Potential problems of emotional design

Summary:
Emotional design is a generally good focus. However, it presents the problem of marginalizing some users who cannot appreciate certain displays of emotions and it also does not address the fact that objects and people take place in a distributed-system.

I went to see Don Norman talk about his book, Emotional Design (a good and thoughtful read, as all Norman books are, in my opinion). Although I agree that stressing emotion in the design of consumer products is worthwhile, I responded to the speech by noting two potential problems with designing products that are emotionally evocative:

The first problem is practical: Some users may have bad reactions to emotions in certain products. They will be unable to use products that are made for the majority "emotional" market.

Some users may find the emotions of the product to be horrific and the lack of product usability to be dumbfounding. For example, a mobile phone that had pleasant tones and "calming" graphics (of swimming fish) was so bizarre for one elderly user who I observed that she had a very negative affective reaction to it. It was so negative that she ended up crying and talking about how she is not able to understand everyday objects anymore. It was a very sad scene, and I blame it partially on the emotional design of the phone. On the flip side, the middle aged woman who owns the phone and thinks it is has cute emotions. I like it more than most phones too, and I anticipate most younger people do. Nevertheless, since emotional design has been stressed, unemotional phones have become so scarce that some (older) users are unable to join the market. Ironically, this market is much more able to cope with the large "car phone"-style phones that were available in the early 1990s than it is modern designs...too bad those old, stupid phones are no longer produced for them. Don't get me wrong--emotional design may reach the largest market, but a substantial non-emotional design market should not be ignored.

The second problem is theoretical: Emotionally-evocative designs may be inappropriate for objects used in distributed tasks.

A broad underpinning of the emotional design thesis is that people find some objects to evoke positive emotions and other objects to evoke negative emotions. A question that needs to be asked is whether these emotions are always evoked or if they only occur during the early stages of being introduced to an object. In Emotional Design, Norman eludes to how some objects are emotionally satisfying upon initial viewing but that this wanes upon subsequent encounters. Perhaps this is a hallmark of bad emotional design for some types of objects because it is a sign that "cuteness" is wearing away. Then again, perhaps it is a hallmark of good emotional design for other objects. I believe that the process of labeling an object as "emotionally good" needs to be broken down into use/time-specific components.

Norman made several elegant contributions to the field of distributed systems, however the proper way to reconcile this field with his new ideas about emotional design is unclear. "Distribution systems" perform tasks using a combination of the internal mind and body of the task-doer, and the external objects in the environment of the task-doer. Therefore, the work is distributed across the person and environment. Why does this present a challenge to emotional design? Let's use an example. When I get onto my bicycle, I notice that I am emotionally drawn to its body style initially. Once I am riding the bike, however, I cease to appreciate this fact. Instead, the bicycle's physical constraints and affordances are integrated into my body--my feet become wheels and my hands become a direction-bearing and braking device. I may emotionally appreciate the bicycle in terms of the way it rides and handles, but I do not appreciate the bicycle as a stand-alone system at this point--I only appreciate the way it works with me on it. People integrate well into their cars--after a long drive, people don't recognize that they are operating automobile controls. Instead they are integrated into the driving system--the car becomes almost an extension of the body. It is as if their human input and output systems no longer exist within the body, but instead because they are routed to the automobile inputs and outputs.

At this integrated level of interaction, object emotion may be a problem. This is because tasks like driving a car are controlled better by a distributed human-object system than by a person who interacts with a separate entity of an object. So being constantly reminded that you are a separate entity from the car may be detrimental to driving. After all, the point of drivers education courses is to get inexperienced drivers to the point of retraining their inputs and outputs to being the car's inputs and outputs. If we are reminded that the object in the distributed system has emotions--especially when these emotions are not central to the distributed sense of being--then this may make the distributed human-object system less effective. Because of this, I believe it may make sense to limit the integration of emotional design to non-distributed systems (i.e. systems that people interact with but don't distribute their body or mind into--although this can sometimes be a gray line). At very least, this should be taken as a warning against designing emotion into the interfaces where it could be dangerous (like automobile interfaces).

These two comments are minor in comparison to the great benefits that I believe are produced by stressing emotional design. Nevertheless, I thought it would be productive to add my two cents to the dialogue.