Let's consider someone who wants to become a member of the sushi cognoscenti, one who is really an expert in sushi. This domain is a good one to discuss because everyone is familiar with learning about new foods. Also, as with most "real-world" domains, this is one for which there is no fixed curriculum. What exactly is it that an expert on sushi knows? What memory structures get built and changed as the person learns?
To start, let's assume that our subject has never tasted sushi before. Having never had sushi, what happens when you eat it for the first time? The answer is that you create a case for that experience. This case contains the story of your sushi-eating experience. It packages information about where you were, what you ordered, how you liked it, and other related details.
Where is this case located in memory? Well, when eating the sushi (or doing any other task), you are comparing it with other things you have eaten before. Since each person has different experiences, each person will have a distinct set of memory structures to use to understand the sushi experience. Typical ways of understanding the experience might be "an example of uncooked food" or "an example of Japanese food" or "an example of seafood." Each of these general categories would then be used to "index" the case in memory. "Indexing" is the process that people use to attach labels like these to the cases they experience. They can then use those labels at a later time to recall relevant cases. So, if you were later asked "When was the first time you ate raw fish?" you would then be able to answer "I had sushi once in such-and-such a place."
What happens if you are asked at this point, "Do you like sushi?" Having only eaten sushi once, you have no generalizations about it. The way you would answer this question is to use case-based reasoning. You would be reminded of your one sushi-eating experience and build an answer around how much you enjoyed it.
One case does not a gourmet make, however. If you wish to become an expert on sushi, then you must get more experience -- you must eat more sushi. You need to experiment with the different variables that affect sushi. If you want to know the extent to which freshness matters, you have to eat some sushi that isn't fresh and some that is especially fresh. To understand how sushi compares to other forms of raw fish, you must try them as well. This means eating sashimi, for example, or eating rare tuna in a restaurant that grills tuna steaks.
What happens as you gain these experiences? One result is that you build additional cases. But you will find, as you gain cases, you begin to build organizational structures in memory that capture what you learn from those cases. You begin to build a theory-of-sushi around your experience, capturing that theory in your organizational memory structures. After becoming experienced with sushi, when asked the question, "Do you enjoy sushi?" you no longer get reminded of your initial sushi eating experience. In fact, you will probably not be reminded of any particular case at all. Instead, you will be able to tap into your organizational memory structures and use rule-based reasoning to give a reply such as, "Normally I don't much care for sushi, but when it's fresh, it's wonderful." The additional cases allow you to distinguish nuances in sushi and the organizational structures represent the rules you have learned from those cases.
People are constantly building theories such as the theory-of-sushi we have been discussing. How do we create such theories? Three processes are critical: expectation-building, expectation-failure, and explanation.
Where am I in the content of the book?
