Mnemonics and Expert Knowledge: Mental Cuing
Francis S. Bellezza
I am particularly interested in how special knowledge facilitates (or sometimes interferes with) the learning and recall of new information. This is reflected in my research on mnemonic devices, which are special techniques used to memorize information ( Bellezza, 1981, 1983). Psychologists have long known that the particular knowledge already in memory somehow mediates the learning of new information, but just how this mediation process takes place is not yet fully understood. The research I discuss here addresses the question of how expert knowledge representations in memory influence learning and recall. The study of the relation between expert knowledge and learning, however, is not as common in psychological research as the study of the role of expert knowledge in decision making and problem solving. But as our understanding of expertise increases, the study of learning by experts will, I believe, contribute to the understanding of how experts use their knowledge.
In my research, the knowledge extracted from experts is declarative knowledge, that is, knowledge that is verbal and factual rather than procedural ( Anderson, 1976; Gordon, this volume). This distinction goes back to Ryle ( 1949) and is the distinction between knowing what (declarative knowledge) and knowing how (procedural knowledge). The expert knowledge typically studied by knowledge engineers is some mixture of these two types. Both types of knowledge are important. Experts solve problems by using knowledge of procedures, by using factual knowledge, and by making inferences from their factual knowledge. They do not simply remember facts. Knowledge engineers typically focus on declarative knowledge in the early stages of the knowledge-elicitation process and use a variety of techniques to construct representations of the declarative knowledge of their experts ( Hart, 1986; Hoffman, 1987, 1989; Olson & Rueter, 1987). However, I have limited my investigations to declarative knowledge only. I ignore, at least temporarily, the problem of how knowledge is used by the expert in a procedural manner. Because of this orientation, the issue of declarative knowledge representation and organization in memory become important in my work. Other psychological researchers have also concentrated on the declarative knowledge of experts. Chi and Koeske ( 1983), for example, studied children's expert knowledge of dinosaurs and how that knowledge was organized in memory. McKeithen, Reitman, Rueter, and Hirtle ( 1981) found that experts in the ALGOL W computer language organized reserved words in memory according to their functional meaning, whereas nonexperts organized them using common-language associations. In addition, early expert computer systems, such as systems involving knowledge about baseball ( Green, Wolf, Chomsky, & Laughery, 1963), dealt primarily with declarative knowledge. Thus,
-204-
