Connectivity and Creativity in Semantic Neural Networks Nagendra Marupaka and Ali A. Minai, Senior Member, IEEE Abstract— Creativity and insight are distinctive attributes of human cognition, but their neural basis remains poorly understood due to the difficulty of experimental study. As such, computational modeling can play an important role in understanding these phenomena. Some researchers have proposed that creative individuals have a “deeper” organization of knowledge, allowing them to connect remote associates and form novel ideas. It is reasonable to assume that the depth and richness of semantic organization in individual minds is related to the connectivity of neural networks involved in semantic representation. In this paper, we use a simple and plausible neurodynamical model of semantic networks to study how the connectivity structure of these networks relates to the richness of the semantic constructs, or ideas, they can generate. This work is motivated, in part, by research showing that experimentally obtained semantic networks have a specific connectivity pattern that is both small-world and scale- free. We show that neural semantic networks reflecting this structure have richer semantic dynamics than those with other connectivity structures. Though simple, this model may provide insight into the important issue of how the physical structure of the brain determines one of the most profound features of the human mind – its capacity for creative thought. I. I NTRODUCTION Creativity and insight are often seen as arising from the ability to connect disparate concepts or ideas, i.e, to make unexpected connections in semantic space [1], [2], [3], [4], [5], [6]. The experience of unusual combinations leading to creativity has been documented by several scientists. For example, Einstein is quoted as saying, “Taken from a psychological viewpoint ... combinatory play seems to be the essential feature in productive thought – before there is any connection with logical construction in words or other kinds of signs which can be communicated to others” (quoted in [2], [4]). Similarly, Poincare, described his own creative thinking process as follows: “Ideas arose in crowds; I felt them collide until pairs interlocked, so to speak, making a stable combination. ... the only combinations that have a chance of forming are those where at least one of the elements is one of those atoms freely chosen by our will. Now, it is evidently among these that is found what I called the good combination. ... among the great numbers of combinations blindly formed almost all are without interest and without utility” [7] (quoted in [4]). The latter quote Nagendra Marupaka is with the School of Electronic and Computer Systems, University of Cincinnati, Cincinnati OH 45221, Email: maru- pana@mail.uc.edu. Ali Minai is with the School of Electronic and Com- puter Systems, University of Cincinnati, Cincinnati OH 45221, Email: Ali.Minai@uc.edu Acknowledgement: This work was supported in part by a National Science Foundation Human and Social Dynamics Program grant to Ali Minai (BCS-0728413), which includes support from the Deputy Director of National Intelligence, and by a National Science Foundation CreativeIT grant to Ali Minai (IIS-0855714) points to both the utility and dangers of unusual combinations – most combinations generated by the creative mind are not useful, or even nonsensical, but this is the price for discovering the rare combination of uncommon value. Since conceptual combinations arise in the minds of indi- viduals, they must be the product of the cognitive dynamics of these individual minds. The key idea is that creativity depends on how associations between concepts are organized in the mind. Presumably, individuals whose minds link concepts that most people do not are likelier to find them when needed. This idea was formalized by Mednick [2] as an associative hierarchy, which relates the uniqueness (or “unusualness”) of associations in the individual’s mind with the strength of these associations. He argued that non-creative individuals (or individuals thinking in contexts where they are non-creative) have a steep associative hierarchy, where common associations are very strong, but uncommon ones are much weaker. Such an individual is likely to think mainly in terms of common associations and generate conventional, non-creative ideas. The creative individual, in contrast, has a flatter association hierarchy where even unusual concepts are fairly strong, and therefore likelier to be discovered during associative search. Another important factor in creativity is the problem of fixation – the inability to break out of conventional thinking or standard practices. This often occurs in individuals with high expertise in narrow domains because the associations underlying their standard methods are much stronger than those that would lead to non-standard ones. Faced with a particular task, such individuals are likelier to keep return- ing to the same standard ideas. One interesting aspect of Mednick’s hypothesis is that creative individuals may be less likely to generate conventional combinations than non- creative ones, thus avoiding fixation. It has been argued that individuals with broad but relatively shallow expertise may be more creative than those with deep and narrow expertise [6]. Indeed, experiments have shown that the inclusion of a few unconventional thinkers can enhance the creativity of a whole group even if these thinkers are not especially knowledgeable [8], [9]. Given the significance of how conceptual associations are organized, it is important to ask three questions: 1) What is the actual organization of associations is in the minds of individuals? 2) How do these associations generate a “train of thought” that may lead to creativity? 3) Are some types of associative organization more con- genial to creative thinking than others? The focus of this paper is on the last question.