关键词: 顿悟, 问题解决, 意识, 执行功能, 无意识信息加工过程假设模型

Abstract: Insight problem-solving, a cornerstone of human creativity, fundamentally relies on the interplay between unconscious information processing and executive function regulation. Despite extensive theoretical emphasis on unconscious contributions, prior research has largely failed to directly capture the dynamic transition of weakly associated information from unconscious to conscious states or elucidate the precise regulatory role of executive control. To address these critical gaps, this study proposes and empirically validates a novel cognitive model that redefines the information processing dynamics in insight. The model posits that the brain undergoes a systematic replacement of dominant strong associations with latent weak associations during unconscious processing, culminating in the conscious “aha” moment when executive function modulation permits this transition. Crucially, we demonstrate that executive function does not merely suppress unconscious processes but dynamically shifts from inhibitory to permissive states, enabling weak associations to surface into awareness. This work significantly advances the field by moving beyond descriptive accounts to provide mechanistic, testable insights into the subconscious underpinnings of creativity.
The primary innovation of this study lies in formulating and empirically testing a comprehensive cognitive model that integrates unconscious dynamics with executive regulation. This model seeks to bridge the theoretical and methodological gaps between these two key aspects of insight problem-solving. Unlike previous models that treated unconscious processing as a black box, ours specifies discrete stages: initial suppression of weak associations by executive control, followed by a critical “release phase” where reduced executive activity allows unconscious weak associations to gain neural prominence. This model was not merely hypothetical; it was systematically operationalized through a multi-method experimental design that uniquely combined behavioral, neuroimaging, and neuromodulation techniques to isolate and quantify the proposed mechanisms. Specifically, we introduced subliminal word detection tasks to present weakly associated problem-relevant cues below conscious awareness, ensuring that any influence on insight could be attributed solely to unconscious processing. This approach overcame the methodological limitations of earlier studies that relied on post-hoc self-reports, which often conflate conscious and unconscious contributions. By embedding these cues within classic insight problems (e.g., compound remote associate tasks), we directly tracked how unconscious weak associations evolve toward conscious realization, providing evidence of this transition.
Methodologically, this research pioneers an unprecedented integration of fMRI with advanced analytical techniques to capture the spatiotemporal dynamics of unconscious processing. We moved beyond standard activation mapping by implementing two innovative neuroimaging strategies: (1) dynamic region-of-interest (ROI) analysis focused on executive control networks (e.g., dorsolateral prefrontal cortex, anterior cingulate cortex), which revealed time-locked deactivation in these regions precisely during the unconscious incubation phase preceding insight; and (2) representational similarity analysis (RSA) applied to fMRI data, which quantified neural pattern shifts from strong to weak association representations over milliseconds. This dual approach allowed us to demonstrate that weak associations initially activate in sensory and default-mode networks during unconscious processing, only becoming dominant in prefrontal regions upon conscious insight—a sequence unobservable with conventional methods. Critically, RSA confirmed that the neural similarity between weak association patterns and the solution state increased incrementally during unconscious processing, directly supporting our model’s replacement hypothesis. These techniques collectively provided the first high-resolution neural signature of unconscious-to-conscious transition, resolving long-standing ambiguities about where and when weak associations gain traction in the brain.
Through targeted behavioral interventions and neuroregulatory methods, we plan to conduct causal tests. We designed implicit cueing paradigms to subtly prime weak associations without participants’ awareness, coupled with transcranial direct current stimulation (tDCS) to selectively inhibit prefrontal executive regions. Further innovation emerged through targeted behavioral and neuromodulation interventions that causally tested the role of executive function deactivation. We designed implicit cueing paradigms to subtly prime weak associations without participants’ awareness, coupled with transcranial direct current stimulation (tDCS) to selectively inhibit prefrontal executive regions. These interventions established a causal chain: executive deactivation enables unconscious weak associations to dominate processing, which then triggers conscious insight. This goes beyond correlational findings in prior work by providing mechanistic leverage for modulating creativity.
Collectively, this study transforms our understanding of insight by reframing it as an executive-regulated information replacement process rather than a sudden, unexplained event. The model we proposed employs methods such as subconscious detection, dynamic functional magnetic resonance imaging, and transcranial direct current stimulation intervention, aiming to provide a unified framework for the processing, storage, and retrieval of weak unconscious associations under the regulation of executive functions.

Key words: insight, problem solving, unconscious processing, executive function, unconscious information processing model