Novel Methods and Perspectives 1 & 2
Thursday May 22, 2:00 - 3:15 PM, ICM Auditorium
An Entropy Modulation Theory of Creative Exploration
Thomas Hills, University of Warwick
Yoed Kenett, Technion - Israel Institute of Technology
Compared to individuals who are rated as less creative, higher creative individuals tend to produce ideas more quickly and with more novelty—what we call faster-and-further phenomenology. This has traditionally been explained either as supporting an associative theory—based on differences in the structure of cognitive representations—or as supporting an executive theory—based on the principle that higher creative individuals utilize cognitive control to navigate their cognitive representations differently. Though extensive research demonstrates evidence of differences in semantic structure, structural explanations are limited in their ability to formally explain faster-and-further phenomenology. At the same time, executive abilities also correlate with creativity, but formal process models explaining how they contribute to faster-and-further phenomenology are lacking. Here, we introduce entropy modulation theory which integrates structure and process-based creativity accounts. Relying on a broad set of evidence, entropy modulation theory assumes that the difference between lower and higher creative individuals lies in the executive modulation of entropy during cognitive search (e.g., memory retrieval). With retrieval targets racing to reach an activation threshold, activation magnitude and variance both independently enhance the entropy of target retrieval and increase retrieval speed, reproducing the faster-and-further phenomenology. Thus, apparent differences in semantic structure can be produced via an entropy modulating retrieval process, which tunes cognitive entropy to mediate cognitive flexibility and the exploration–exploitation trade-off.
Make to Create: Creativity beyond the Brain
Frédéric Vallée-Tourangeau, Kingston University
A cognitivist narrative on the nature of creativity is steeped in exceptionalism: people with exceptional cognitive capacities and skills are simply more adept at solving problems and generating original ideas. It is from this exceptionalist perspective that commentators offer a distinction between so called Big-Creativity of breakthroughs and little-creativity displayed in more pedestrian activities. From this perspective, interventions to enhance creativity aim to improve people’s cognitive abilities. These abilities are measured through standardized tests of divergent and convergent thinking, and the logic of such measurement practices in turn validate—and perform—the exceptionalist agenda. In contrast, a post-cognitivist perspective casts creative problem solving as a situated, enacted and interactive process. Here the emphasis is not so much on ideation before engaging with a problem, or on ways to promote more creative ideation, but rather on the material engagement with the elements that configure a problem. Laboratory methodologies designed to foster this form of interactive engagement are better placed to reveal how novel ideas emerge through a recursive cycle of making and observing. Participants are transformed as they transform physical solution prototypes; we witness a double process of becoming. While the term ‘insight’ is often employed (anecdotally and in more controlled laboratory conditions), a more appropriate term might be ‘outsight’, that is people see in the world the results of their engagement. The positive phenomenology of insight (relief and pleasure) also accompanies moments of outsight.
Creativity, curiosity, and optimal level of arousal
Paula Ibáñez de Aldecoa, Institute for Advanced Study in Toulouse
Emily Burdett, University of Nottingham
Erik Gustafsson, Université de Franche-Comté
Creativity is a key resource that helps individuals manage and adapt to varying levels of cognitive and environmental stimulation. We revisit the “optimal-level of stimulation” (OLS) model, highlighting its relevance in the context of changing brains across different developmental stages, moods, and arousal levels. The OLS model posits that engagement, attention, and curiosity are driven by stimuli that balance simplicity and complexity. However, two critical but often overlooked factors are the interplay between stimuli within their context and the relativity of stimulus complexity to an individual’s cognitive resources. While familiar contexts tend to lower overall stimulation and encourage novelty-seeking, novel contexts can elevate stimulation and foster a preference for familiarity. To maintain an optimal arousal state when overstimulated or under-stimulated, individuals may shift their focus, adopt alternative processing strategies, or engage in creative thought and action. These dynamics are profoundly influenced by factors such as age, mood, arousability, and the broader context. By integrating these variables, the updated OLS model provides a holistic framework for understanding exploratory behaviors, the rewarding nature of curiosity, and the role of creativity in self-regulation. This model enriches existing theories by explaining how changes in cognitive resources and external stimulation shape the creative processes of individuals with developing or aging brains. Ultimately, it sheds light on the adaptive mechanisms underlying creativity and offers insights for interventions addressing learning and cognitive development in diverse populations.
Pareto optimality reveals the building blocks of creative search
Dan Amir, Hebrew University of Jerusalem
Roee Schur, Hebrew University of Jerusalem
Shir Nehamkin, Hebrew University of Jerusalem
Roy Gutglick, Hebrew University of Jerusalem
Yuval Hart, Hebrew University of Jerusalem
Creative search is a multifaceted phenomenon, as it balances inherent trade-offs such as exploring vs. exploiting or avoiding the common vs. being understood. How can we disentangle this complex cognitive process to its underlying drivers? To do so requires to go beyond end products to quantify the entire dynamic search trajectories and to examine creativity in more than one dimension of low-vs-high creativity scores as we risk lumping together contradicting computations. Here, we apply a novel principled approach, Pareto optimality, to decompose creative search to its core building blocks. Pareto optimality theorem posits that systems optimize performance under multiple competing goals by balancing trade-offs within a low-dimensional polytope, spanned by archetypes that optimize each goal individually. Pareto optimality ties individual differences to different balances between competing optimization goals and thus explains the utility of these individual differences. We applied Pareto optimality to data from the creative foraging game (CFG, N=643) which describes the entire search trajectories in many dimensions. We found that the behavioral data lies in a significant polytope with 6 archetypes (p<0.001). The 6 core search strategies (goals) correspond to: 1) commonality detection 2) exploration 3) non-biphasic search 4) efficient search 5) high coverage search and 6) originality detection. Each goal is characterized by behavioral, functional, and structural features. Pareto optimality connects behavioral, neuronal, and computational aspects of creativity under one theoretical framework to assess creativity changes across individuals and within individuals across time.
Varieties of divergent production: A network analysis of 27 divergent thinking tasks
Baptiste Barbot, UCLouvain
Lea Marie Naczenski, University of Amsterdam
Matthijs Baas, University of Amsterdam
Claire E. Stevenson, University of Amsterdam
Grounded in the historical development of Guilford’s Structure of Intellect model (SOI), this study examines the structure of divergent thinking (DT) and its relationship with other cognitive abilities using psychometric network analysis. We reanalyzed data from an unpublished report by Guilford, Merriefield and Cox (1961), encompassing 27 DT tasks (i.e., different operationalizations of DT which varied according to their content and product dimensions) and six cognitive ability tasks administered among 677 adolescents: verbal comprehension, verbal reasoning, word meaning, paragraph meaning, numerical ability, and processing speed. Results showed that (1) the association between different DT tasks across a range of operationalizations varies greatly, with partial correlations ranging from -.21 to .41; (2) DT task performance within a content domain could be closely related to performance on other DT tasks in the same content domain, but not systematically; (3) performance on DT tasks and other cognitive ability tasks are intertwined according to content (as opposed to product dimension); (4) no strong clusters of tasks with related performance could be found in the network; (5) the most central DT task, the one that best represents performance across all operationalization of DT, was the Alternate Use Task; and (6) verbal comprehension and verbal reasoning played a central role in the network. This study initiates a new approach to investigate the nature of DT, and provides new insights on the issue of generalizability of findings in the field of creativity research, as well as the relationship between DT and other cognitive abilities.
Many Routes to Insight, Many Routes away from Impasse
Wendy Ross, London Metropolitan University
George Georgiou, University of Hertfordshire
Thomas C Ormerod, University of Sussex
This presentation will present preliminary data from a large scale study undertaken in collaboration with the Copernicus Science Centre, Warsaw. Data were collected from 62,000 participants as part of an interactive exhibition on creativity. Participants were invited to solve the ten coins problem presented on an interactive computer display. The data presented here focus on a subset of 17,000 individuals over 18 who took part in the experiment between July and December 2023. We will demonstrate how the cognitive and affective states vary between participants according to the paths they select towards solution. In particular, we will focus on the feelings of insight and impasse experienced by participants. The data suggest that both insight and impasse are multidimensional and can be used to predict problem solution and withdrawal.
EPoC - a multifaceted measure of creatiuve potential
Todd Lubart Université Paris Cité
Tools to evaluate the potential for creative thinking are needed in both research and practice to advance the field of creativity. Since 2010, the EPoC tool (Evaluation of potential for creativity) is available. It measures creative potential in terms of both divergent-exploratory and convergent-integrative process components, across several content domains. These domains are verbal, graphic, social, math, science, music and body movement. This presentation wil provide an overview of the measurement approach, with concrete examples of the tasks and the range of responses that children, adolescents and adults give; The psychometic properties will be described with particular attention to the structural model, which has been found to be stable across samples from severql different countries. . The EPoC tool has been used in numerous studies as a pre - post test to see if creativity training has an effect, It has also been used in reserach on the detection of high creative potential, creative giftedness. This work will be reviewed, and results on the relationship between creativity and intelligence, bazsed on EPoC will be highlighted.; Finally, EPoC offers a dynamic assesment mode, which allows both the test-taker's ability level to be assessed, but also the test-taker's capacity to develop creative ability thanks to hints and advice in real time: Details on this dynamic testing mode will be provided. In a concluding part, the EPoC battery will be situated in the larger literature on creativity assessment, based on a recent review of a century of creativity measurement.
Music-Induced Spontaneous Thought Reflects Individual Differences in Creativity
Ben Kubit, Northeastern University
Peter Benson, Northeastern University
Cara Turnbull, Princeton University
Ziyan (Annie) Zhao, Northeastern University
Oscar von Rekowsky, Northeastern University
Bolor Amgalan, Northeastern University
Claire Bauer, Northeastern University
Noelle Carter, Northeastern University
Elizabeth Margulis, Princeton University
Psyche Loui, Northeastern University
Most would agree that creativity is a desirable trait in humans, but it has traditionally been difficult to define and measure, in part, because creativity is linked to a diverse set of mental processes that are considered task-independent and even stimulus-independent, like mind-wandering, daydreaming, and imagining. Music, however, provides a tractable window into these varieties of spontaneous thought. While listening to music, people regularly imagine fictional scenarios that feel highly personal, but are actually reliably prompted by features of the music. Here, we show in two studies that creativity is reliably predicted by music-induced imaginings. We applied natural language processing tools to derive psychometric measures of fluency, originality, and appropriateness out of imagined stories from a highly unconstrained music listening task. These measures predict independent trait-level measures of divergent thinking in a separate task. Our results demonstrate a relationship between music-induced imaginings and creativity that reflects not only a participant’s ability to imagine an elaborate story, but also the specificity of the story’s semantic content relative to other stories imagined by the same participant for other pieces of music. Thus, our task captures people’s ability to constrain creative thoughts based on what is practical and relevant to a problem. Results have implications for understanding mind-wandering, stimulus-independent thought, episodic future simulations / prospective memory, semantic vs. autobiographical thinking, rumination, and problem solving in a highly naturalistic and ecologically valid context.