A cognitive framework for the categorisation of auditory objects in urban soundscapes

Woodcock, JS; Davies, WJ; Cox, TJ

Abstract

Categorisation is a fundamental cognitive process that plays a central role in
everyday behaviour and action. Whereas previous studies have investigated the
categorisation of isolated everyday sounds, this paper presents an experiment to
investigate the cognitive categorisation of everyday sounds within their original
context. A group of eighteen expert and non-expert listeners took part in a free
sorting task using 110 sounds identified within ambisonic reproductions of urban
soundscapes. The participants were asked to sort the objects into groups of
sounds that served a similar purpose in the overall perception of the soundscape.
Following this, the participants were asked to provide descriptive labels for the
groups they had formed. The results were analysed using hierarchical agglomerative
clustering and non-metric multidimensional scaling (MDS) to explore
both the structure and dimensionality of the data. The resulting hierarchical
clustering of objects show three top level categories relating to transient sounds,
continuous sounds, and speech and vocalisations. Sub-categories were identified
in each of the top level categories which included harmonic and non-harmonic
continuous sounds, clear speech, unintelligible speech, vocalisations, transient
sounds that indicate actions, and non-salient transient sounds. The first two
dimensions revealed by the MDS analysis relate to temporal extent and intelligibility
respectively. Interpretation of the third dimension is less clear, but may be related to harmonic content.

• Journal Article Type: Article
• Acceptance Date: Jan 24, 2017
• Online Publication Date: Feb 10, 2017
• Publication Date: Jun 1, 2017
• Deposit Date: Jan 30, 2017
• Publicly Available Date: Feb 10, 2017
• Journal: Applied Acoustics
• Print ISSN: 0003-682X
• Publisher: Elsevier
• Volume: 121
• Pages: 56-64
• DOI: https://doi.org/10.1016/j.apacoust.2017.01.027
• Publisher URL: http://dx.doi.org/10.1016/j.apacoust.2017.01.027
• Related Public URLs: https://www.journals.elsevier.com/applied-acoustics/