RESEARCH ARTICLE The time course of breaking mental sets and forming novel associations in insight-like problem solving: an ERP investigation Junlong Luo · WenFu Li · Andreas Fink · Lei Jia · Xiao Xiao · Jiang Qiu · Qinglin Zhang Received: 26 October 2010 / Accepted: 3 June 2011 / Published online: 19 June 2011 © Springer-Verlag 2011 Abstract In the present study, high-density event-related potentials were recorded to examine the electrophysiolog- ical correlates of logogriph problem solving in using a new experimental paradigm (learning-testing model) that was adopted in order to enable subjects to find a solution on their own initiative. For each trial, subjects were given a target logogriph followed by three types of base logog- riphs: surface similarity logogriphs (SUSL, the base logogriph and target logogriph share some same words), structural similarity logogriphs (STSL, the base and target logogriphs do not have any words in common), and base- line logogriphs (BSL, the base and target logogriphs are all simple character-generation tasks). The results demon- strated a more negative event-related potential deflection during STSL than during SUSL in both the 300–500 and 1,100–1,300 ms time windows, most likely reflecting the breaking of mental sets during insight-like problem solv- ing. Moreover, SUSL and STSL demonstrated greater positivity than BSL in a time window between 900 and 1,700 ms, possibly reflecting processes such as forming novel associations. Keywords Insight-like problem solving · Breaking mental sets · Forming novel association · Event-related potentials Introduction The term “insight” has been used to designate the sudden and unexpected flash knowing of how to solve a problem (Bowden et al. 2005). There was a persistent debate on the question whether insight problem solving involves unique processes or is merely a process based on the same cog- nitive mechanisms as non-insight solutions (Weisberg and Alba 1981; Perkins 1998). Brain imaging techniques such as functional magnetic resonance imaging (fMRI) or electrophysiological mea- sures such as event-related potentials (ERPs) have facilitated the reliable and precise measurement of brain activity associated with many high-level cognitive pro- cesses (e.g., insight problem solving). For example, Luo et al. recorded neural activity using fMRI and correlated activity with cognitive insight by providing a trigger (the solution) to catalyze insightful riddle solving processes (Luo and Niki 2003; Luo et al. 2004). Results showed that insight riddle solving was associated with activity primar- ily in the hippocampus, anterior cingulate cortex (ACC) and the prefrontal cortex (PFC). In studies using the compound remote associates problem (CRA, e.g., boot, summer, ground; solutions: camp), fMRI results revealed an increased signal in the right anterior superior temporal gyrus for insight but not noninsight solutions (Jung-Be- eman et al. 2004; Bowden et al. 2005). Using ERPs methodology, Lavric et al. (2000) found that P300 were more frontal during analytical problem (wason’s selection task) solving as compared to insight Junlong Luo and WenFu Li have equally contributed to this work. J. Luo · W. Li · L. Jia · X. Xiao · J. Qiu (&) · Q. Zhang (&) Department of Psychology, School of Psychology, Southwest University, Beibei, Chongqing 400715, China e-mail: luo831023@163.com Q. Zhang e-mail: luojl@swu.edu.cn J. Qiu · Q. Zhang Key Laboratory of Personality and Cognition, Ministry of Education, Southwest University, Chongqing 400715, China A. Fink Institute of Psychology, University of Graz, Graz, Austria 123 Exp Brain Res (2011) 212:583–591 DOI 10.1007/s00221-011-2761-5