P300 and creative chunk decomposition

Abstract

Creativity plays a critical role in human development, and insight problem solving is a crucial aspect of creativity. One way to access insight is chunk decomposition, which hypothesizes that the difficulty of chunk decomposition is determined by chunk tightness: it is more difficult to decompose tight chunks than loose chunks because tight chunk decomposition involves the removal of meaningless elements, whereas loose chunk decomposition involves the removal of meaningful elements. However, previous studies did not exclude certain confounding factors (eg, spatial relation) considering that the process of chunk decomposition can also be influenced by them. The current study aimed to examine the neural basis of chunk decomposition. Specifically, we focused on addressing the question upon whether this element-type-based chunk decomposition is related to the P300 effect after controlling for spatial relation. To this aim, an adapted version of the chunk decomposition task in Chinese characters was introduced in this study. Participants completed the chunk decomposition task in both tight and loose chunk conditions in which the tightness of a chunk is determined by the type of the removed part in the source character: whereas the removed parts were meaningless strokes in the tight chunk condition, the removed part was a meaningful character in the loose chunk condition. For both conditions, the removed part and the left part were spatially kept in an up-down structure in the source characters. For example, in the tight chunk condition, the source character “买” consists of a stroke and a character “大”. In the loose chunk condition, the source character “奇” consists of a character “大” and another character “可”. In both tight and loose chunk conditions, the removed element, the source character and the left element were presented one after another and participants were asked to remove one part from a source character to get another valid character. Meanwhile, the event-related potentials (ERPs) were recorded after the onset of the source character. The behavioral data including solution rates as well as reaction times and the ERPs data including the amplitudes of P300 were collected in both tight and loose chunk conditions. The repeated measures of ANOVAs were run for the comparisons of these data. The behavioral results showed that: (ⅰ) there was no significant difference on the solution rates between the tight chunk condition and the loose chunk condition; (ⅱ) tight chunk decomposition took longer time to complete than loose chunk decomposition. The ERPs results showed that the element-type-based chunk decomposition elicited a significant P300 effect: the amplitude of P300 deflections was greater in the tight chunk condition than in the loose chunk condition. Maps of the P300 manifested strong activations over the parietal scalp regions. Our study suggests that element type is one of the difficulties in chunk decomposition. Moreover, chunk decomposition might involve representation updating in perceptual transition.