The Design of Everyday Things

by Don Norman

  But wait: I’m not finished. Is the time line relative to the person or relative to the environment? In some Australian Aborigine societies, time moves relative to the environment based on the direction in which the sun rises and sets. Give people from this community a set of photographs structured in time (for example, photographs of a person at different ages or a child eating some food) and ask them to order the photographs in time. People from technological cultures would order the pictures from left to right, most recent photo to the right or left, depending upon how their printed language was written. But people from these Australian communities would order them east to west, most recent to the west. If the person were facing south, the photo would be ordered left to right. If the person were facing north, the photos would be ordered right to left. If the person were facing west, the photos would be ordered along a vertical line extending from the body outward, outwards being the most recent. And, of course, were the person facing east, the photos would also be on a line extending out from the body, but with the most recent photo closest to the body.

  The choice of metaphor dictates the proper design for interaction. Similar issues show up in other domains. Consider the standard problem of scrolling the text in a computer display. Should the scrolling control move the text or the window? This was a fierce debate in the early years of display terminals, long before the development of modern computer systems. Eventually, there was mutual agreement that the cursor arrow keys—and then, later on, the mouse—would follow the moving window metaphor. Move the window down to see more text at the bottom of the screen. What this meant in practice is that to see more text at the bottom of the screen, move the mouse down, which moves the window down, so that the text moves up: the mouse and the text move in opposite directions. With the moving text metaphor, the mouse and the text move in the same directions: move the mouse up and the text moves up. For over two decades, everyone moved the scrollbars and mouse down in order to make the text move up.

  But then smart displays with touch-operated screens arrived. Now it was only natural to touch the text with the fingers and move it up, down, right, or left directly: the text moved in the same direction as the fingers. The moving text metaphor became prevalent. In fact, it was no longer thought of as a metaphor: it was real. But as people switched back and forth between traditional computer systems that used the moving window metaphor and touch-screen systems that used the moving text model, confusion reigned. As a result, one major manufacturer of both computers and smart screens, Apple, switched everything to the moving text model, but no other company followed Apple’s lead. As I write this, the confusion still exists. How will it end? I predict the demise of the moving window metaphor: touch-screens and control pads will dominate, which will cause the moving text model to take over. All systems will move the hands or controls in the same direction as they wish the screen images to move. Predicting technology is relatively easy compared to predictions of human behavior, or in this case, the adoption of societal conventions. Will this prediction be true? You will be able to judge for yourself.

  Similar issues occurred in aviation with the pilot’s attitude indicator, the display that indicates the airplane’s orientation (roll or bank and pitch). The instrument shows a horizontal line to indicate the horizon with a silhouette of an airplane seen from behind. If the wings are level and on a line with the horizon, the airplane is flying in level flight. Suppose the airplane turns to the left, so it banks (tilts) left. What should the display look like? Should it show a left-tilting airplane against a fixed horizon, or a fixed airplane against a right-tilting horizon? The first is correct from the viewpoint of someone watching the airplane from behind, where the horizon is always horizontal: this type of display is called outside-in. The second is correct from the viewpoint of the pilot, where the airplane is always stable and fixed in position, so that when the airplane banks, the horizon tilts: this type of display is called inside-out.

  In all these cases, every point of view is correct. It all depends upon what you consider to be moving. What does all this mean for design? What is natural depends upon point of view, the choice of metaphor, and therefore, the culture. The design difficulties occur when there is a switch in metaphors. Airplane pilots have to undergo training and testing before they are allowed to switch from one set of instruments (those with an outside-in metaphor, for example) to the other (those with the inside-out metaphor). When countries decided to switch which side of the road cars would drive on, the temporary confusion that resulted was dangerous. (Most places that switched moved from left-side driving to right-side, but a few, notably Okinawa, Samoa, and East Timor, switched from right to left.) In all these cases of convention switches, people eventually adjusted. It is possible to break convention and switch metaphors, but expect a period of confusion until people adapt to the new system.

  CHAPTER FOUR

  KNOWING WHAT TO DO: CONSTRAINTS, DISCOVERABILITY, AND FEEDBACK

  How do we determine how to operate something that we have never seen before? We have no choice but to combine knowledge in the world with that in the head. Knowledge in the world includes perceived affordances and signifiers, the mappings between the parts that appear to be controls or places to manipulate and the resulting actions, and the physical constraints that limit what can be done. Knowledge in the head includes conceptual models; cultural, semantic, and logical constraints on behavior; and analogies between the current situation and previous experiences with other situations. Chapter 3 was devoted to a discussion of how we acquire knowledge and use it. There, the major emphasis was upon the knowledge in the head. This chapter focuses upon the knowledge in the world: how designers can provide the critical information that allows people to know what to do, even when experiencing an unfamiliar device or situation.

  Let me illustrate with an example: building a motorcycle from a Lego set (a children’s construction toy). The Lego motorcycle shown in Figure 4.1 has fifteen pieces, some rather specialized. Of those fifteen pieces, only two pairs are alike—two rectangles with the word police on them, and the two hands of the policeman. Other pieces match one another in size and shape but are different colors. So, a number of the pieces are physically interchangeable—that is, the physical constraints are not sufficient to identify where they go—but the appropriate role for every single piece of the motorcycle is still unambiguously determined. How? By combining cultural, semantic, and logical constraints with the physical ones. As a result, it is possible to construct the motorcycle without any instructions or assistance.

  FIGURE 4.1.Lego Motorcycle. The toy Lego motorcycle is shown assembled (A) and in pieces (B). It has fifteen pieces so cleverly constructed that even an adult can put them together. The design exploits constraints to specify just which pieces fit where. Physical constraints limit alternative placements. Cultural and semantic constraints provide the necessary clues for further decisions. For example, cultural constraints dictate the placement of the three lights (red, blue, and yellow) and semantic constraints stop the user from putting the head backward on the body or the pieces labeled “police” upside down.

  In fact, I did the experiment. I asked people to put together the parts; they had never seen the finished structure and were not even told that it was a motorcycle (although it didn’t take them long to figure this out). Nobody had any difficulty.

  The visible affordances of the pieces were important in determining just how they fit together. The cylinders and holes characteristic of Lego suggested the major construction rule. The sizes and shapes of the parts suggested their operation. Physical constraints limited what parts would fit together. Cultural and semantic constraints provided strong restrictions on what would make sense for all but one of the remaining pieces, and with just one piece left and only one place it could possibly go, simple logic dictated the placement. These four classes of constraints—physical, cultural, semantic, and logical—seem to be universal, appearing in a wide variety of situations.

  Co
nstraints are powerful clues, limiting the set of possible actions. The thoughtful use of constraints in design lets people readily determine the proper course of action, even in a novel situation.

  Four Kinds of Constraints: Physical, Cultural, Semantic, and Logical

  PHYSICAL CONSTRAINTS

  Physical limitations constrain possible operations. Thus, a large peg cannot fit into a small hole. With the Lego motorcycle, the windshield would fit in only one place. The value of physical constraints is that they rely upon properties of the physical world for their operation; no special training is necessary. With the proper use of physical constraints, there should be only a limited number of possible actions—or, at least, desired actions can be made obvious, usually by being especially salient.

  Physical constraints are made more effective and useful if they are easy to see and interpret, for then the set of actions is restricted before anything has been done. Otherwise, a physical constraint prevents a wrong action from succeeding only after it has been tried.

  The traditional cylindrical battery, Figure 4.2A, lacks sufficient physical constraints. It can be put into battery compartments in two orientations: one that is correct, the other of which can damage the equipment. The instructions in Figure 4.2B show that polarity is important, yet the inferior signifiers inside the battery compartment makes it very difficult to determine the proper orientation for the batteries.

  Why not design a battery with which it would be impossible to make an error: use physical constraints so that the battery will fit only if properly oriented. Alternatively, design the battery or the electrical contacts so that orientation doesn’t matter.

  Figure 4.3 shows a battery that has been designed so that orientation is irrelevant. Both ends of the battery are identical, with the positive and negative terminals for the battery being its center and middle rings, respectively. The contact for the positive polarity is designed so it contacts only the center ring. Similarly, the contact for negative polarity touches only the middle ring. Although this seems to solve the problem, I have only seen this one example of such a battery: they are not widely available or used.

  FIGURE 4.2.Cylindrical Battery: Where Constraints Are Needed. Figure A shows the traditional cylindrical battery that requires correct orientation in the slot to work properly (and to avoid damaging the equipment). But look at Figure B, which shows where two batteries are to be installed. The instructions from the manual are shown as an overlay to the photograph. They seem simple, but can you see into the dark recess to figure out which end of each battery goes where? Nope. The lettering is black against black: slightly raised shapes in the dark plastic.

  FIGURE 4.3.Making Battery Orientation Irrelevant. This photograph shows a battery whose orientation doesn’t matter; it can be inserted into the equipment in either possible direction. How? Each end of the battery has the same three concentric rings, with the center one on both ends being the “plus” terminal and the middle one being the “minus” terminal.

  Another alternative is to invent battery contacts that allow our existing cylindrical batteries to be inserted in either orientation yet still work properly: Microsoft has invented this kind of contact, which it calls InstaLoad, and is attempting to convince equipment manufacturers to use it.

  A third alternative is to design the shape of the battery so that it can fit in only one way. Most plug-in components do this well, using shapes, notches, and protrusions to constrain insertion to a single orientation. So why can’t our everyday batteries be the same?

  Why does inelegant design persist for so long? This is called the legacy problem, and it will come up several times in this book. Too many devices use the existing standard—that is the legacy. If the symmetrical cylindrical battery were changed, there would also have to be a major change in a huge number of products. The new batteries would not work in older equipment, nor the old batteries in new equipment. Microsoft’s design of contacts would allow us to continue to use the same batteries we are used to, but the products would have to switch to the new contacts. Two years after Microsoft’s introduction of InstaLoad, despite positive press, I could find no products that use them—not even Microsoft products.

  Locks and keys suffer from a similar problem. Although it is usually easy to distinguish the smooth top part of a key from its jagged underside, it is difficult to tell from the lock just which orientation of the key is required, especially in dark environments. Many electrical and electronic plugs and sockets have the same problem. Although they do have physical constraints to prevent improper insertion, it is often extremely difficult to perceive their correct orientation, especially when keyholes and electronic sockets are in difficult-to-reach, dimly lit locations. Some devices, such as USB plugs, are constrained, but the constraint is so subtle that it takes much fussing and fumbling to find the correct orientation. Why aren’t all these devices orientation insensitive?

  It is not difficult to design keys and plugs that work regardless of how they are inserted. Automobile keys that are insensitive to the orientation have long existed, but not all manufacturers use them. Similarly, many electrical connectors are insensitive to orientation, but again, only a few manufacturers use them. Why the resistance? Some of it results from the legacy concerns about the expense of massive change. But much seems to be a classic example of corporate thinking: “This is the way we have always done things. We don’t care about the customer.” It is, of course, true that difficulty in inserting keys, batteries, or plugs is not a big enough issue to affect the decision of whether to purchase something, but still, the lack of attention to customer needs on even simple things is often symptomatic of larger issues that have greater impact.

  Note that a superior solution would be to solve the fundamental need—solving the root need. After all, we don’t really care about keys and locks: what we need is some way of ensuring that only authorized people can get access to whatever is being locked. Instead of redoing the shapes of physical keys, make them irrelevant. Once this is recognized, a whole set of solutions present themselves: combination locks that do not require keys, or keyless locks that can be operated only by authorized people. One method is through possession of an electronic wireless device, such as the identification badges that unlock doors when they are moved close to a sensor, or automobile keys that can stay in the pocket or carrying case. Biometric devices could identify the person through face or voice recognition, fingerprints, or other biometric measures, such as iris patterns. This approach is discussed in Chapter 3, page 91.

  CULTURAL CONSTRAINTS

  Each culture has a set of allowable actions for social situations. Thus, in our own culture we know how to behave in a restaurant— even one we have never been to before. This is how we manage to cope when our host leaves us alone in a strange room, at a strange party, with strange people. And this is why we sometimes feel frustrated, so incapable of action, when we are confronted with a restaurant or group of people from an unfamiliar culture, where our normally accepted behavior is clearly inappropriate and frowned upon. Cultural issues are at the root of many of the problems we have with new machines: there are as yet no universally accepted conventions or customs for dealing with them.

  Those of us who study these things believe that guidelines for cultural behavior are represented in the mind by schemas, knowledge structures that contain the general rules and information necessary for interpreting situations and for guiding behavior. In some stereotypical situations (for example, in a restaurant), the schemas may be very specialized. Cognitive scientists Roger Schank and Bob Abelson proposed that in these cases we follow “scripts” that can guide the sequence of behavior. The sociologist Erving Goffman calls the social constraints on acceptable behavior “frames,” and he shows how they govern behavior even when a person is in a novel situation or novel culture. Danger awaits those who deliberately violate the frames of a culture.

  The next time you are in an elevator, try violating cultural norms and see how uncomfor
table that makes you and the other people in the elevator. It doesn’t take much: Stand facing the rear. Or look directly at some of the passengers. In a bus or streetcar, give your seat to the next athletic-looking person you see (the act is especially effective if you are elderly, pregnant, or disabled).

  In the case of the Lego motorcycle of Figure 4.1, cultural constraints determine the locations of the three lights of the motorcycle, which are otherwise physically interchangeable. Red is the culturally defined standard for a brake light, which is placed in the rear. And a police vehicle often has a blue flashing light on top. As for the yellow piece, this is an interesting example of cultural change: few people today remember that yellow used to be a standard headlight color in Europe and a few other locations (Lego comes from Denmark). Today, European and North American standards require white headlights. As a result, figuring out that the yellow piece represents a headlight on the front of the motorcycle is no longer as easy as it used to be. Cultural constraints are likely to change with time.

  SEMANTIC CONSTRAINTS

  Semantics is the study of meaning. Semantic constraints are those that rely upon the meaning of the situation to control the set of possible actions. In the case of the motorcycle, there is only one meaningful location for the rider, who must sit facing forward. The purpose of the windshield is to protect the rider’s face, so it must be in front of the rider. Semantic constraints rely upon our knowledge of the situation and of the world. Such knowledge can be a powerful and important clue. But just as cultural constraints can change with time, so, too, can semantic ones. Extreme sports push the boundaries of what we think of as meaningful and sensible. New technologies change the meanings of things. And creative people continually change how we interact with our technologies and one another. When cars become fully automated, communicating among themselves with wireless networks, what will be the meaning of the red lights on the rear of the auto? That the car is braking? But for whom would the signal be intended? The other cars would already know. The red light would become meaningless, so it could either be removed or it could be redefined to indicate some other condition. The meanings of today may not be the meanings of the future.