On the dual nature of creativity: Same same but different?

Abstract

The creativity literature is replete with dualistic constructs, suggesting shared mechanisms but also tempting overinterpretation of their interrelations. An explicit list of relevant concept associations indicates substantial commonality, yet also exposes certain inconsistencies. Dual-process accounts (A and B is relevant) hold promise in resolving discrepancies to the extent that we understand the relative contributions and conditions of A and B.

Creativity research has substantially progressed over the last few decades by amassing findings of how creativity is related to other psychological constructs. The target article by Ivancovsky et al. adds to this work by reviewing relevant literature and proposing novelty seeking as a shared mechanism underlying creativity and curiosity. This claim is substantiated with an extensive review suggesting similar correlates of both constructs across diverse literatures. Reading the target article, it occurred to me that what the authors have done is to perform a “mental factor analysis” on the correlates of creativity and curiosity to extract the factor explaining most shared variance. Identifying underlying factors and mechanisms that describe observations in a more parsimonious way is a central principle of science; however, this endeavor can run risk of oversimplifying matters by being selective in what evidence is considered or being overinclusive in its interpretation (Benedek & Jauk, 2014). Hence, the question arises whether dualistic constructs, which are very popular in the creativity literature, refer to the same underlying factors, or not.

The extensive synthesis of the target article inspired me to explore what dual constructs are commonly associated in the creativity literature. Table 1 lists constructs distinguishing complemental or opposing concepts mentioned by this review as well as by other works (Baas, De Dreu, & Nijstad, 2008; Boot, Baas, van Gaal, Cools, & De Dreu, 2017; Campbell, 1960; De Dreu, Baas, & Nijstad, 2008; Kounios & Beeman, 2014; Neuberg & Newsom, 1993; Smallwood & Schooler, 2015; Troyer, Moscovitch, & Winocur, 1997; Zuckerman, 1984). Concepts listed in column A are often considered related whereas constructs listed in B are commonly considered distinct or opposing to those in A. The table further organizes constructs by overarching themes such as creative thinking, its neuroscientific basis, neophile traits and states (Griffin, 2016), motivation and moods, cognitive processes including (memory, attention, and cognitive control), and psychopathology. Aligning constructs in this way is helpful to uncover underlying, implicit assumptions and to eventually identify new links but also potential inconsistencies. It struck me that we tend to use different, sometimes vague labels for largely the same thing (e.g., defocused, broad, global attention), but then again, we are also very generous in presuming tight relationships across different constructs, together known as the jingle-jangle fallacy (Kelley, 1927). Importantly, although positive associations have been assumed between many constructs in A, it does not seem to work well for them all.

Table 1. Common dualistic constructs (distinguishing complemental or opposing concepts A and B) in creativity research: Are they (all) related?

Note: Concepts in italics were not considered in this context in the target article.

For instance, the target article as well as other works assume that idea generation versus evaluation are related to activation of the default mode network (DMN) and executive control network (ECN), and spontaneous versus controlled forms of thought, respectively. However, on the one hand, recent work suggests that episodic simulations associated with DMN structures also contribute to idea evaluation (Benedek, Beaty, Schacter, & Kenett, 2023; Ren et al., 2020). On the other, there is abundant evidence linking idea generation to executive control (Benedek & Jauk, 2019), as it involves core- and metacognitive control processes associated with overcoming prepotent responses and cognitive fixation as well as with finding and implementing goal-directed strategies (Lebuda & Benedek, 2023; Smith & Blankenship, 1991); in contrast, the empirical evidence on the association between idea evaluation and cognitive control is still relatively scarce (but, see, Benedek et al., 2016; Karwowski, Czerwonka, & Kaufman, 2020). Similarly, exploration versus exploitation of cognitive spaces corresponds to switching versus clustering tendencies (Troyer et al., 1997), with the former being related to higher cognitive control (Ovando-Tellez et al., 2022; Rubinstein, Meyer, & Evans, 2001) – again opposing a simple exploration–spontaneity relationship.

It seems that Ivancovsky et al. also came to the conclusion that cognitive control does not fit well with the other constructs, which motivated them to eventually propose a two-dimensional model with exploration/exploitation and inhibition/disinhibition (akin to cognitive control) as independent factors. This makes sense in my view, although it was not clear why convergent thinking ended up on the exploration side. Another important move was to acknowledge that creativity and curiosity are strongly associated with A, but actually rely on both A and B. This view is consistent with long-standing dual-process models that emphasized the interplay of exploration and exploitation modes (Hart et al., 2018; Hills, Jones, & Todd, 2012) also known as flexibility versus persistence pathways to creativity (De Dreu et al., 2008; Nijstad, De Dreu, Rietzschel, & Baas, 2010), as well as arguing for the relevance of an interplay between spontaneous and controlled processes (Benedek & Jauk, 2018; Sowden, Pringle, & Gabora, 2015; Zabelina & Robinson, 2010). Future research is challenged to study the very nature and conditions of these interplays in more detail to avoid oversimplistic conclusions such as that everything is important for creativity. Another question refers to the actual independency of exploration/exploitation with the level of cognitive control, which appears supported by recent work showing that clustering and switching can occur fast and slow (Ovando-Tellez et al., 2023). The list in Table 1 may prove useful in reconsidering our theorizing on these dualisms toward enhancing our understanding of the nomological network of creativity (Kenett et al., 2020).