acm - an acm publication
Articles

Emotion and affect

Ubiquity, Volume 2002 Issue May, May 1- May 31, 2002 | BY Don Norman 

|

Full citation in the ACM Digital Library


Don Norman on the value of beauty, fun and pleasure in design.

Don Norman has a special interest in usability and human-centered design. He is co-founder of the Neilsen Norman Consulting Group, professor of Computer Science at Northwestern University, and a trustee of the Chicago Institute of Design. He is a former head of Apple's Advanced Technology Lab and the author of several books.


UBIQUITY: Tell us about the book you're working on now, "The Future of Everyday Things."

DON NORMAN: It's about the impact that the modern technologies of computation and telecommunications have on our lives. We have always strapped a wristwatch to our bodies. Now, we strap the cell phone. Someday, we may embed it. We're always in touch with people.

UBIQUITY: Is that a positive or a negative?

NORMAN: It's a virtue to be always in touch with friends and colleagues, no matter where they are. We move and we still keep in touch. But whether or not we're continually in touch, there's no focus. There's no depth. There's no uninterrupted time.

UBIQUITY: How does it balance out?

NORMAN: I think it will balance out by having different means of interacting. We will learn to negotiate the phone calls, chats, e-mails and two-way pagers. Just as when we pass a person's office, if we see they're busy we don't interrupt. If it's important that we speak to the person, we interrupt only to say, "It's really important, I have to see you," and then negotiate a time and move on. We should do the same with our messages and phone calls.

UBIQUITY: What are some of the other themes of your book?

NORMAN: Another major theme is emotion and affect. An emotion has three different aspects to it. One is the scientific understanding of affect and emotion. The second aspect is the display of emotion. Facial expression in people is a rich, multidimensional display of internal feelings. It's a very important part of social interaction. The third aspect is the emotional impact that devices have on us. Affect actually changes how the brain processes information: If you make something more pleasant, it's easier to use. (My first article on this subject is going to be published in the July/August 2002 issue of ACM's Interactions magazine.)

The usability community has not paid enough attention to beauty, to fun, or to pleasure. I'd like to change that. The theme of affect and emotion is growing so much within me that I' considering changing the title and focus of the book to "Emotion and Design."

UBIQUITY: Incidentally, your own Web site, http://www.jnd.org/, is very nice, especially the trailing balls that follow cursor movement on the "Gratuitous Graphics" page. Do many Web sites achieve the right balance of fun and seriousness?

NORMAN: Not enough. For a lot of Web sites, it's all or none. It's rare to find a Web site that's really fun.

UBIQUITY: Does your Web site get criticized?

NORMAN: When I first brought this one out, I got a huge amount of criticism because of little errors that I had made. My response was to thank everybody and immediately correct the errors. This current Website was helped by the critiques that I have received over my past Websites.

UBIQUITY: Were any of the critiques especially interesting to you? In other words, beyond things like typos or wrong numbers?

NORMAN: Yes. Here's a trivial example. I recommend a lot of books in my Web site. I used to cite the books in traditional academic citation form, which is the author's last name first, followed by first name, parenthesis, year, city, publisher, title, and so on. Someone suggested why didn't I just give the title and the author's name, which is what everyday people look for. That made me stop and think that if you actually look at the way academics list their references, it's for the convenience of the librarian or some such person. I've never understood why we include the city of publication, for example. The ordering of information is clearly artificial. It's a small critique but one I thought telling. And so I followed it. I don't use traditional scientific citation; instead I list the name of the person and the name of the book. That's what people care about.

UBIQUITY: Is usability more science than art, or more art than science?

NORMAN: Much of all practices are art. Even science is an art. A laboratory experiment is an art. That's why you learn it at a mentorship or as an apprentice. That's what graduate school is about.

UBIQUITY: Going back to the theme of emotion, what made you realize the importance of the emotional impact of devices?

NORMAN: I started wondering why it was that I prefer pleasant Web sites, pleasant industrial design, and nice looking products to ones that are bad looking. When color computer displays first came out. I knew that color didn't make a difference for stuff like word processing and e-mail. There were many scientific studies showing that color didn't matter for readability. You could read just as fast without it. I decided that I needed to understand this because more business people insisted on color. So I got myself a color display and took it home for a week. When the week was over, I had two findings. The first finding was that I was right, there was absolutely no advantage to color. The second was that I was not going to give it up.

UBIQUITY: Talk about your work with robots.

NORMAN: I became interested in the design of robots, which will be a theme of the book, as well. The question is, how do you make a home robot that is autonomous, that lives by itself, that won't get stuck in corners, and doesn't have to be reminded so that it doesn't run out of power? I decided that what it needed was emotions, or affect.

UBIQUITY: What kinds of emotions?

NORMAN: It had to get frustrated. Being frustrated gets it out of deadlocks. If it's stuck somewhere trapped in a corner, it has intelligent algorithms trying to get it out. But if they fail, it says, "the hell with it," and goes off and does something else. It should be afraid of heights so that it doesn't fall down the stairs. It should get fatigued so that it won't wear out the battery. As its battery level gets lower, it should travel more slowly and not do some tasks. It should always make sure its close enough to the recharging station so that it can get back.

UBIQUITY: But are those emotions not essentially algorithms?

NORMAN: So what is an emotion? Let's instead use the word affect. I started trying to understand what affect was. Humans have two processing systems. One of them is cognitive, which is interpreting and understanding the world. The other is affect, which is evaluating the world, making quick value judgments, mainly positive or negative. The two processing systems work differently. The affect system works by chemical infusion. It changes the operating parameters of cognition. The cognitive system works more slowly, mostly by electrical signals. It interprets.

UBIQUITY: Without getting too clinical, explain how these systems process information and how it relates to your theme of emotions.

NORMAN: There are three levels of processing. The lowest level is reflexive. That's what makes your hand jerk away when you burn it. The third level is cognition, which is slow and analytical, reflective. The middle level is always evaluating where you are versus your expectations, which is how you get to hope, anticipation and dread . . .. So when I'm anxious and worried, I flood the brain with stuff that makes me focus on the task, which makes me think harder about whatever path I'm thinking about. If I try to open a door that won't open, I push it again harder. It's called tunnel vision. If I give you an unexpected gift, then I flood the frontal lobes with dopamine.

There have been a number of studies showing that people solve difficult problems better when they're happy than when they're neutral or when they're angry, which leads me to the original point. In other words, the brain changes when we are happy, making pleasant objects easier to use. We are global, breadth-first thinkers when happy, local, depth-first thinkers when stressed. Affect is truly an important factor in how we live in the world. There are a lot of exciting new findings. I want to bring them to the attention of designers -- and engineers who build large, complex, autonomous systems.

UBIQUITY: And yet, when you started thinking about this, when you evaluated the black-and-white vs. color monitors, you called the non-pleasant objects equally usable.

NORMAN: This is an error that those in the field of usability have made, including me. We looked solely at the cognitive aspect of usable. If you're trying to distinguish say, highlighting from non-highlighting, is there any difference in the ability of a person to distinguish a gray background from a yellow background? If there is any difference, it's minuscule. We looked at information processing. We didn't look at affect. We didn't look at judgmental things. We also didn't think of looking at overall behavior.

UBIQUITY: Thinking of color in terms of the early television sets, I recall being appalled because the color was often bad. It makes me wonder whether quality plays a role. In other words, would bad color be more usable or pleasant than black and white?

NORMAN: A quick answer is no. Color TV is another example, by the way, that when it first came out, there were many issues. What's the matter with black and white? The movie is the same movie, isn't it? But today, would you go back to black-and-white TV? No. The story is unchanged. The actors are the same. But color adds something that we find difficult to measure.

UBIQUITY: Wouldn't the movie director Woody Allen, who films many movies in black and white, disagree with that statement?

NORMAN: Actually, Woody Allen would agree. That is to say, making something in color changes your perception. If you're an artist, you manipulate people's perceptions. You may wish to use black and white because you want people to focus and think in a certain way. Many photographers insist on using black and white. They rightly feel that it gives a very different picture. But the vast majority of us in most situations prefer color. Now, I think there are different levels of bad color. There's color the experts says is bad and there's color where the faces are all green. That's the worst because it jars you. I'd like to say that bad color can't be good.

But for some things, inferior quality seems to be acceptable. For instance, we now have high definition TV, which is quite superior to standard definition, but most people don't give a damn because they're used to their TV and don't see a need to switch. That's a standing problem in the industry. Manufacturers always ask why should they spend money to convert to better sound and better picture when the evidence is that the public doesn't care that much?

UBIQUITY: Does that attitude eventually change?

NORMAN: I think it does. After you've experienced it awhile, you don't want to go back. But it takes a while. Color TV was the same way. It took a long time to be adopted. It was a combination of people not seeing why they needed it and it being quite expensive.

UBIQUITY: Going back to your robots and the question of getting them to have affects. How is that accomplished? How do you get them to feel frustrated?

NORMAN: There are two levels. The bottom level has things like fear of heights (cliff detection) or injury. Those are built-in detectors. They send a strong signal to avoid those locations. The second level is comparison of expectations, and that's where frustration comes in. You build up repeated attempts to do a task with little or no progress towards completion -- a frustration index, if you like. You need a way of measuring progress towards task to compare with how much effort you're putting in. You also need sensors of things such as internal temperature, whether your belts are slipping and how much energy you're putting into the motors versus how fast you're progressing, which is a measure of slippage. All those factors would help build up the frustration. Then you need an architecture that uses this their natural way. Fortunately, the new behavior-based robotic architectures allow that. It fits directly into the way they are already orchestrated.

UBIQUITY: How would you actually implement this?

NORMAN: Several architectures lend themselves nicely to this, including fuzzy logic, neural networks, and hidden Markov marker modes. You simply add activation, or change the weights, or the equivalent. Changing weights and thresholds is probably what happens within the human nervous system.

UBIQUITY: Are there any robotic applications that are particularly interesting to you, right now?

NORMAN: Home robots. I think they are going to be big in entertainment. Sony has focused its robots, the Aibo, on home entertainment. Entertainment is a clever thing to do because that way, when your robot screws up, it's part of the fun. But I think they will be caretakers and educators, playthings and companions.

UBIQUITY: Let's go back to the themes of your book. What's another theme?

NORMAN: I believe entertainment is going to change dramatically. Computer games are a good example. Computer gaming is now a huge industry, bigger in sales than Hollywood box office receipts. They make heavy use of simulation and of artificial beings. I think that's going to be very powerful and important. It's going to affect the way we even hear the news. I mean, half the news sets we see today are fake. The sets themselves are computer-generated and superimposed over the newscasters. One of these days, the newscasters will be fake. For all I know, they already are. Could we tell the difference? And then there's the new genre of literature, adventure games and God games.

UBIQUITY: What is your take on God games and adventures games?

NORMAN: Adventure is like the Dungeons and Dragons where you have quests and you search out and overcome many obstacles. A God game is where you create a universe. Any of the Sims games such as Sims City, Sims Earth, et cetera, are God games because you create the environment or the family. Then, if it isn't evolving to your liking, you change it. But you don't have direct control. You can only change high-level parameters.

Then there are games that start to take on their own life. The games will play even when you're not there. Sometimes they'll call you on the phone or send you faxes or e-mail asking for help or advice. I think this is a new literature form. It's what interactive literature is about. At one point, it was thought interactive literature was this, oh, we'll tell you a story and then every so often, we'll ask which alternative do you want at this point? I think that's a failed notion. Video games got the interaction right: the way a video game progresses depends upon what you do. In the God games, the high-level parameters that you set up determine what happens. So far, none of these are literature. But we're also in the early infancy.

UBIQUITY: In thinking about games as literature, think for a second about sports as ballet. Basketball or some other sport can be very balletic. It can be all kinds of graceful emotion. But it's not normally thought of as an art. It's still somehow a sport or a game.

NORMAN: What do you think the Harlem Globetrotters are? Usually in sports, there's a goal, which is to win. The goal is not to entertain. But the Harlem Globetrotters basketball team is interesting. Although they like to win, their real goal is to entertain the audience. And so, that, I think, moves it closer to a theater. Another question is what's the difference between ice skating and ballet, and circus performers and ballet? That line in recent years is diminishing. Now what does that tell us about video games relevant to literature? I think the line will go away. The early days of movies were akin to video games today. They told simple-minded stories. People did not treat them as art.

UBIQUITY: When video games get even better 10 or so years from now, will that change not just the games and the art but also the way people think? Will they change the character of management and society and organizations and so forth?

NORMAN: I doubt it. But let me move to a different topic, which flows naturally from this question: education. What's the difference between a simulation game and education? Answer: How it's used. Many of the games have powerful simulations of nonsensical societies or something. But if they were real, you would learn from them. The most dramatic example is a flight simulator or a driving simulator. They are games that people play and enjoy. But the flight simulators are accurate enough that they are actually used for flight training. Instructors use the games as initial flight training and then switch over to more powerful, more realistic simulators. But they're still simulators.

I think that education will be enhanced dramatically. The very same technologies that are used in games will be used in education. There will be simulators and enhanced tutors, some with AI components. A partial answer to your question is that we can pull out of games things of great value to education.

COMMENTS

POST A COMMENT
Leave this field empty