Design of Everyday Things - Norman, Don [66]
The connectionist approach to memory might also be called the “multiple-exposure” theory of memory.
Suppose, unbeknownst to you, your camera broke so that the film wouldn’t wind Every picture you took went right on top of all the others. If you had taken pictures of different scenes, you might still be able to make out the individual parts. But suppose you had taken a picture of a high-school graduating class, one person at a time. Each person took a turn sitting in the chair in front of the fixed camera; each smiled; each had a picture taken. Afterward, when you developed the film, you would find just one picture, a composite of all those faces. All the individual records would still be there, but on top of one another, difficult to separate out. You’d have the average high-school graduate.
Throw everything into memory on top of one another. That is a crude approximation of the connectionist approach to memory. Actually, things aren’t thrown together until after a lot of processing has gone on. And memory isn’t really like a multiple exposure. Still, this is not a bad characterization of the connectionist approach.
Consider what happens when two similar events are experienced: they merge together to form a kind of average, a “prototypical event.” This prototype governs interpretations and actions related to any other event that seems similar. What happens when something really discrepant occurs? If it is quite different from the prototype, it still manages to maintain its identity when thrown into memory. It stands out by itself.
If there were a thousand similar events, we would tend to remember them as one composite prototype. If there were just one discrepant event, we would remember it, too, for by being discrepant it didn’t get smudged up with the rest. But the resulting memory is almost as if there had been only two events: the common one and the discrepant one. The common one is a thousand times more likely, but not to the memory; in memory there are two things, and the discrepant event hardly seems less likely than the everyday one.
So it is with human memory. We mush together details of things that are similar, and give undue weight to the discrepant. We relish discrepant and unusual memories. We remember them, talk about them, and bias behavior toward them in wholly inappropriate ways.
What has this to do with everyday thought? A lot. Everyday thought seems to be based upon past experiences, upon our ability to retrieve an event from the past and use it to model the present. This event-based reasoning is powerful, yet fundamentally flawed. Because thought is based on what can be recalled, the rare event can predominate. Think about it. Think of your experiences with computers, or VCRs, or home appliances; what probably come to mind are the unusual experiences, things that are discrepant. It doesn’t matter that you may have used the device a hundred times successfully—it is the one time you got embarrassed that will come to mind.8
The limitations of human thought processes have important implications for everyday activities, and in fact can be called in to distinguish everyday activities from others.
The Structure of Tasks
Everyday activities are conceptually simple. We should be able to do most things without having to think about what we’re doing. The simplicity lies in the nature of the structure of the tasks.
WIDE AND DEEP STRUCTURES
Consider the game of chess, an activity that is neither everyday nor simple, at least, not for most of us. When it is my turn to play, I have a number of possible moves. For each of my moves, my opponent has a number of possible responses. And for each of my opponent’s responses, I have a number of possible