Motivating the Notion of Generic Design within Information-Processing Theory: The Design Problem Space

conceptual knowledge here seems to be less fragmentary and more theoretical than with the general schemas. 9 (This fact is not surprising considering that it was acquired as an organized, systematic body of knowledge.) For example, Subject-A had an elaborate minitheory about the use and organization of space between buildings. His first sentences on viewing the site were as follows: (PF3) S-A: Well, what comes to my mind immediately, as I told you before when I was waiting [for] you, I was looking at, how this is set by pathways, this, this open space in between the sports court yard and these three buildings. And in thinking about the missed opportunity that people had here, of having a sort of more relaxed plaza, instead of being just a cross between these two directions. Which makes it very efficient, ah, but for sure it didn’t, ah, give any contribution to the urban open space. Similarly, Subject-I had a minitheory about motivating, teaching, and imparting knowledge: (PF4) S-I: The first thing we want to do with these people is try and sell them on a system. Any time you change somebody from an old system to a new system, or from what they are doing to what they’re going to be doing, or what you’re expecting them to be doing, you’ve got to give them a good positive reason. Why do I really? What’s in it for me, you know. . . . This is positive reinforcement. Several of these protocol fragments (PF1, PF2, PF3) are also examples of what we call scenario immersion. Scenarios are frequently occurring episodes in which designers recall and immerse themselves in rich, intricate images from their past experience. The experience in question could have been acquired directly or vicariously through some symbolic medium (for example, reading, watching television). These episodes seem to play an absolutely crucial role in the process of generation and evaluation. For instance, the scenario in PF1 is used to generate the functional requirement “protection from rain.” In PF2, the scenario is used to evaluate a proposed spatial configuration of APTMs. We say more about scenario immersion in Extensive Performance Modeling. Application of Personal Knowledge. Personal knowledge structures and procedures are stored in LTM. Their indexing and retrieval are not well understood. Problem structuring is the process of finding and retrieving relevant schemata and instantiating them into the problem space. By instantiation, we mean nothing more than the process by which a proposition of the content “All public buildings are required to have a ramp access for the handicapped” is transformed into the proposition that “This building requires ramp access.” As it is construed here, problem structuring is not itself a problem-solving activity. However, the extent to which it is successful does determine the amount of problem solving that needs to occur. Subject-I was able to find, retrieve, and instantiate a single powerful schema for designing training programs. The template came with slots marked for lessons, sections, subsections, and so on. He merely had to fill in the blanks with the content of the particular course. He generated the required content by (1) asking the client (experimenter) for a list of tasks the secretary would be required to perform, (2) drawing on his own personal knowledge of Viewpoint, and (3) consulting the Viewpoint manuals.10 Finally, the selection of content was guided by an idealized cognitive model (ICM) (Lakoff, 1987) of what a secretary is; for example: (PF5) S-I: All this isn’t going to stay in this create and edit documents [lesson]. This is just looking at what’s available, and what we are going to have to do. Because within this table of contents [of viewpoint manual] we’ve got related information—hardware requirements and so forth that has nothing to do with the secretaries, and foundation and environment. Secretaries couldn’t care less. . . . And the logoff sheet properties, I, I wouldn’t even teach the secretaries. That’s none of their business. They have no need for that information. That I would teach your systems administrator. Subject-A, however, seemed to find his design problem more of a challenge and exhibited somewhat different behavior. His initial structuring process took 20 minutes and resembled a brainstorming session. If the protocol for this phase is recorded as a directed graph, with the nodes forming individual ideas as they are uttered in temporal sequence and the arcs connecting related nodes, then the result is a lattice structure. The density and distribution of the links suggest there are really four smaller structures. First are some site-related