The Design of Everyday Things (20 page)

The design implications are clear: provide meaningful structures. Perhaps a better way is to make memory unnecessary: put the required information in the world. This is the power of the traditional graphical user interface with its old-fashioned menu structure. When in doubt, one could always examine all the menu items until the desired one was found. Even systems that do not use menus need to provide some structure: appropriate constraints and forcing functions, natural good mapping, and all the tools of feedforward and feedback. The most effective way of helping people remember is to make it unnecessary.

Approximate Models: Memory in the Real World

Conscious thinking takes time and mental resources. Well-learned skills bypass the need for conscious oversight and control: conscious control is only required for initial learning and for dealing with unexpected situations. Continual practice automates the action cycle, minimizing the amount of conscious thinking and problem-solving required to act. Most expert, skilled behavior
works this way, whether it is playing tennis or a musical instrument, or doing mathematics and science. Experts minimize the need for conscious reasoning. Philosopher and mathematician Alfred North
Whitehead stated this principle over a century ago:

          
It is a profoundly erroneous truism, repeated by all copy-books and by eminent people when they are making speeches, that we should cultivate the habit of thinking of what we are doing. The precise opposite is the case. Civilization advances by extending the number of important operations which we can perform without thinking about them
. (Alfred North Whitehead, 1911.)

One way to simplify thought is to use simplified models, approximations to the true underlying state of affairs. Science deals in truth, practice deals with approximations. Practitioners don't need truth: they need results relatively quickly that, although inaccurate, are “good enough” for the purpose to which they will be applied. Consider these examples:

EXAMPLE 1: CONVERTING TEMPERATURES BETWEEN FAHRENHEIT AND CELSIUS

It is now 55°F outside my home in California. What temperature is it in Celsius? Quick, do it in your head without using any technology: What is the answer?

I am sure all of you remember the conversion equation:

°C = (°F–32) × 5 / 9

Plug in 55 for °F, and ºC = (55–32) × 5 / 9 = 12.8°. But most people can't do this without pencil and paper because there are too many intermediate numbers to maintain in STM.

Want a simpler way? Try this approximation—you can do it in your head, there is no need for paper or pencil:

°C = (°F–30) / 2

Plug in 55 for °F, and ºC = (55–30) / 2 = 12.5º. Is the equation an exact conversion? No, but the approximate answer of 12.5 is close
enough to the correct value of 12.8. After all, I simply wanted to know whether I should wear a sweater. Anything within 5ºF of the real value would work for this purpose.

Approximate answers are often good enough, even if technically wrong. This simple approximation method for temperature conversion is “good enough” for temperatures in the normal range of interior and outside temperatures: it is within 3ºF (or 1.7ºC) in the range of –5º to 25ºC (20º to 80ºF). It gets further off at lower or higher temperatures, but for everyday use, it is wonderful. Approximations are good enough for practical use.

EXAMPLE 2: A MODEL OF SHORT-TERM MEMORY

Here is an approximate model for STM:

          
There are five memory slots in short-term memory. Each time a new item is added, it occupies a slot, knocking out whatever was there beforehand
.

Is this model true? No, not a single memory researcher in the entire world believes this to be an accurate model of STM. But it is good enough for applications. Make use of this model, and your designs will be more usable.

EXAMPLE 3: STEERING A MOTORCYCLE

In the preceding section, we learned how Professor Sayeki mapped the turning directions of his motorcycle to his turn signals, enabling him to remember their correct usage. But there, I also pointed out that the conceptual model was wrong.

Why is the conceptual model for steering a motorcycle useful even though it is wrong? Steering a motorcycle is counterintuitive: to turn to the left, the handlebars must first be turned to the right. This is called countersteering, and it violates most people's conceptual models. Why is this true? Shouldn't we rotate the handlebars left to turn the bike left? The most important component of turning a two-wheeled vehicle is lean: when the bike is turning left, the rider is leaning to the left. Countersteering causes the rider to lean
properly: when the handlebars are turned to the right, the resulting forces upon the rider cause the body to lean left. This weight shift then causes the bike to turn left.

Experienced riders often do the correct operations subconsciously, unaware that they start a turn by rotating the handlebars opposite from the intended direction, thus violating their own conceptual models. Motorcycle training courses have to conduct special exercises to convince riders that this is what they are doing.

You can test this counterintuitive concept on a bicycle or motorcycle by getting up to a comfortable speed, placing the palm of the hand on the end of the left handlebar, and gently pushing it forward. The handlebars and front wheel will turn to the right and the body will lean to the left, resulting in the bike—and the handlebars— turning to the left.

Professor Sayeki was fully aware of this contradiction between his mental scheme and reality, but he wanted his memory aid to match his conceptual model. Conceptual models are powerful explanatory devices, useful in a variety of circumstances. They do not have to be accurate as long as they lead to the correct behavior in the desired situation.

EXAMPLE 4: “GOOD ENOUGH” ARITHMETIC

Most of us can't multiply two large numbers in our head: we forget where we are along the way. Memory experts can multiply two large numbers quickly and effortlessly in their heads, amazing audiences with their skills. Moreover, the numbers come out left to right, the way we use them, not right to left, as we write them while laboriously using pencil and paper to compute the answers. These experts use special techniques that minimize the load on working memory, but they do so at the cost of having to learn numerous special methods for different ranges and forms of problems.

Isn't this something we should all learn? Why aren't school systems teaching this? My answer is simple: Why bother? I can estimate the answer in my head with reasonable accuracy, often good enough for the purpose. When I need precision and accuracy, well, that's what calculators are for.

Remember my earlier example, to multiply 27 times 293 in your head? Why would anyone need to know the precise answer? an approximate answer is good enough, and pretty easy to get. Change 27 to 30, and 293 to 300: 30 × 300 = 9,000 (3 × 3 = 9, and add back the three zeros). The accurate answer is 7,911, so the estimate of 9,000 is only 14 percent too large. In many instances, this is good enough. Want a bit more accuracy? We changed 27 to 30 to make the multiplication easier. That's 3 too large. So subtract 3 × 300 from the answer (9,000 – 900). Now we get 8,100, which is accurate within 2 percent.

It is rare that we need to know the answers to complex arithmetic problems with great precision: almost always, a rough estimate is good enough. When precision is required, use a calculator. That's what machines are good for: providing great precision. For most purposes, estimates are good enough. Machines should focus on solving arithmetic problems. People should focus on higher-level issues, such as the reason the answer was needed.

Unless it is your ambition to become a nightclub performer and amaze people with great skills of memory, here is a simpler way to dramatically enhance both memory and accuracy: write things down. Writing is a powerful technology: why not use it? Use a pad of paper, or the back of your hand. Write it or type it. Use a phone or a computer. Dictate it. This is what technology is for.

The unaided mind is surprisingly limited. It is things that make us smart. Take advantage of them.

SCIENTIFIC THEORY VERSUS EVERYDAY PRACTICE

Science strives for truth. As a result, scientists are always debating, arguing, and disagreeing with one another. The scientific method is one of debate and conflict. Only ideas that have passed through the critical examination of multiple other scientists survive. This continual disagreement often seems strange to the nonscientist, for it appears that scientists don't know anything. Select almost any topic, and you will discover that scientists who work in that area are continually disagreeing.

But the disagreements are illusory. That is, most scientists usually agree about the broad details: their disagreements are often about tiny details that are important for distinguishing between two competing theories, but that might have very little impact in the real world of practice and applications.

In the real, practical world, we don't need absolute truth: approximate models work just fine. Professor Sayeki's simplified conceptual model of steering his motorcycle enabled him to remember which way to move the switches for his turn signals; the simplified equation for temperature conversion and the simplified model of approximate arithmetic enabled “good enough” answers in the head. The simplified model of STM provides useful design guidance, even if it is scientifically wrong. Each of these approximations is wrong, yet all are valuable in minimizing thought, resulting in quick, easy results whose accuracy is “good enough.”

Knowledge in the Head

Knowledge in the world, external knowledge, is a valuable tool for remembering, but only if it is available at the right place, at the right time, in the appropriate situation. Otherwise, we must use knowledge in the head, in the mind. A folk saying captures this situation well: “Out of sight, out of mind.” Effective memory uses all the clues available: knowledge in the world and in the head, combining world and mind. We have already seen how the combination allows us to function quite well in the world even though either source of knowledge, by itself, is insufficient.

HOW PILOTS REMEMBER WHAT AIR-TRAFFIC CONTROL TELLS THEM

Airplane pilots have to listen to commands from air-traffic control delivered at a rapid pace, and then respond accurately. Their lives depend upon being able to follow the instructions accurately. One website, discussing the problem, gave this example of instructions to a pilot about to take off for a flight:

          
Frasca 141, cleared to Mesquite airport, via turn left heading 090, radar vectors to Mesquite airport. Climb and maintain 2,000. Expect 3,000 10 minutes after departure. Departure frequency 124.3, squawk 5270
.

          
(Typical Air traffic control sequence, usually spoken extremely rapidly. Text from “ATC Phraseology,” on numerous websites, with no credit for originator.)

“How can we remember all that,” asked one novice pilot, “when we are trying to focus on taking off?” Good question. Taking off is a busy, dangerous procedure with a lot going on, both inside and outside the airplane. How do pilots remember? Do they have superior memories?

Pilots use three major techniques:

       
1.
   
They write down the critical information.

       
2.
   
They enter it into their equipment as it is told to them, so minimal memory is required.

       
3.
   
They remember some of it as meaningful phrases.

Although to the outside observer, all the instructions and numbers seem random and confusing, to the pilots they are familiar names, familiar numbers. As one respondent pointed out, those are common numbers and a familiar pattern for a takeoff. “Frasca 141” is the name of the airplane, announcing the intended recipient of these instructions. The first critical item to remember is to turn left to a compass direction of 090, then climb to an altitude of 2,000 feet. Write those two numbers down. Enter the radio frequency 124.3 into the radio as you hear it—but most of the time this frequency is known in advance, so the radio is probably already set to it. All you have to do is look at it and see that it is set properly. Similarly, setting the “squawk box to 5270” is the special code the airplane sends whenever it is hit by a radar signal, identifying the airplane to the air-traffic controllers. Write it down, or set it into the equipment as it is being said. As for the one remaining item, “Expect 3,000 10 minutes after departure,” nothing need be done. This is just reassurance that in ten minutes, Frasca 141 will probably
be advised to climb to 3,000 feet, but if so, there will be a new command to do so.

How do pilots remember? They transform the new knowledge they have just received into memory in the world, sometimes by writing, sometimes by using the airplane's equipment.

The design implication? The easier it is to enter the information into the relevant equipment as it is heard, the less chance of memory error. The air-traffic control system is evolving to help. The instructions from the air-traffic controllers will be sent digitally, so that they can remain displayed on a screen as long as the pilot wishes. The digital transmission also makes it easy for automated equipment to set itself to the correct parameters. Digital transmission of the controller's commands has some disadvantages, however. Other aircraft will not hear the commands, which reduces pilot awareness of what all the airplanes in the vicinity are going to do. Researchers in air-traffic control and aviation safety are looking into these issues. Yes, it's a design issue.