A Science Of Design?


“Design” means many different things. Some think of interior design, or fashion— creating things that are pleasing to the eye. Engineers think of design more pragmatically—designing new circuits, materials, or chemicals. At the Segal Design Institute, we combine these two approaches to focus on the design of systems, goods and services that involve people and technology, and how technological concerns interact with societal and individual ones.

In many ways, we consider the relationship of design to the university as akin to that of engineering to science. If engineering is the application of scientific knowledge and methods, then design is the application of knowledge from across the university: science and engineering, the arts and humanities, the cognitive, social, and behavioral sciences. We call our approach human-centered design.

There are a number of important principles that guide human-centered design derived from psychology, sociology, anthropology, and related fields. One important principle is that of the “conceptual model,” or a person’s understanding of how something works. Whether correct or flawed, conceptual models dictate how people use a system. A good design, then, is one that encourages the development of a good conceptual model by making an object’s operation visible or otherwise clear. This allows users to predict the result of their actions and figure out what to do when problems arise.

Want an example of an effective conceptual model? Just look at your computer screen. You have files and folders, even a trash can. You can put files into folders, and then move one folder into other folders. You can move things to the trash. All this is visible on the screen in a nice, pictorial format.

But this is all a myth. Inside the computer there is no such thing as a file or folder, no such thing as a trash can. Instead, information is scattered throughout the computer’s memory systems, not even in coherent locations. The information that we think of as belonging to a single file is usually broken up into multiple small segments, with elaborate internal mechanisms that point the end of each segment to the beginning of the next. In many cases, what we consider “deleting” does not actually remove the material—the system simply sets those pointers so these pieces are skipped over.

However, the average person is unaware of the elaborate data structures inside the machine. Instead, there is a simple conceptual model that shows files in an orderly structure—a nice model, but highly simplified. Thus, if you type some nasty text, and later think better of it and delete it, the expert who has a more complex and accurate conceptual model can find it.

Another excellent conceptual model is displayed on the dashboard of the Toyota Prius. The Prius has a simple diagram showing the gasoline engine, electrical engine, battery, generator, and power train (wheel), with moving lines showing just what the source and destination is of the power sources at any moment. Like the computer example, this is also a tremendous simplification, but it provides the driver with a valuable conceptual model of how the system operates (see Figure 1).

Now consider an object that offers no model of how it works—your thermostat. Some people, when cold, crank it up to some extremely high setting, hoping the temperature will reach the comfort zone more quickly. Unfortunately, it won’t. The heat will just remain on, expelling the same amount and temperature of air, until the programmed temperature is reached. But it’s not the person’s fault—the thermostat makes no attempt to provide a conceptual model to people. So, they must develop their own, often quite inappropriately.

“Signifiers” are another principle of design that send important signals to people. A signifier is an indicator, a signal in the physical or social world that can be interpreted meaningfully. It “signifies” critical information, even if the signifier itself is an accidental byproduct of the world. Social signifiers are those that are the result of the behavior of others. Let me illustrate.

Suppose you are rushing to catch a train. You don’t recall the exact time the train was scheduled to depart, but you know it’s soon. You run across the city, up the stairs in the train station and on to the platform. There is no train—did you miss it, or has it simply not arrived yet? How can you tell? The state of the platform serves as an signifier (see Figure 2).

Social signifiers do not guarantee the result, but they are strongly suggestive. For example, the crowd of waiting people on the left can be considered evidence that the train has not yet arrived. But, in a busy station, like one at the center of a large city, there are many trains arriving at frequent intervals, so the presence of a crowded platform provides no information about a particular train. The same problem applies to the empty platform on the right, which suggests the train has already left. But what if no one else wanted to catch this particular train?

Doors present wonderful examples of good and bad signifiers in terms of design. For years I’ve been tormented by doors, light switches, and water faucets. Look at the doors in Figure 3. Those are more than mere doors—they are a test. These particular doors happen to guard the entrance to the Dean of Engineering at Northwestern (although I bet he changes them as soon as this gets posted).

Which door would you choose? The door on the left has a flat plate that signifies “push me!” You can’t grab, twist, pull or lift. It looks like a door, it looks like it should be pushed—but no, it’s locked. What about the one on the right? It has a nice, grab-able handle. Twist to rotate the handle and pull. Nope—it’s a push door.

When I explained this to the Dean’s administrative assistant (whose hand is shown in Figure 4), her face lit up and she smiled: “Oh, that’s why everyone has trouble with those doors.” Inappropriate signifiers, frustrating interactions.

Of course, there is a lot more to design, many more principles. For example, as devices become more intelligent, the interaction between technology and its user will become increasingly critical. In this case, some of the initiative comes from people, but some from the devices. Consider automobiles—as technologies progress, the car takes more and more control of the situation when it perceives danger ahead. Adaptive cruise-control that brakes if your car is too close to the one ahead, lane-keeping, and automatic braking for collision avoidance are just a few examples. Some cars can even parallel park automatically. The design of the interaction between car and driver and one car with all the other cars and their environment is a complex topic. Why, one could write a book about it (which I did: The Design of Future Things).

Unhelpful thermostats, inexplicable doors, and cars that park for you are just snapshots of design at work. Design is a challenging, difficult, and little-understood field, mixing aesthetics and function, individual and social behavior, and technological advance. Its principles don’t just guide the creation of objects we use everyday, but the systems and technologies used to address world-wide issues, like energy use and production. It is this broad range that makes our work at Segal so exciting.



Add new comment

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.