The Cognitive And Neural Mechanism Of Prototype Activation During Insight Problem Solving

Admittedly, Representation Change (RC) Theory and Progress Monitoring (PM) Theory could explain the cognitive mechanism of insight problem solving to some degree. However, both theories failed to tell the critical factors and to provide detailed explanations of the cognitive progress. Recently developed brain imaging techniques such as functional magnetic resonance imaging (fMRI) and event-related potentials (ERPs) have made it possible for us to record precisely the brain activity associated with many high-level cognitive processes (e.g., insight problem resolving). However, as for now, investigators (e.g., Luo et al., 2003; Mai et al., 2004; Bowden et al., 2003, 2005) tried their best to adopt the advanced brain-imaging techniques to discuss its neural mechanism. The difficulties still lie in: (1)We could not control the time spot of insight occurrence because it is developed when one breaks the unwarranted mental "fixation" and novel, task-related associations are formed on the old nodes of concepts or cognitive skills. (2)There are very few insight problems in the previous studies so that the neural basis of insight problem solving could not be attained by overlapped and averaged operation using fMRI or ERPs.Thus, on the basis of past studies and views, the present study proposed a theory of Prototype Activation during Insight Problem Solving. The prototype matters were the cognitive events that could facilitate the insight problem solving. The prototype matters were noticed by the solver or appears in the mental scope of the solver was named activation. The activation of prototype matters couldn't predicate the happen of insight, only the special information embodied in the prototype matters was noticed by the solver, the insight was happening. The special information was the key heuristic information. The key heuristic information noticed by the solver was named the activation of the key heuristic information. The insight was the activation of the key heuristic information embodied in the prototype matters.The present study tried to establish a new material—a logogriph store and investigated the theory of activation of prototypal matters in insight problem solving through the learning and testing experimental paradigm. In experiments 1 and 2, the results indicated: Learning prototypal logogriph could facilitate logogriph problem solving. And we established a new material (a logogriph store) which included some sortable indexes to each base logogriph, such as difficulty, accuracy and reaction time. Each pair of riddles included one base logogriph and a target logogriph. A new experimental paradigm was also adopted: a two-stage model for learning-testing, so as to research the time course of successful insight problem solving. Firstly, subjects learned a logogriph (a base logogriph with the answer offered)—learning stage, and then they were asked to solve a homotypical logogriph (a target logogriph, and the base logogriph learned beforehand would provide heuristic information for solving the target logogriph)—testing stage.Experiment 3 explored whether activation of prototypal matters was effected by the increasing of the amount of learning through "X vs. X (X=3, 6, 9)" experimental paradigm. And experiment 4 explored whether the activation of prototypal matters was effected by the increasing of the amount of learning and the elicitation index. The results showed that "Difficulty" and "Elicitation" was an effective index and the new experimental paradigm simulated the process of insight. The activation of prototypal matters showed a parallel processing tendency. And these results supported the idea that the activation of the prototypal matters came as a result of automatic processing and the activation of heuristic information in prototypal matters maybe the result of automatic processing.Thus, the two-stage model for learning-testing was designed to make subjects find a solution on their own initiative rather than just receive and comprehend the given answers passively. The learning stage helped subjects obtain useful heuristic information so that they can successfully solve the target logogriph as soon as possible when the target logogriph appeared. We recorded and analyzed high-density ERPs elicited by logogriphs guessed or not. Thus, ERP data allow for more precise examinations of the time course of activation for different stages of successful insight problem solving and provide more valuable results for determining cognitive mechanism of insight.In Experiment 6, the electrophysiological correlates of the "Aha" effects in resolving insight problems were studied using high-density Event-Related Potentials (ERPs) by adopted "one to one" two-stage model for learning-testing. Results showed that Aha logogriphs which subjects guessed correctly (insight occurred) elicited a more positive ERP deflection than did No-aha logogriphs which subjects could not guess (no insight occurred) in the time window from 200-600 ms (P200-600) after onset of the riddle. Aha logogriphs elicited a more negative ERP deflection than did No-aha logogriphs in the time window within 1500-2000 ms (N1500-2000) and 2000-2500 ms (N2000-2500) after onset of the logogriph. Maps of the difference wave (P200-600) showed strong activity in the midline parieto-occipital scalp. Dipole analysis localized the generator of P200-600 in the left superior temporal gyrus and near parietotemporo-occipital cortex areas. N1500-2000 and N2000-2500 all had a distinct activation over left frontal scalp. Dipole analysis localized the generator of the N1500-2000 in the anterior cingulate cortex (ACC) and of the N2000-2500 in the posterior cingulate cortex (PCC). In Experiment 7, the electrophysiological correlates of the "Aha" effects in resolving insight problems were studied using high-density Event-Related Potentials (ERPs) by adopted "five to five" two-stage model for learning-testing. Results showed that Aha logogriphs elicited a more negative ERP deflection than did No-aha logogriphs in the time window within 1400-1700 ms (LNC1), 1700-2000 ms (LNC2) and 2000-2500 ms (LNC3) after onset of the logogriph. Dipole analysis localized the generator of LNC1 in the left Gyrus frontalis superior, LNC2 in the left frontal cortex and LNC3 in the posterior cingulate cortex (PCC).In a word, these results indicate that there were no any difference of spatiotemporal cortical activation patterns underlying the early stage of insight problem solving between Aha logogriphs and No-aha logogriphs. In the process of activating heuristic information and then hastening the onset of insight, LPFC/ACC played an important role, which may be associated with the breaking of mental set successfully and the forming of novel associations in insight. The "Aha" feeling mainly provoked the PCC brain region in the solving the logogriph at last.