Design of Everyday Things - Norman, Don [64]
SOME MODELS OF HUMAN THOUGHT
Psychologists have chronicled the failures of thought, the nonrationality of real behavior. Even simple tasks can sometimes throw otherwise clever people into disarray. Even though principles of rationality seem as often violated as followed, we still cling to the notion that human thought should be rational, logical, and orderly. Much of law is based upon the concept of rational thought and behavior. Much of economic theory is based upon the model of the rational human who attempts to optimize personal benefit, utility, or comfort. Many scientists who study artificial intelligence use the mathematics of formal logic—the predicate calculus—as their major tool to simulate thought.
But human thought—and its close relatives, problem solving and planning—seem more rooted in past experience than in logical deduction. Mental life is not neat and orderly. It does not proceed smoothly and gracefully in neat, logical form. Instead, it hops, skips, and jumps its way from idea to idea, tying together things that have no business being put together; forming new creative leaps, new insights and concepts. Human thought is not like logic; it is fundamentally different in kind and in spirit. The difference is neither worse nor better. But it is the difference that leads to creative discovery and to great robustness of behavior.
Thought and memory are closely related, for thought relies heavily upon the experiences of life. Indeed, much problem solving and decision making takes place through attempts to remember some previous experience that can serve as a guide for the present. There have been many theories of human memory. For example, every method of filing things has shown up somewhere along the line as a model for human memory. Do you file photographs neatly in a scrapbook? One theory of memory has postulated that our experiences are neatly encoded and organized, as if in a photo album. This theory is wrong. Human memory is most definitely not like a set of photographs or a tape recording. It mushes things together too much, confuses one event with another, combines different events, and leaves out parts of individual events.
Another theory is based on the filing cabinet model, wherein there are lots of cross references and pointers to other records. This theory has a good deal going for it, and it is probably a reasonable characterization of the most prominent approach today. Of course, it is not called a file cabinet theory. It goes by the names of “schema theory,” “frame theory,” or sometimes “semantic networks” and “propositional encoding.” The individual file folders are defined in the formal structure of the schemas or frames, and the connections and associations among the individual records make the structure into a vast and complex network. The essence of the theory consists of three beliefs, all reasonable and supported by considerable evidence: (1) that there is logic and order to the individual structures (this is what the schema or frame is about); (2) that human memory is associative, with each schema pointing and referring to multiple others to which it is related or that help define the components (thus the term “network”); and (3) that much of our power for deductive thought comes from using the information in one schema to deduce the properties of another (thus the term “propositional encoding”). 6 To illustrate the third concept: once I learn that all living animals breathe, I know that any live animal I will ever meet will breathe. I don’t have to learn this separately for all animals. We call this the “default value.” Unless told otherwise, anything I learn for a general concept applies to all of its instances by default. Default values do not have to apply to everything—I can learn exceptions, such as that all birds fly except for penguins and ostriches. But defaults hold true unless an exception shows otherwise. Deduction is a most useful and powerful property of human memory.
THE CONNECTIONIST APPROACH
We still