Thought Experiments and Imagery in Expert Protocols

This paper focuses on case studies from think-aloud protocols with expert scientists solving explanation problems in which they appear to make predictions for novel systems they have never seen by “imagining what will happen” or “running” the system in their heads. Nersessian [2002] has proposed, based on her reading of historical records of investigations in scientific work such as Maxwell’s work on electromagnetic field theory, that thought experiments can play a role in scientific theory formulation. Such thought experiments are intriguing because (1) they appear to play a powerful role in science; and (2) the subject appears to gain something like empirical information without making any new observations. This raises what I call the fundamental paradox of thought experiments, expressed as: “How can findings that carry conviction result from a new experiment conducted entirely within the head?” Here I will analyze examples of thought experiments from think aloud protocols in order to examine some approaches to resolving the paradox and to begin to explore the breadth of circumstances in which thought experiments (TE’s) of different types are used. Scholars have long been intrigued with the nature of thought experiments in science but the definition and scope of the term thought experiment has remained controversial, as have theories of the mechanisms by which they work. This motivates developing a taxonomy and theory of thought experiment processes based on observations from expert protocols. Findings from such studies may also give us a more systematic way to analyze the types of TE’s used in the history of science and in instruction.