Abstract:Six subjects were asked to read and try to understand the first part of Einstein's 1905 paper on special relativity. On the basis of the diagrams drawn by the subjects while they read the text, their thinking aloud protocols, their responses to questions, and other observations, we have inferred the processes they used to understand the paper, and the role that mental images (mental representations) and their paper-and-pencil drawings played in achieving the understanding. In particular, we discuss the characteristics of the representations the subjects formed and the processes they used to form them, as well as the ways in which they derived or reached an understanding of the equations of relativity, using the mental images to assist them. From our analysis we infer that subjects translating the natural-language text into images are able to manipulate, control, and observe these representations to run simple "mental experiments," and by these means draw out the qualitative and quantitative implications of the text. The mental representations and drawings appear to mediate between the initial natural language text and the final equations. In no case did any subject achieve an understanding of the equations without using this kind of intermediate representation. For some decades, imagery has been an active research area in cognitive science (c.f., Paivio 1971 ; Shepard and Metzler 1971; Simon 1972; Pylyshyn 1973; Kosslyn 1980, 1983; Shepard and Copper 1982; Pinker 1985; Finke 1989). Research on imagery's neuro psychological base is also progressing (Farah 1985; Kosslyn 1987; Kosslyn, SokoIv and Chen 1989; etc.). And the role of imagery in education has been examine (e.g., Marte11991). However, there are only a few publications of experimental work concerned mainly with imagery in problem solving (e.g., Simon and Barenfeld 1969; Hayes 1973; Novak 1976; and some aspects in Larkin and Simon 1987). Our study, in which subjects were asked to read and try to understand the first part of Einstein's 1905 paper on special relativity, focuses on imagery in understanding and problem solving, in mental modeling (c.f. Gentner and Stevens 1983; Johnson-Laird 1983; Norman 1988) and in qualitative reasoning (c.f. Iwasaki 1989; Weld and de Kleer 1990). In Qin and Simon (1990, 1991), we discussed some results of our project relating to problem solving success. In this paper, we summarize very briefly our major findings about imagery and mental models.

Methods of Thinking with Imagery for Fault Diagnosis

Intelligent Fault Diagnosis of CNC Machine Tools Based on Rough Set Theory

FAULT DIAGNOSIS FOR POWER ELECTRONIC CIRCUITS BASED ON HYBRID SYSTEM THEORYAND EVENT IDENTIFICATION

A Method Of Printed Circuit Board Failure Diagnosis Based On Infrared Image Processing

A Fault Diagnosis Expert System for Hydraulic System of Injection Moulding

Visual method of fault diagnosis for a complicated process

Power System Fault Diagnosis Method Based on Intuitionistic Fuzzy Sets and Incidence Matrices

Imagery and Mental Models in Problem Solving

Intelligent fault diagnosis using rough set method and evidence theory for NC machine tools

A Practical Approach To The Integrated Expression Of Shallow And Deep Knowledge In Fault Diagnosis Systems

An Analytic Model and Optimization Technique Based Methods for Fault Diagnosis in Power Systems

Research on Multiple Faults Diagnosis Method Based on Case Reasoning

A New Comprehensive Fault Diagnosis Approach Based on Fuzzy Theory

The research and application of intelligent fault diagnosis methods

A general design technique for fault diagnostic systems

The Optimization Method of Mechanical Fault Diagnosis Based on Artificial Intelligence Technology

New Approach of Integrated Intelligent Fault Diagnosis Based on Rough Set Theory

Study on Fault Multi-Symptom-Domain Integrated Diagnosis Based on Fuzzy Integral Theory

The visual mental image base for thinking by image

Process fault diagnosis with model- and knowledge-based approaches: Advances and opportunities

Enhancing Fault Troubleshooting through Human-Machine Collaboration: A Multi-Stage Reasoning Approach