A recent study published in Nature Communications explores the dynamics of higher-order novelties, shedding light on how humans combine existing elements to generate new ideas, technologies, and discoveries. Led by Prof. Vito Latora from Queen Mary University of London, the research investigates both first-order novelties—the discovery of single elements like a new song or artist—and higher-order novelties, which emerge when multiple elements are combined in novel ways.
Traditional studies have focused on first-time occurrences of individual elements, but this research emphasizes the importance of combinatorial creativity. For example, rather than just discovering a new word, writers often form fresh word associations. Similarly, musicians can combine existing notes to create unique compositions. To model this phenomenon, the research team developed the Edge-Reinforced Random Walk with Triggering (ERRWT) framework, which simulates the process of making novel connections between elements.
ERRWT builds upon the random walk model, which describes how systems evolve through a sequence of steps. In this case, the model treats all potential discoveries as nodes in a network, with links representing relationships between them. As the “walker” moves through the network, it not only traverses existing paths but also creates new connections, reflecting how real-world discoveries expand the landscape of possibilities. The strengthening of frequently used connections, known as edge reinforcement, and the creation of new ones, known as edge triggering, both contribute to the emergence of higher-order novelties.
To validate their model, the researchers analyzed data from three domains: music listening patterns (Last.fm), literary texts (Project Gutenberg), and scientific articles (Semantic Scholar). They found that users on Last.fm followed distinct pathways in their discovery of new songs, even when their overall rates of discovery were similar. In literature, authors were more likely to create new word associations rather than introduce completely new words. Meanwhile, scientific papers, particularly their titles, demonstrated a high frequency of novel word combinations, highlighting the role of combinatorial innovation in academic research.
The study’s findings align with Heaps’ law, a mathematical principle describing how new combinations emerge over time. The researchers observed that different processes can have similar rates of discovering individual elements but vastly different rates of generating new combinations. This suggests that the structure of a network and the way it is explored both influence the rate of innovation.
Beyond theoretical insights, this research has significant implications for understanding creativity, scientific progress, and cultural evolution. The ERRWT framework provides a quantitative basis for studying how new ideas take shape and how some innovations succeed while others fade. It also offers potential applications in education, policy-making, and technological development by helping to design systems that foster creative breakthroughs.
Looking ahead, the research team aims to refine the model further, incorporating social influences into the discovery process. As Prof. Latora emphasized, understanding the mechanisms behind creativity and novelties is essential for fostering sustainable innovation and societal progress. By studying how new ideas emerge and how discoveries trigger further innovations, this work paves the way for a deeper comprehension of human creativity and its far-reaching impact.