Explanation is ubiquitous. We wonder why events unfold in particular ways, why people behave as they do, and why objects have some properties rather than others. Moreover, we have strong and systematic intuitions about what counts as a satisfying explanation, and we often improve our understanding of the world as a consequence of seeking explanations. Why is explanation such a basic part of cognitive life, and what is it about explaining that generates these cognitive consequences?

The funded research developed and tested an account of explanation motivated by these questions. The core idea is that when children or adults engage in explanation, they seek explanations that are “good” in the sense that they appeal to explanatory hypotheses that are simple and broad. As a result, explainers are more likely to go beyond the obvious in search of subtle patterns in the world. Our research has found that both adults and preschool-aged children are more likely to discover such patterns when they are prompted to explain their observations (versus engaging in a control task, such as describing their observations or thinking out loud). In many contexts this is beneficial, but explaining can also lead adults to perseverate in looking for patterns that are not there, and it can lead children to overlook perceptual details.

This research has potential implications for education and for human computer interaction. Within education, explanations are used to assess student understanding, communicate content, and even as a way to foster understanding. Maximizing the effectiveness of explanation as a pedagogical tool requires an understanding of when and why explanation contributes to learning. The funded work contributes to this understanding. Second, explanation is relevant to computer science, where explanation-based learning has been developed as a mechanism for learning from small samples, and where explanations can serve as the input to or output from expert systems and complex algorithms. These enterprises can inform the psychology of explanation and in turn benefit from empirical findings.

Finally, this body of research has important theoretical implications. Explanation is often invoked in theories of conceptual representation and reasoning, and explanation has been shown to play an important role in categorization, causal reasoning, generalization, probability assignment, and learning. A better theoretical and empirical understanding of explanation is not only valuable in its own right, but as a window onto these foundational cognitive processes.

Explanation is particularly central to one approach in cognitive science that draws an analogy between cognition and science, with people as “folk scientists” who develop intuitive theories about the world. This approach typically appeals to an unanalyzed and untested notion of explanation in articulating the content and function of intuitive theories. A precise characterization of explanation is therefore of value in developing this general approach, which has proven fruitful in understanding many aspects of cognition, particularly those that involve inductive inferences. 

Last Modified: 05/31/2018
Modified by: Tania Lombrozo