Kurashige Hiroki, Yamashita Yuichi, Hanakawa Takashi, Honda Manabu
Graduate School of Informatics and Engineering, The University of Electro-Communications, Tokyo, Japan.
National Center of Neurology and Psychiatry, National Institute of Neuroscience, Tokyo, Japan.
Front Psychol. 2019 Mar 28;10:600. doi: 10.3389/fpsyg.2019.00600. eCollection 2019.
The results of many studies have suggested that we actively select information from the environment. However, the functional consequences of such selectivity in knowledge acquisition remain unclear, even though it is a vital factor in determining the characteristics of our future knowledge and cognition. We hypothesized that spontaneous selectivity in knowledge acquisition results in effective augmentation of productivity, especially in creativity-demanding task. To test this, we conducted experiments in which subjects acquired novel compositional words during their rapid presentation, evaluated memory confidence rates for the acquired words, and then produced essays based on these words. First, in experiment 1, we showed that the level of confidence in the recognition memory for the words positively related with the length of the essays (a measure of creativity-involving productivity in quantity). Additionally, we found that the semantic distance from the essay to the components of the compositional word (a measure of creative-productivity in quality) was farther for the word with higher memory confidence than for the word with lower memory confidence, suggesting creative leaps when writing the former. While this result supported our hypothesis, it might also reflect better memory that was independent of spontaneous selection. Thus, in a different subject group, we conducted a similar experiment (experiment 2) in which two of the 20 compositional words were presented more often (five times per block) to force memorization. Again, consistent with our hypothesis, essays based on spontaneously memorized words (presented once per block) were significantly longer than those produced using the forcedly memorized words. Therefore, better memory did not explain the higher productivity. Instead, these results suggested that the higher creativity-involving productivity was consequent to spontaneous selectivity in the knowledge acquisition. Additionally, we propose a possible mechanism for the observed results based on the results of a neural network simulation. In this simulation, we found that novel information that was assigned to locations more easily accessible to the entire network was better assimilated and therefore selectively acquired. Based on this simulation, we moderately suggest that spontaneously acquired knowledge effectively confers productivity because it effectively activates large parts of the neural networks.
许多研究结果表明,我们会主动从环境中筛选信息。然而,尽管这种筛选在知识获取中是决定我们未来知识和认知特征的关键因素,但其在知识获取中的功能后果仍不明晰。我们假设,知识获取中的自发筛选会有效提高效率,尤其是在需要创造力的任务中。为了验证这一点,我们进行了实验,让受试者在快速呈现过程中获取新的合成词,评估对所获取单词的记忆置信度,然后基于这些单词撰写文章。首先,在实验1中,我们发现对单词的识别记忆置信度水平与文章长度呈正相关(文章长度是衡量创造力相关效率的一个指标)。此外,我们发现,与记忆置信度较低的单词相比,记忆置信度较高的单词所构成文章与合成词各部分之间的语义距离更远(语义距离是衡量创造力相关质量的一个指标),这表明在撰写前者时存在创造性飞跃。虽然这一结果支持了我们的假设,但它也可能反映了与自发筛选无关的更好的记忆。因此,在另一组受试者中,我们进行了类似的实验(实验2),在20个合成词中,有两个词被更频繁地呈现(每块呈现5次)以促使记忆。同样,与我们的假设一致,基于自发记忆单词(每块呈现1次)所撰写的文章明显长于使用强制记忆单词所撰写的文章。因此,更好的记忆并不能解释更高的效率。相反,这些结果表明,更高的创造力相关效率是知识获取中自发筛选的结果。此外,基于神经网络模拟的结果,我们提出了一个可能的机制来解释观察到的结果。在这个模拟中,我们发现被分配到整个网络更容易访问位置的新信息能被更好地同化,因此被选择性地获取。基于这个模拟,我们适度地认为,自发获取的知识能有效提高效率,因为它能有效激活神经网络的大部分区域。