Pete Docter & Dacher Keltner
Creating feelings and lovable beings
Why do we feel so much during Pixar films? How can a wordless robot or one-eyed monster win our hearts and gather our enthusiastic support? What is it at work in these creatures and characters that makes them so easy to love, easy to cheer for, and easy to feel connected to, what makes them just so darn…relatable?
Emotion is at the heart of Empathy Lab’s work. We’re fascinated by the study of connection, the things and beings that stir and evoke our deepest human feelings. We are massive Pixar fans, taking total child-like joy in their many worlds of wonder. So naturally we jumped at the chance to talk to the folks who created the magic of Inside Out, to learn about their journey breathing life into the emotions as characters…and also to hear a little about the secret sauce of creating such beloved beings across all their films.
We were bubbling with excitement the day Pete Docter, Pixar’s Academy Award-winning director and writer, and Dacher Keltner, professor of psychology and director of UC Berkeley’s Interaction Lab and Center for Greater Good came to hang and talk with us about expression, emotion, and their work together making of Inside Out.
Dacher, you’re a decorated professor of psychology, tell us how Inside Out and Pixar found their way to you.
Dacher: I got this phone call from Pete, who I’d known from different contexts, and he said, “Hey, I’m making a movie about emotion,” I was like, it’s about time, and he’s like, “and it’s about how emotions work in the mind” and I was like, This is amazing!” And so he brought me in and it was an incredible privilege to be part of Pete’s and Ronnie’s producing of Inside Out and just to be a scientific voice in their journey. There’s this really interesting process by which science, which is a foundation to the film, becomes converted to design. So it was really interesting to be part of this conversation and to feed science into it.
One thing we’ve loved about the universal response to Inside Out is that it’s given permission and understanding to conversations about our messy, lovely human feelings. How do you feel Inside Out has influenced our culture around emotion?
Dacher: Inside Out was, I think, the most powerful revision of how we think about emotion for 2000 years. From Plato to Kant, we have thought about emotions as disruptive dysfunctional irrational processes, and we should suppress them. And Inside Out comes along and says they are the very foundation of story and relationships and purpose in life.
Inside Out comes along and says [emotions] are the very foundation of story and relationships and purpose in life.
Some of the core scientific insights that are undergird Inside Out are really relevant to how we design things out in the world based on what we know about the human mind.
Can you share with us some of those core scientific insights?
Dacher: Principle one is you really are your emotions. Descartes said, “I think therefore I am”. That is wrong. It really is “I feel, therefore I am”. Emotions are narratives, they’re part of your identity. Half of who you are as an individual from the first moment of life and before in the womb is based on that emotional temperament.
You really are your emotions
I think one of the most striking themes in the film is that your past is seen through the lens of your present emotion. It becomes this really complex lens by which we retrieve information filtered through present emotion. And this is not only true of how we construct the past, it’s true, of how we construct every judgment. Do I want to buy this home? Is this somebody I should fall in love with? Those basic judgments come through emotion.
Emotions are not irrational disruptive forces, they are the grammar of social living.
The second principle is that emotions are not irrational disruptive forces, they are the grammar of social living. For example, when a parent hugs a child or adolescent flirting, it’s an interesting unfolding of little brief emotional expressions and tones of voice and facial expressions. You see that throughout Inside Out, the science and the Pixar genius. The portrayal of emotional expression is so precise and accurate and artistic.
In some sense, Inside Out was this meditation on accepting and embracing negative emotions.
The third principle is the power of acceptance. Inside Out has the most unlikely hero in the history of filmmaking — Sadness — and this becomes an opportunity for our culture, which loves to suppress sadness, to medicate sadness, to not talk about it. And suddenly it opened up this conversation that sadness is okay, you move through it. It transforms human relationships. In some sense, Inside Out was this meditation on accepting and embracing negative emotions.
Beautiful. What happens when we embrace our emotions?
Dacher: At the Greater Good Science Center we do a lot of work on the benefits of not reacting against negative states, like stress or sadness or anger. One study shows the more acceptance you practice in handling stress, the less stress that you feel over time.
We have to recognize the wisdom of the emotions.
Having these feelings at the right times about the right things, toward the right people in the right way is proper virtue. The good life rests in moderation, being open to the emotions, letting them take you on the brief journeys that they take you on. The richer your emotional life, and the more you embrace your emotions, the better your health profile. We have to recognize the wisdom of the emotions.
Pete, the relationship between Sadness and Joy seems to be where the “juice” of the film really is. We’re curious, as you’re feeling your way through a story, how do you intuit where those moments are? How do you know that’s “the thing”?
Pete: Well we didn’t at first. Like I said, we just keep making mistakes. I think, you know, you try stuff out. Having some way to mock things up and road-test them is pretty key. You need to have cultural permission to make blunders and errors, you know, even if they look like beginner kind of mistakes. Without that sort of permission, you’re never gonna get it right the first time.
You need to have cultural permission to make blunders and errors. Without that sort of permission, you’re never gonna get it right the first time.
We’re curious, how did the team’s own emotions play a part in the team’s creative process. How much is gut instinct and how much is the art and science of story?
Pete: I’d say it’s 99–100% gut instinct at the beginning, and then you put the thing up and you’re either surprised or justified in what you’ve done. Andrew Stanton, who directed Nemo and Wall-e has something he’s said from the beginning, “just make me care!” That’s the job of the storytelling, to make you care, which is basically another way of saying, “make me feel an emotion of some kind”. It doesn’t have to be happiness, it could be sadness, could be anger, could be anything, but that turns out to be surprisingly difficult sometimes.
That’s the job of the storytelling, to make you care…[to make you] feel an emotion
We’ve heard that the team explored logic as an emotion in Inside Out, but it felt too robotic. Can you tell us about that?
Pete: Yeah, we had a character we called Frank, who was from the head office and he was always rolling his eyes at the emotions, like they’re a bunch of amateurs that get us into trouble, and he was trying to shut them down. He was kind of the antagonist at first, but then we realized we’d basically created a robot, an automaton who’s just responding to the whims of Frank or the emotions, and we wanted to create room for Riley as a character. As storytellers we wanted Joy’s relationship to Riley to be parental and loving, and she couldn’t care deeply about a robot. It has to be a being that has its own authority, and in the end we can’t choose our emotions, they’re unbidden. They come on, but we can choose what we do with them.
On that note, Wall-E literally is a robot, yet we fall in love with him immediately. Help us understand what’s going on in our weird humanness with these dynamics?
Pete: Yeah, maybe that was a poor choice of words “caring for a robot”. There are studies they’ve done with bomb detection robots that were walking out into minefields and having their legs blown off and the general would be like, “We can’t. We have to stop, this is inhumane,” and they would stop these robots from going further because they empathize so strongly with them.
Purposely giving yourself restrictions sometimes, I think is a good thing.
In the case of Wall-E, one of the keys was the eyes. From the early on we were thinking of it as having no dialogue. None of the short films have dialogue and yet audiences totally get everything. We believed in the power of the acting and animation and Wall-E turned out to be a really giving character. I know the animators had a great time with it.
I always feel like when you can, you should give people creative restrictions, for example Mike Wazoski who has one eye. So much of our expressions are about the asymmetry in our brows and our face, so we had to figure out a way of squishing two eyes into one and being able to get skepticism and all that out of this one eye. Wall-E’s the same thing. You don’t have a mouth and that’s a big expressive tool that we all use, so you just have to rely on some of the other stuff, and yet as an animator, that’s what makes those characters specific and unique. So, purposely giving yourself restrictions, sometimes I think is a good thing.
Is there any sort of pattern in the way Pixar creates these characters? Is there a magic recipe behind the feelings that rise up as we join the journey of each film?
Pete: Well if I knew what it was, they wouldn’t take five years to make! John Lasseter has been really big since the beginning on being in the shoes of your main character, so you understand things and learn information at the same pace as your main character. The whole thing with storytelling is that you’re taking a character who’s flawed, and then you’re watching them change and grow and become sort of healed of that.
You have to make them palatable and relatable.
But if we as the audience recognize them as flawed we won’t like them. So you have to make them palatable and relatable. And it doesn’t mean the character has to be perfect and nice, you know, you try to find some edge to the character. And then that’s what the character has to learn.
How do you find the path to take risks and push the creative process by integrating new technology? How does that work for the teams?
Well, we’re not afraid to. Also, we are afraid. I don’t really think of them as being risks. A friend of mine who’s a filmmaker has this great sentiment that “there’s nothing riskier than playing it safe, and there’s nothing safer than taking a risk”. In other words, if you do something new, even if it doesn’t 100% work, the audience will be like, “Well I’ve never seen that before!” Whereas if you try to do things we know like, they’re gonna be like “Seen it.”
It’s especially difficult with the sequels, because people come in with a certain expectation of what they liked last time and want again. So you have to figure out which elements do we have in this story that we can reconfigure to make it seem new and familiar at the same time?
The universal love of Pixar characters is inspiring, because they touch everyone. As we turn our attention to the space of digital assistants, we’re curious what sort of advice you’d have for character creation?
Pete: Half the time I talk to the thing it doesn’t understand me and I’m super frustrated. If it can’t understand what I’m saying, then it doesn’t matter what the character is. The functionality has to get there first, but then after that it’s really interesting.
Dacher: I think this is where you really need to draw upon the genius of Pixar and remember that humans have massively powerful imaginations. I would make them really playful and weird and fun and as opposed to this normalized mean of human encounter which ends up being nothing.
Pete: We’ve found over and over in storytelling that you need to expose an element of vulnerability to the audience. That’s a huge thing I think, even in real life.
We’ve found over and over in storytelling that you need to expose an element of vulnerability to the audience.
Let’s say you go to a restaurant and you have lousy service but if the waiter comes over and says “I’m sorry, this my first day” you’re like “Oh! Well, of course go.” You’re much more forgiving. An element like that connects you and allows you to forget some stuff.
Dacher: I was going to say the same thing. It’s the foibles and the mistakes and the flaws of humans that make you love them, make you forgive them, make you patient, and make you trust them. Part of why we love animated characters is that they’re quirky and that’s why we love them. So I would be thinking not about perfection, but about mistakes.
[Think] not about perfection, but about mistakes.
We didn’t want to steal all the precious time with Pete and Dacher, so handed the mic over to the curiosities of Googlers in the audience next:
Googler: After spending five years thinking about emotions and how they interact with one another, how did it change you or the way that you live?
Pete: Yeah, definitely. For one, I’m just way more aware. Emotions trick us into acting. The other day I was finding myself really angry…and I was able to open the hood and look at the engine a little bit.
Dacher: Aristotle had this really interesting idea about having insight into the emotions of yourself and other people. Emotions are narratives. They’re the story of your life and you see little pieces of them as that story unfolds.
Googler: If emotions are a huge part of our mind process and intelligence, is it possible to make artificial intelligence without emulating emotions?
Dacher: I think that you could design non-living systems that would function at 80% in terms of detecting stimuli in the environment and extracting themes and acting upon those things, but you would miss the subjective layer to emotion. Metaphors and connections to past memories and all the distributive processes that happen in the brain that are really complicated you would miss, but you can get 80% of this stuff.
Googler: I want to follow up on the idea of making assistants more quirky and emotional. How do you dial in that sort of personality without it seeming fake?
Dacher: Pragmatically you’re going to need to take into consideration the temperament and present state of the user. For example, a highly fearful person is not going to have fun with a quirky character, but somebody who’s wild and open and creative would really like an assistant that has quirky dimensions. That being said, humans have always liked art and quirkiness no matter what form it’s represented in.
Pete: If I’m super stressed and I’m on the freeway, I don’t want somebody to be jokey and funny, but there might be other times when I do.
Googler: Can you talk a bit about the idea of emotion as the birth of feeling something that happens to you in the short-term versus emotion as a character trait?
Pete: Emotions are short. They’re reactions. And mood is like when you wake up kind of grouchy and it’s not necessarily a reaction to anything. It’s just you’re grouchy and that’s it. We wanted to make sure that our characters were not moods, they were emotions.
Dacher: We have temperaments that persist through the day and you can drop in and measure that scientifically. And you’ll feel that somebody is kind, right? You’ll sense that in them. And that’s joy, temperamentally. These fluctuations are what emotions do.
