We know what stops you from thinking forever: A metacognitive perspective

Abstract

This commentary addresses omissions in De Neys's model of fast-and-slow thinking from a metacognitive perspective. We review well-established meta-reasoning monitoring (e.g., confidence) and control processes (e.g., rethinking) that explain mental effort regulation. Moreover, we point to individual, developmental, and task design considerations that affect this regulation. These core issues are completely ignored or mentioned in passing in the target article.

Effort Regulation in Meta-reasoning: Monitoring and Control

“…These models also predict that as the amount of effort invested in a particular task item increases, the confidence level rises until it reaches or exceeds the solver's pre-determined confidence target [30,31]. However, the model does not explain why some of the reasoners' confidence levels did not rise as they invested more time, which could be because a person lacked the knowledge needed to solve the problem, or simply because no acceptable solution came to mind [32]. After a long period of reflection, judgements are often made with less confidence, leading to an inverse relationship between time and confidence [25].…”

Metacognitive Effort Regulation across Cultures

“…These models also predict that with increased effort investment in a particular task item, confidence levels should rise until they meet or exceed the solver’s preset confidence goal. However, these models do not address cases in which confidence does not rise as more time is invested, either because one lacks the knowledge needed to solve the problem or simply because no acceptable solution comes to mind ( Ackerman and Morsanyi 2023 ). Indeed, numerous studies have shown that after lengthy thinking time, people tend to provide low confidence judgments, resulting in persistent inverse relationships between time and confidence (e.g., Koriat et al 2006 ).…”