Home
Department of Economics
RESEARCH SEMINAR

Caterina Mauri: Originality, influence and success: A model of style

Main content

Welcome to seminar with PhD Candidate Caterina Mauri, University of Southern Denmark.

Title: Originality, influence and success: A model of style

Abstract: Human creativity is central to all innovation but remains poorly understood. In economics, many central questions around growth, regulation and even inequality touch upon creativity. In this project, we contribute to the study of this elusive concept by providing and applying rigorous measures of two concepts that relate to creative production. We apply our framework to study creative production to classical music. 

These measures, built on a general model of creative style and computed on the basis of data on the content of creative works, allow us to study two key aspects of creative production. First, creative production involves novelty. For any work to be considered creative, it must deviate, at least, from that which was previously known to its authors. We capture this divergence as originality. Second, it nearly always involves the recombination of other creative work. This aspect may seem to be in conflict with the first, but is in fact central to the value of creative production. It is the transmission of creative output through recombination that makes progress possible. We capture the extent to which creative output is taken up in later work as influence.

Using extensive data on the content of musical compositions as well as composers' popular success and biographical information, we apply these concepts to classical music composed since the 15th century. The core of our data consists of information about themes, i.e. short recognizable sequences of notes, in classical compositions. The standardized nature of classical music is useful in the construction of the feature space and thus the characterization of style. We then explore, using regression analysis, the relationship between originality, influence and success of composers, as measured by e.g. their popularity on Spotify.

We find that more original composers tend to be more influential upon the work of their later peers and to be more successful with present-day audiences. These findings are robust to the inclusion of a range of controls. A positive association between originality and influence also holds across works by a given composer.

The paper is written with Karol Jan Borowiecki, University of Southern Denmark.