City Research Online

The design of note-based algorithmic systems through the use of mental models

Dos Santos Agostinho, G. (2021). The design of note-based algorithmic systems through the use of mental models. (Unpublished Doctoral thesis, City, University of London)


This practice-based research, which consists of this dissertation, a portfolio of compositions, and a programming library for Python, explores the role played by metaphors and mental models in providing a framework for algorithmic composers to operate within. When working with algorithms, the composer approaches the act of music creation through an algorithmic lens which, in turn, can often suggest specific compositional ideas. To put it simply, in order to compose with algorithms, the composer must think algorithmically, which, in turn, affects how they conceive their musical ideas in the first place. We are shaped by our tools.

This research argues that this algorithmic musical thinking is often realised with the aid of metaphors and mental models. These cognitive mechanisms are crucially important not only for algorithmic conceptualisation but also for their potential to suggest specific ways of organising and manipulating algorithmic ideas, directly influencing the final artwork. As such, this work presents a novel way of approaching algorithmic composition, one in which the composer is consciously designing specific (and possibly idiosyncratic) mental models. From this perspective, composition becomes an investigation of the potentials of these mental models, aligning itself with the notion that algorithmic music can serve as a form of exploration to reach musical results not entirely planned a priori.

This approach is utilised throughout the accompanying portfolio of compositions, which is also analysed in this dissertation. The mental models employed in these compositions—particularly those involving musical repetition—are connected to a specific set of aesthetic concepts that underpins this research. The accompanying programming library, Auxjad, provides classes and functions written in Python that directly implement these specific mental models. This library is publicly available online under a permissive free software license, allowing other composers to adapt and incorporate its code into their own practices.

Publication Type: Thesis (Doctoral)
Subjects: M Music and Books on Music > M Music
M Music and Books on Music > MT Musical instruction and study
Departments: Doctoral Theses
Doctoral Theses > School of Arts and Social Sciences Doctoral Theses
School of Communication & Creativity > Performing Arts > Music
Text - Accepted Version
Download (4MB) | Preview
[img] Archive - Accepted Version
Download (3MB)



Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login