An empirical investigation of the concept of memes in music using mass data analysis of string quartets

Dawkins introduced the concept of the meme as the cultural equivalent to the gene (1989, pp. 189-201). To illustrate the concept, Dawkins cited ‘tunes, ideas, catch-phrases, clothes, fashions, ways of making pots or of building arches’ (1989, p. 192) as examples of memes. All of Dawkins’ examples are elements of culture that have evolved over time. Because music is a part of culture, then under Dawkins’ hypothesis, memes should exist in music. After all, the first of Dawkins’ examples was a ‘tune’.

Jan expanded on Dawkins’ ideas with a thorough investigation into memes in music (2007). This was done on a number of different levels within music, from melodic lines to overall structure, using a range of examples within music. Whilst providing a strong case for memes, Jan was not able to provide evidence from an analysis encompassing a large dataset of music. However, Jan does provide a number of possible methodologies for analysing memes in music, including investigating memes across time periods using single lines of notes (2007, p. 211). The present research expands on Jan’s suggested methodology by looking at short monophonic three- to eleven-note patterns in music across five different non-traditional musicological time periods within a large dataset of string quartets.

A search for memes in music is conducted using a range of scores. These are converted to MusicXML documents, which are then imported into a relational database. Data mining is then implemented on the resultant dataset to produce a series of ranking positions for monophonic note patterns within the music based upon the relative frequencies of their appearances within specified time periods. Additionally, a similarity algorithm is used to investigate the possible ancestral relationships between different monophonic note patterns. Within the limitations of the working definitions and assumptions made in the research, it was shown that there is evidence for the evolutionary properties of selection, replication and variation, and the replicator properties of longevity, fecundity and copying fidelity for some monophonic note patterns within the dataset.