Corpus-based rhythmic pattern analysis of ragtime syncopation. Koops, H. V., Volk, A., & Bas de Haas, W. In Proceedings of the 16th International Society for Music Information Retrieval Conference, ISMIR 2015, pages 483–489, Malaga, Spain, 2015.
abstract   bibtex   
This paper presents a corpus-based study on rhythmic patterns in the RAG-collection of approximately 11.000 symbolically encoded ragtime pieces. While characteristic musical features that define ragtime as a genre have been debated since its inception, musicologists argue that specific syncopation patterns are most typical for this genre. Therefore, we investigate the use of syncopation patterns in the RAG-collection from its beginnings until the present time in this paper. Using computational methods, this paper provides an overview on the use of rhythmical patterns of the ragtime genre, thereby offering valuable new insights that complement musicological hypotheses about this genre. Specifically, we measure the amount of syncopation for each bar using Longuet-Higgins and Lee's model of syncopation, determine the most frequent rhythmic patterns, and discuss the role of a specific short-long-short syncopation pattern that musicologists argue is characteristic for ragtime. A comparison between the ragtime (pre-1920) and modern (post-1920) era shows that the two eras differ in syncopation pattern use. Onset density and amount of syncopation increase after 1920. Moreover, our study confirms the musicological hypothesis on the important role of the short-long-short syncopation pattern in ragtime. These findings are pivotal in developing ragtime genre-specific features.
@InProceedings{    koops.ea2015-corpus-based,
    author       = {Koops, Hendrik Vincent and Volk, Anja and {Bas de Haas},
                   W.},
    year         = {2015},
    title        = {Corpus-based rhythmic pattern analysis of ragtime
                   syncopation},
    abstract     = {This paper presents a corpus-based study on rhythmic
                   patterns in the RAG-collection of approximately 11.000
                   symbolically encoded ragtime pieces. While characteristic
                   musical features that define ragtime as a genre have been
                   debated since its inception, musicologists argue that
                   specific syncopation patterns are most typical for this
                   genre. Therefore, we investigate the use of syncopation
                   patterns in the RAG-collection from its beginnings until
                   the present time in this paper. Using computational
                   methods, this paper provides an overview on the use of
                   rhythmical patterns of the ragtime genre, thereby offering
                   valuable new insights that complement musicological
                   hypotheses about this genre. Specifically, we measure the
                   amount of syncopation for each bar using Longuet-Higgins
                   and Lee's model of syncopation, determine the most
                   frequent rhythmic patterns, and discuss the role of a
                   specific short-long-short syncopation pattern that
                   musicologists argue is characteristic for ragtime. A
                   comparison between the ragtime (pre-1920) and modern
                   (post-1920) era shows that the two eras differ in
                   syncopation pattern use. Onset density and amount of
                   syncopation increase after 1920. Moreover, our study
                   confirms the musicological hypothesis on the important
                   role of the short-long-short syncopation pattern in
                   ragtime. These findings are pivotal in developing ragtime
                   genre-specific features.},
    address      = {Malaga, Spain},
    booktitle    = {Proceedings of the 16th International Society for Music
                   Information Retrieval Conference, ISMIR 2015},
    isbn         = {9788460688532},
    keywords     = {music information retrieval},
    mendeley-tags= {music information retrieval},
    pages        = {483--489}
}

Downloads: 0