Research reference
Silas, S., Robinson, M., Baker, D., Harrison, P. M. C., Jacoby, N., & Müllensiefen, D (2026). The recall and recognition of sonic logos. Psychology of Aesthetics, Creativity, and the Arts (In Review).
Joint score definition: The joint score is calculated as the sum of the rescaled Recognisability and Recallability scores, where each component is first linearly transformed to a 0–1 range.
For legal reasons we are not able to display the original collection of sonic logos that served as stimuli in the research paper.
Instead, this app uses melodic fragments from the MeloSol corpus (Baker, 2021), derived from Berkowitz et al.’s sight-singing materials (Berkowitz et al., 2017), to demonstrate the statistical relationship between recallability and recognisability.
To approximate the statistical properties of the original sonic logo corpus, we selected 100 melodic fragments by minimising Mahalanobis distance to the centroid of the original corpus in a five-dimensional feature space (four melodic PCA components and melody length).
Surface comparability:
Mean melody length = 6.35 (Berkowitz selection) vs. 6.19 (sonic logos)
PCA component means (sonic logos centred at 0):
Structural comparability: Mantel r = 0.97 between feature correlation matrices, indicating near-identical internal feature structure.
This reflects the stability of melodic grammar constraints across corpora, even when surface statistics differ.
Corpus references
Baker, D. (2021). MeloSol Corpus. Empirical Musicology Review, 16, 106–113. https://doi.org/10.18061/emr.v16i1.7645
Berkowitz, S., Fontrier, G., Goldstein, P., & Smaldone, E. (2017). A new approach to sight singing (6th ed.). W. W. Norton & Company.