Statistical Properties of Musical Creativity: Roles of Hierarchy and Uncertainty in Statistical Learning

Abstract

Creativity is part of human nature and is commonly understood as a phenomenon whereby something original and worthwhile is formed. Owing to this ability, humans can produce innovative information that often facilitates growth in our society. Creativity also contributes to esthetic and artistic productions, such as music and art. However, the mechanism by which creativity emerges in the brain remains debatable. Recently, a growing body of evidence has suggested that statistical learning contributes to creativity. Statistical learning is an innate and implicit function of the human brain and is considered essential for brain development. Through statistical learning, humans can produce and comprehend structured information, such as music. It is thought that creativity is linked to acquired knowledge, but so-called "eureka" moments often occur unexpectedly under subconscious conditions, without the intention to use the acquired knowledge. Given that a creative moment is intrinsically implicit, we postulate that some types of creativity can be linked to implicit statistical knowledge in the brain. This article reviews neural and computational studies on how creativity emerges within the framework of statistical learning in the brain (i.e., statistical creativity). Here, we propose a hierarchical model of statistical learning: statistically chunking into a unit (hereafter and shallow statistical learning) and combining several units (hereafter and deep statistical learning). We suggest that deep statistical learning contributes dominantly to statistical creativity in music. Furthermore, the temporal dynamics of perceptual uncertainty can be another potential causal factor in statistical creativity. Considering that statistical learning is fundamental to brain development, we also discuss how typical versus atypical brain development modulates hierarchical statistical learning and statistical creativity. We believe that this review will shed light on the key roles of statistical learning in musical creativity and facilitate further investigation of how creativity emerges in the brain.

Keywords: abstraction; autism spectrum disorder; creativity; development; hierarchy; integration; prediction; statistical learning.

FIGURE 1

Statistical creativity in musical improvisation. Misty by Errol Garner, composed in 1954. (A) The arrangement, chord names, and symbols are simplified (just major/minor, flat, and 7th note) to account for the two-five-one (II–V⁽⁷⁾–I) progression. The statistical characteristics of jazz improvisation played by Bill Evans, Herbie Hancock, and McCoy Tyner. (B) Adapted from a figure of a previous article (Daikoku, 2018b). The component loading of principal component analyses showed that statistically coherent units have general characteristics shared among the three improvisors, whereas large-scale statistical units provide individualities unique to each improvisor. This suggests that abstraction (i.e., statistical learning within words) may fundamentally provide general knowledge, while integration (i.e., deep statistical learning between words) contributes to musical creativity and individuality, as well as common knowledge.

FIGURE 2

Statistical creativity in the uncertainty fluctuation of musical composition. Figure adapted from a previous article (Daikoku, 2019b). From the early to the late periods of Beethoven’s lifetime, the predictable patterns that ubiquitously appear in all of his piano sonatas (familiar sequence) were decreased, whereas the uncertainties were gradually increased. Further, these findings were more prominent in higher- (deeper), rather than lower-order statistical learning models (right). This may suggest that higher-order statistical learning reflects novelty-seeking (creative) behavior over a composer’s lifetime.

FIGURE 3

A hypothesis of statistical creativity. Statistical creativity may, at least, be achieved via two potential mechanisms in a hierarchical statistical learning. The first is the interplay between the chunking of statistically coherent events into a unit and integration of the several units. This process forms a hierarchical structure in statistical learning (i.e., hierarchical statistical learning). The second is a perceptual uncertainty as shown in each of the bell-shaped distribution in the figure. The brain appears to seek a suboptimal solution of uncertainty for creativity based on prior distribution in the internal predictive model. It is assumed that a perceptual uncertainty at not very small- and large-scale statistical learning may induce statistical creativity.