hide
Free keywords:
Digital humanities, Stylometry, Topic modelling, Lyrics, Distant reading, Clustering
Abstract:
The digital humanities offer many new methods to scholars interested in metal studies. This chapter demonstrates two such methods, that is, stylometric clustering and topic modelling, by interpreting the collected lyrics of over 8,000 metal bands. This allowed the authors to identify general trends that would be hard to derive by personally reading lyrics. This analysis showed several recurring medieval topics in metal lyrics, for example, a Fight for Glory expressed with words like ‘die’, ‘glory’, ‘warriors’, and ‘victory’. The authors were also able to distinguish metal bands with medieval references from those without – a line of research that could help in categorising (new) bands – and track the stylistic development of bands over time. Such statistical methods might help scholars not only by allowing for large-scale studies, yet also by providing inter-subjective feedback for theories.
