention, to non-human beings – that sucks us into the characters and narratives in animation. How motion is used to create a believable being, regardless of its form, led Philippe to investigate the notion of the “uncanny.”

The “uncanny valley” was a term used by the Japanese robotics professor, Masahiro Mori, to describe the limits of belief and appeal in human-like automatons. The term “uncanny” was used by 19th century German philosophers, continued into early 20th century psychology by the German psychiatrist, Ernst Jentsch, and elaborated and popularized by Sigmund Freud. Freud was referring to the feeling of horror or repulsion when experiences we thought as being in the past, hidden or repressed, suddenly appear in the present reality. Mori extends this notion to human appearances. But Philippe notes that Mori and others dealing with the “uncanny valley” rarely address “motion.”

“In animation, it’s when we see a character that’s not quite realistic, almost real life but not quite, so there is something we feel that’s off, and kind of revolting, seeing something that’s almost but not quite right. There are a lot of theories of why we get that feeling, and scientists have tried to identify what causes it. But they have mostly been looking at form, concentrating on the appearance of the character. In part, I think it’s because it’s easier from a scientific point of view to control, measure and test a character’s visual appearance rather than test a character’s movements. That’s kind of easy.”

There’s not a lot of research on what’s the impact of a not quite like realistic movement on a viewer. And it’s weird because Mori in his original articles believed the effect of the uncanny valley would be increased with movement…. That’s why it’s an interesting area of research. It’s also tough to study because when you’re animating something it’s hard to animate something badly (but believable). Like what constitutes badly? I can see why scientists have stayed away from this because it’s really a tricky question. Most of the stuff in animations that we associate with the uncanny have to do with appearances, like with the Polar Express, where critics said that viewers were freaked out because the characters looked uncanny. Maybe. But it could also be in the way it used the motion capture.”

There’s a historical link between the Nine Old Men and Disney to the roots of animation. Some of the principles were already outlined in E.G. Lutz’s Animated Drawings, the 1920 book that Walt Disney famously read and referenced. Much of his book references “sight gags” and depictions of humorous motion. Significantly, the rendering of motion within the context of “cartoon” animation was tied to the then current theatrical showing of animations along with or accompanied by vaudeville acts.

Ironically (and possibly, uncannily), it was only in my re-reading Animated Drawings more closely for this blog, did I see a name I seem to have missed, and thought I would never see in a technical manual of animation: Henri Bergson.

Henri Bergson was one of the first philosophers to use the cinematic apparatus – the technologies of recording and projecting a sequence of still images – as a metaphor for how humans see and think about – and most of all experience – events, space and time. (A machine metaphor for thought continues with today’s use of “computers” as a metaphor for human intelligence).  One of the influential takeaways from Bergson’s philosophy was the difference between “objective” measured and mathematical time and “subjective” experienced time, and between discrete moments and temporal flow. The fluidity of “durée” – duration as lived time -- had an impact on Continental Philosophy, Gilles Deleuze, the entire field of phenomenology, film studies, and even physics.   

Lutz describes Bergson’s “treatise” (probably, On Laughter: An Essay on the Meaning of the Comic) in explaining the comic effect. I don’t know if Lutz read Bergson’s book in the original French (published in 1907), or one of its English translations at the time (1911 or 1914). But Bergson wrote about laughter in the context of (then) modern technologies, of the machines that have been injected into our lives. Lutz summarizes Bergson by describing “one law of the ludicrous… is that the human body appears laughable when its movements give a similitude of a machine in operation.” Lutz then expounds on this by explaining that this effect is probably because “machine-like movements in organic bodies amuse us because of the rhythmic, orderly, or periodic occurrences of these movements themselves…” (p. 230). Lutz follows this up with drawings of men comically spinning and running.

While Lutz was focusing on the motions themselves as being funny, Sergei Eisenstein revered Disney’s “plasmatic” cartoons as being subversive and irreverent comments on mechanized, ordered, modern society. The humorous appeal of Disney’s anthropomorphic animals that have been mechanized is also highlighted in Russell Merritt’s and J.B. Kaufman’s 1993 book, Walter in Wonderland. Here, the authors comment on the “hyperkineticism” in Disney’s early cartoon motion.

It’s one thing to have an animate human or animal-like object move mechanically when there is some obvious hardware present. Or demonstrate unusual, in a sense, exaggerated, flexibility, speed and agility – as do dancers, mimes, and Jim Carrey. We readily accept that a human can act entirely illogically and literally “out of mind,” or “losing control.” But animated human characters – (or androids, as Mori pointed out) that move in “un-humanly” ways cross a boundary that provokes an uncomfortable eeriness or repulsion. Maybe it’s one of those hard-wired responses that our cave-dwelling ancestors passed on to us to the dangers of faulty, but necessarily quick, assumptions. Maybe it’s due to the conflict we have when confronted with an object in front of us that has been judged to have human-like intentions (aka “theory of mind”) but then behaves like an automaton. Hopefully, Philippe will solve the mystery of the uncanny in animation.

Eisenstein, when not busy crafting what would be the basis of Russian film and the foundations of film theory still used today, worked with leading Russian scientists and psychologists in their interests in learning about language and thought, and in how we understand art and film.  I asked Philippe if he separates being an animator from being a scientist.

“I wear different shirts. (Laughs). But I wouldn’t necessarily call myself a scientist just yet, not because I don’t want to be a scientist. I want to be a mad scientist. Right now, I’m just an animator who is using science.

So, I don’t necessarily call myself a scientist just yet, but I do think the creative process and the scientific method of thinking are very similar. Scientists will find a problem or a question that needs investigating, so they might perform an experiment or use a certain methodology to explore it, while animators working on a film or a shot or scene are kind of confronted with similar problems. The animators will test and try different ways, and it could be in the planning or storyboard or an animatic (to test things out) so I think the methods (of simulation) are very similar. So, in regard to me wearing different hats, I think it’s really the same hat, but with different ear flaps. And I have scientist friends and I feel we have a lot in common. Also, I know that some of the greatest discoveries were done by very creative scientists, and some of the best films were made by very [logical] artists. I think they’re both sides of the same process, but applied to different things using different languages.

So, staging, another of the 12 principles, basically says one thing at a time. If one character is talking don’t make the other characters move. It’s kind of like everyone freezes because the main character is talking now, and you don’t want to create other movements that distract from it. The principle recognizes that we seem to be very easily distracted by other movements, and we’re not good at paying attention to more than one thing at a time. And that’s what I’m reading generally from the scientific literature.  It’s confirming what’s in the 12 principles of animation, it’s just expressed in a different way.”

Animators are often scientists by other means.

Here are links to some of Philippe animations:

The Song-Catcher / Chasse-papillon
https://www.nfb.ca/film/song_catcher/

The Well
https://www.tfo.org/fr/univers/cinetfo/100307609/le-puits

Red & Blue
https://youtu.be/8UK4rNTA_3w

And these were some of the books referenced in this blog:

Sergei Eisenstein.  On Disney (trans Alan Upchurch),  (Edited by Jay Leyda)., 2017 (originally 1986), Seagull Books, London.

E. G. Lutz, Animated Cartoons, Applewood Books, Bedford MA (originally 1920, Scribner’s, NY).

Russell Merrit & J.B. Kaufman, Walt in Wonderland: The Silent Films of Walt Disney, 1994. The Johns Hopkins University Press, Baltimore, MD.