Cultural Models, Schemas, and Scripts

From WikEd

In this section of Chapter 5, Cole examines the role of schemas and scripts.

A schema, in the sense it's used here, is a sort of "default model" of dealing with the world. While (like all models), schemas sometimes fail to capture the essence of a situation, they do serve as a simplifying mechanism. We assume that otherwised unsupported objects will fall when they are released. We assume that turning on a light switch (even in a room we've never been in before) will produce light. We assume that a hammer-shaped object is intended for striking things.

Scripts, in the sense the word is used here, are a subtype of schema. They specifically relate to everyday activities. Cole mentions Schank and Abelson, who did a great deal of work with scripts while working on artificial intelligence in the 1970s. Schank and Abelson used a restaurant scenario as one of their examples (as did Nelson, also mentioned in this section). I suspect that this is because a restaurant is a familiar situation to most people in our culture, and is a social situation in which the expected behavior can be easily modeled by a script (I don't have Schank and Abelson's work in front of me, so the examples below are probably not exactly the same as theirs).

• John went to a restaurant and ate a steak. Then he went home.

We all assume that a waiter came and took John's order, that John found the steak acceptable, and that he paid the bill. Note that this is very similar to Nelson's three-year-old's description of the restaurant. The difference lies in the unspoken assumptions of what went on.

• Jane went to a restaurant, took a tray from a stack, and got in line.

Here we infer that this is a cafeteria-style restaurant. There are no waiters. The tablecloths (if any) and napkins probably aren't linen. The tableware may well be plastic. However, we do assume that Jane got a meal and ate it, and that the meal was at least marginally edible.

• John went to a restaurant and sat down at a table. After 45 minutes he got up and left without paying.

Perhaps John is a thief, but we probably assume that either the waiter never came to take John's order, that the food never arrived, or that the food was unacceptable.

In Nelson's cited work, a five-year-old's description of the restaurant is much richer than the three-year-old's, but Cole points out that even the richest description leaves out much of what goes on (such a description would be unreadably long and tedious, even if it were feasible to write).

Cole argues that schemas and scripts aren't sufficient to account for thought and action, and points out that the idea of scripts as "in the head" knowledge doesn't fully mesh with his theory of artifact mediation. This seems right to me. While we do use scripts, those scripts are transformed according to the cultural context in which we find ourselves.

Suppose we visit a restaurant that serves cuisine from a culture that's unfamiliar to us. We may be faced with utensils and social conventions that we don't recognize. Some parts of the default "restaurant script" will probably still be valid (e.g., we probably won't assume that the food is free), but others won't. Faced with this situation, a sophisticated adult will most likely fall back on observing the behavior of others and doing as they do, or asking questions of companions or the restaurant staff. A child or a less-sophisticated adult may stick to the script, to the extent of demanding food items or utensils that aren't appropriate, or even consuming items that aren't actually food (the canonical social gaffe here is assuming that a finger bowl with a lemon wedge is some type of soup).

Question: what other ways (if any) do we have of generating new scripts, other than direct observation or questioning of other people?