Empirical Advances in the Neuroscience of Creativity
Thursday, May 22, 3:30 - 4:50 PM, ICM Auditorium
Individual and Neural Differences in Learning Preferences Point to A Neurocognitive Framework for Design Fixation
Alexandra E. Kelly, Drexel University
Dong Ho Kim, Northwestern University
Julie Milovanovich, University of North Carolina at Charlotte
John Gero, University of North Carolina at Charlotte
&
Evangelia G. Chrysikou, Drexel University
Design fixation refers to the tendency to reproduce elements of examples during problem solving. Although mechanical engineers frequently show evidence of design fixation, product designers are less susceptible to fixation effects during design problem solving. We hypothesized that—beyond disciplinary differences—individual differences in how participants approach learning tasks may partially underlie these effects: Abstraction learners tend to extract generalizable rules and apply them to new situations thus resisting fixation, whereas exemplar learners tend to memorize specific examples and reproduce them during learning thus increasing their susceptibility to fixation. In this study our goal was to provide behavioral and neural evidence of the two different learning profiles, and relate them to behavioral evidence of design fixation. Mechanical engineering and product design students were administered behavioral learning and creative problem solving tasks. They also completed a category learning task and design tasks while undergoing functional magnetic resonance imaging (fMRI). Our results point to neural differences in large-scale brain networks during categorization learning and design problem solving between participants who are classified as either abstraction or exemplar learners based on their behavioral performance on a learning task outside of the scanner. Our results further suggested that these neural differences are associated with participants’ tendency to fixate to pictorial examples during problem solving. We discuss a neurocognitive framework for fixation to pictorial examples and its implications for theories of learning and creative problem solving
Default and Executive Networks Support Category Switching: A representational similarity analysis (RSA) of memory search during an associative fluency task
Simone A. Luchini, Pennsylvania State University
John D. Patterson, Pennsylvania State University
Catherine Carpenter, Pennsylvania State University
Heath E. Matheson, Memorial University of Newfoundland
Roger E. Beaty, Pennsylvania State University
Creativity involves the association of distant concepts in semantic memory, yet how the brain supports semantic association remains unclear. When associating two concepts, people tend to engage in either category switching—linking concepts from different semantic categories (e.g., vegetables, animals, furniture)—or category clustering—linking concepts from the same category. We devised an associative fluency task based on polysemous words—words with multiple meanings based on the context they appear in (e.g., server; computer-server, restaurant-server)—and administered this task to participants during fMRI (N = 18). For each trial, participants saw a polysemous word and generated a creative associate to that word, and then generated a second associate to their first response. Participants were instructed to either switch categories between their first and second association, or to cluster responses within the same category. As expected, category switching led to more semantically distant prompt-response pairs than clustering. Using representational similarity analysis (RSA), we identified key brain regions of the default mode and executive control networks which respond differently to category switching and category clustering, with stronger reinstatement of information between successive clustering responses. These findings contrasted with primary sensory regions, such as the primary visual cortex, which were insensitive to clustering and switching. Our results suggest that switching categories during an associative task involves the coordination of default and executive brain regions, highlighting a neurocognitive process relevant for creativity.
EEG Analysis Reveals Neural Mechanisms of Design Creativity
Morteza Zangeneh Soroush, Concordia University
Yong Zeng, Concordia University
Design creativity emerges from the design process and involves tackling ill-defined problems through iterative, open-ended exploration. While general creativity may generate novel ideas without specific constraints, design creativity requires problem-solving within functional and practical requirements, making it inherently complex. Models such as the recursive logic of design state that design creativity arises from complex, nonlinear, and recursive design dynamics. Despite extensive theoretical models, the neural mechanisms of creativity remain largely unexplored. Neuroimaging modalities provide novel insights into brain dynamics, enabling the experimental analysis of neural characteristics and offering valuable tools to evaluate theoretical models. In this talk, we will present the recursive logic of design and demonstrate how this model captures the complex and dynamic nature of design creativity. We will discuss our EEG-based experiments and findings, highlighting key insights from our research. Additionally, we will introduce our EEG dataset, recorded from loosely controlled design creativity experiments. This loosely controlled framework enhances ecological validity, ensuring that the data remains realistic, reliable, and reflective of authentic design processes. We propose this talk for the 2025 Annual Meeting of the Society for the Neuroscience of Creativity, believing it will make a significant contribution to the field of creativity. By reinforcing the recursive logic of design, introducing our EEG dataset, and shedding light on the brain dynamics associated with design creativity, our talk paves the way for new approaches in creativity neuroscience.
Neural Dynamics of Mind-Wandering: Evidence from Intracranial Recordings During Naturalistic Film Viewing
Christine Chesebrough, Feinstein Institutes of Biomedical Research
Maximilian Nentwich, Feinstein Institutes of Biomedical Research
Joshua Assi, Feinstein Institutes of Biomedical Research
Ashesh Mehta, Feinstein Institutes of Biomedical Research
Stephan Bickel, Feinstein Institutes of Biomedical Research
Intracranial EEG (iEEG) recordings from clinical patients provide an exceedingly rare opportunity to examine creative cognition with millisecond precision and direct access to deep brain structures. We present analyses from intracranial recordings as they watched different types of movies where we examine periods of mind-wandering and insight. In our first set of analyses, we use recordings from N = 13 patients to identify neural signatures of internal versus external attention states during narrative and ambient films. We found significant differences in alpha and high frequency (70-150 Hz) power in the default mode, attention, and visual networks that fluctuate with attentional states, using eye behaviors and self-reports to detect periods of engagement and disengagement. These differences varied depending on movie type, where watching an ambient, dreamscape film produced longer periods of mind-wandering characterized by attenuated sensory processing but greater activity in the DMN, suggesting an active internally-oriented psychological state. Non-linear measures revealed differences in entropy and complexity between internal and external states differed on the stimuli, where the ambient film showed lower entropy in attention networks but more in limbic networks, suggesting a stable attentional state possibly characterized by heightened emotional variability. These non-linear measures can capture rapid changes in momentary attention and inform creativity neuroscience.
Dynamic rewiring of electrophysiological brain networks during creative thinking
Coralie Ramos, MINDIG
Fatima Chhade, University of Rennes, LTSI - U1099, F-35000
Manon Auffret, University of Rennes, LTSI - U1099, F-35000, France Développement Electronique
Mahmoud Hassan, MINDIG, School of Science and Engineering, Reykjavik University
Marc Vérin, University of Rennes, LTSI - U1099, F-35000, B-CLINE, Laboratoire Interdisciplinaire pour l’Innovation et la Recherche en Santé d’Orléans (LI²RSO), Université d’Orléans
Ahmad Mheich, MINDIG, CHUV-Vaudois University Hospital Center, Department of Autism Spectrum and Related Disorders, Lausanne University Hospital
Creativity is a cornerstone of human innovation, shaping both cultural and technological landscapes. Its multifaceted nature, combined with the challenges in neuroscience to understand the functional and evolutionary dynamics of brain networks, has fueled extensive research in creative cognition. This study examines the dynamic rewiring of the brain, at the millisecond resolution of creative thoughts generation, using high-density electroencephalography (HD-EEG) data, from 93 healthy subjects with diverse creative potential, performing an Alternative Uses Task (AUT) and an Object Characteristics Task (OCT). A source-space functional connectivity method was employed to investigate the dynamics of functional brain networks during these tasks. Using a clustering-based approach, we identified five distinct Brain Network States (BNS). Non-creative states (associated with the OCT) were predominantly frontal, whereas creative states (associated with the AUT) engaged broader brain regions, including greater activation of the default mode network (DMN). Creative states exhibited more frequent reconfigurations and uniquely activated the cingulate isthmus, a key DMN node. Additionally, interactions between the DMN, the frontotemporal network (FTN), and the dorsal attention network (DAN) expanded during creative states, characterized by unique reciprocal connections. Our findings reveal that creativity involves higher dynamic rewiring complexity, increased DMN involvement, and enhanced communication between DMN, FTN, and DAN. These results support the theory that creativity extends beyond standard cognition, requiring specialized neural components to produce novel and relevant ideas.
Neural and genetic contributions to creative performance: Exploring Semantic Distance in Word Associations
Örjan de Manzano, Max Planck Institute for Empirical Aesthetics
Vivien Vach, Karolinska Institutet
Kate Schwarz-Roman, Karolinska Institutet
Fredrik Ullén, Max Planck Institute for Empirical Aesthetics
Previous research has shown that when people name unrelated words, their responses have individual differences in semantic distance, which correlate with psychometric creativity. The present study began with an online experiment where twin participants (N = 8165) generated 10 unrelated words and completed a creative achievement questionnaire. A subset of participants (N = 4565) also had data on educational level, cognitive ability, and personality from a prior study. The results convincingly showed that semantic distance between responses predicted creative achievement, even when controlling for the other measures. Following this, we conducted an fMRI experiment with twin- and non-twin individuals (N = 63) performing a similar task. Participants generated 4 words that were either as related or as unrelated as possible. When comparing these two conditions, we found greater activation for the unrelated associations in the occipital lobe bilaterally, particularly in the lingual gyri. This replicates previous work that highlights visual search strategies for word associations. Notably, semantic distances were correlated with individual differences in the activation of this cluster. Further analyses illustrated a condition-dependent increase in functional connectivity for the same contrast between the occipital cluster (seed) and the left posterior superior temporal gyrus (“phonological retrieval”), and the precuneus (“mental imagery and episodic memory”). Brain activity patterns during word generation correlated significantly higher within twin pairs than between twin pairs. The heritability of divergent thinking and creative achievement will be further discussed.
The White Matter of Aha! Moments
Carola Salvi, Department of Psychology and Social Sciences, John Cabot University, Rome, Italy
"Aha!" moments are a crucial aspect of idea generation in creative cognition. While functional neuroimaging studies have identified brain regions involved in insight, the white matter structure underlying individual differences during idea generation remains unexplored. Our study bridges this gap by employing Diffusion Tensor Imaging (DTI) to ivestigate how white matter microstructure supports insight and analytical problem-solving abilities. Participants were administered with Compound Remote Associate (CRA) problems. We examined Fractional Anisotropy (FA) and Mean Diffusivity (MD) across different brain tracts in relation to their preferred idea-generation strategies. Our findings reveal distinct connectivity patterns: insight is associated with lower FA (and higher MD) in the left Posterior Arcuate Fasciculus and bilateral Superior Longitudinal Fasciculi III, suggesting diffuse connectivity that fosters broad semantic activation and cognitive flexibility. In contrast, analytical thinking relies on stronger, more directed connections, reflected in higher FA (and lower MD) in the anterior Corpus Callosum and left Vertical Occipital Fasciculus, supporting focused executive control. These results provides novel evidence for distinct structural connectivity patterns underlying different idea generation approaches, contributing to a more comprehensive understanding of the neural architecture supporting creative cognition.
Men Use Fists, Women Use Whispers?: A Behavioral and Brain Investigation of Gender Differences in Malevolent Creativity
Corinna M. Perchtold-Stefan, University of Graz, Austria
Katharina Sattler, University of Graz, Austria
Christian Rominger, University of Graz, Austria
Andreas Fink, University of Graz, Austria
Malevolent creativity – when creative ideas are intentionally used to damage others – is a novel research field with broad applications for understanding social and affective contexts of creativity in real life. As one ongoing debate, it is still unclear whether the capacity for creative harm differs between women and men, as behavioral studies report diverging findings for self-reported behavior, and even fluency, malevolence, and originality in performance tests. EEG studies on the other hand suggest that women and men may have a similar potential for malevolent creativity, yet utilize different neurocognitive processes to achieve creative harm, with women preferring social-relational aggression, and men preferring direct physical aggression. In a current pre-registered project, we aim to replicate this observed, brain-based gender difference in malevolent creativity strategies by re-analyzing data from five previous behavioral studies (total n ~ 750) regarding gender differences in malevolent creativity strategies of physical harm, financial harm, social (reputational) harm, coercive aggression (threats), and pranks. Second, we implement an fMRI resting-state approach using multivariate pattern recognition and diffusion MRI connectometry (assessing the integrity of white matter fiber tracts), to examine if intrinsic functional and structural connectivity in women’s and men’s brains show different links to aspects of malevolent creativity (n = 131). This ongoing work uniquely bridges research on aggression and creativity to understand and in the future, counteract interindividual differences in the versatile manifestations of aggression in society.
Intra- and inter-individual differences in spectral power volatility are associated with creative task performance
Yuhua Yu, University of Arizona
Yongtaek Oh, Children's Hospital of Philadelphia
John Kounios, Drexel University
Vicky T. Lai, University of Arizona
Mark Beeman, Northwestern University
The temporal dynamics of neural oscillations are important to human cognition. Spectral power volatility, a measure of moment-to-moment fluctuation in EEG spectral power, can reflect the stability of transitions between neural oscillatory synchronization and desynchronization. Previous research has associated trial-to-trial variability in volatility with verbal problem-solving outcomes, with lower volatility predicting solving success and lower response time (Yu et al., 2024). To further evaluate volatility as an index for cognition, we conducted two studies addressing inter- and intra-individual differences. The first study examined EEG data collected from one rest and two task sessions where participants attempted anagrams and compound remote association (CRA). We found that individual volatility was stable across rest and task sessions, supporting volatility as an individual trait for characterizing oscillation dynamics. Importantly, volatility during eyes-closed rest was negatively correlated with the solving rate in both anagrams and CRAs, extending previous findings with intra-individual variability to an inter-individual level. The second study examined trial-to-trial variability of volatility in an idea generation task when people generated metaphors for science concepts. Again, lower volatility was associated with higher response quality measured by metaphor aptness and novelty. Together, lower spectral power volatility is associated with better performance across problem-solving and idea-generation tasks, at both intra- and inter-individual levels. Volatility can complement existing measures and offer valuable insights into cognition-related oscillation dynamics