Friday Lunch Seminar (English)
November 12, 2021
12:15 〜 13:00
Apply for participation from here by noon, November 11.
You will be notified of participation details by e-mail on November 11.
Talk Title: Insight problem solving as meta problem solving to find what to be a problem
School of Knowledge Science
Japan Advanced Institute of Science and Technology (JAIST)
Host: Satoshi Nishida
Insight problem solving is one of major topic in cognitive science, and past literature has investigated cognitive mechanism underlying the “Aha!” experience in problem solving. Although the literature has accumulated findings on the phenomenological aspect of the “Aha!” experience, we still have little theoretical understanding of the underlying mechanism. In this talk, we propose the meta-problem solving hypothesis: insight or “Aha!” experience would occur, when the class of problems, called the first and second order problem, almost simultaneously solved. The first order problem is a well-posed problem, while the second order problem is a problem to specify the first order problem from a given non-well-posed problem. Specifically, we report an behavioral study on the binary image task, in which the subject is asked to find and report what and where the animal is in an black-and-white image. In this task, the first order problem is to answer what the target is for a specified sub-area of the image, while the second order problem is to specify which sub-area of image may have an object to answer. The binary image task often induces the “Aha!” experience at the moment of finding the target object, and we analyzed eye-movement profile a moment before the finding with or without a Aha experience. Our statistical analysis revealed the average eye movement patterns, which are different for the two sets of trials with or without Aha moment. This result is consistent with the meta-problem solving hypothesis: eye-moment often converged into the taget rapidly, when the subject experienced Aha. In such a case, the two types of the problems, to find where and what the animal is, are considered almost simultaneously solved.