As recent studies of cognitive psychology (Lave; Beach) pointed out, there is deep-rooted gap between the uses of cognition in the classroom and everyday situation. Previous survey on the elementary schoolers' math problem-making and-solving at the University of Tokyo indicated that over 40 percent of children produced nonsensical or unrealistic word problems. The survey also showed that although it was informed in advance that there might be some queer problems in the questionnaire, about 60 percent of elementary schoolers calculated and 'solved' these queer problems. Attempting a breakthrough to the Language Game of School Math (Ueno), I developed a simple software named TEACHER, which solves any math word problems whether it is queer or not. It solves problems made and input by children in a dialogical fashion, but naturally it can not interpret the meaning of the problems. Regardless of their meaning, TEACHER checks problems only in terms of syntactics, then picks up numbers and gives calculation to them. Children can gradually see its rule-based idiocy and its lack of mind. In this study, I put this software into use in the experimental class of sixth graders in order to challenge the children's ways with math in the classroom. The children in the class were asked to "Become a Computer" and were engaged in the 'human simulation of computer' activity. Acting as a computer, the children solved problems in line with six simple rules that drive TEACHER. As the children go on to solve queer problems according to these rules, they become aware of the fact that the rule-governed silly computer solves problems just the same way as they had themselves solved before. The experimental class is effective enough to bring the children to realize the nature of their ways with math in school, and to decrease the number of children who solve and make queer problems.
抄録全体を表示