Abstract: Cognitive insight phenomenon is one of the core components of creativity. Depending on brain imaging techniques such as functional magnetic resonance imaging (fMRI) or electrophysiological measures such as event-related potentials (ERPs), in the past ten years, neural correlates study of insight has been developed rapidly. During the time, numerous valuable studies have adopted RAT (CRAT) or Chinese logogriph to explore the neural basis of insight problem solving. Although the two style tests have been viewed as excellent materials, most studies using RAT or logogriph fail to report consistent findings. Inconsistent findings may come from different analysis method, experimental paradigm or other control variable. We think that the two style test based different cultures may have similar cognitive process, but distinct cognitive component would be existed in the two style cognitive insight. To explore the common and distinct cognitive component underlying logogriph and remote associate test, we selected Chinese remote associates test (RAT) and logogriph of Chinese characters as the materials and adopted spontaneous insight within-task comparison paradigm in the study. Fourteen participants (6 males) were recruited in the ERP experiment. Each trial began with a central fixation cross for 500ms, and then the insight questions (logogriph or RAT) was presented in the center of screen for 10 sec. Participants were instructed to try to work out the solution to the question within this 10ms. After a 500ms fixation cross, the possible answer or solution to the question was then presented in the center of the screen for 3 sec, followed by a random interval ranging from 300 to 500ms. Participants were required to press the “1” or “2” key of number keyboard to indicate whether they got the right answers. If they think of the answer but it might not be correct, they were asked to press “1” key with their index finger of right hand; if they did not think of a answer, they do not need to press any key. When the right answer was displayed on the screen, they were required to press “1” key if their own answer were consistent to the right answer; if their own answer were not consistent to the right answer, they were asked to press “2” key. In the present study, we mainly forced on the correct reaction, which the participants guessed an answer and it was consistent with the subsequent correct answer. So the dependent variables including accurate rate and right reaction time. Behavioral data showed that logogriph problems were more difficult for subjects to solve, compared to solving RAT, F(1,13)=8.40, p<0.001, the average number of problems that subjects solved successfully for logogriph, RAT were 32±6、48±6, respectively. However, there were no differences between logogriph (RT(logogriph)=4273±367ms) and RAT(RT(RAT)=4198±737ms) across mean right RTs ( F(1,13)=0.72, p>0.05),which indicated that both of insight problems right RTs were consistent. ERP waveform analysis revealed that the anterior P170 were elicited by logogriph and RAT, there were no main effects of task type for amplitudes of the component. Successfully solving logogriph elicited a more positive defection (P650) than RAT responses did in 600~700ms, but there was no significant difference among the two responses in -1600~-1000ms with the baseline prior-response 200 ms. It is remarkable that successfully solving logogriph elicited a more positive defection than RAT responses did in -1000~-300 ms, and voltage maps of the difference waves (logogriph–RAT) showed strong activity in the right frontal regions. These results imply that two style problems have common cognitive components in the whole cognitive insight, but in restructuring stage and Aha experience stage, both of the cognitive components have relative differences. We inferred that the different of feeling of suddenness may result from inconsistent restructuring.

Key words: insight, logogriphs, RAT, event related potential

摘要： 采用自发顿悟范式，通过事件相关电位(ERPs)探讨字谜和远距离联想（RAT）两类顿悟问题解决中的认知差异。结果发现，两类任务的正确反应时均在4000ms左右；解决两类任务的初期都在170ms时出现了正成分，且两类任务在此成分上没有显著差异；在600~700ms内，字谜比RAT诱发了一个更正的ERP成分，主要激活了中后部的脑区；在按键前的800~400ms内，字谜较RAT在右前额诱发了更正的脑电成分。结果表明，字谜顿悟和RAT顿悟存在相似的加工过程，但在重构阶段和啊哈体验上存在差异；两类任务在解决问题时重构过程的不同可能是造成情绪体验差异的原因

关键词: 顿悟, 字谜, 远距离联想, ERP