A New Approach to the Classification of Sound-Producing Instruments

The systematic classification of musical instruments requires a fresh approach. Most existing systems classify museum artifacts, employing a rigid set of distinctions inappropriate for systematic integration of worldwide instrumentalia and human behavior included in producing sound. Central to our approach is a reconsideration of the object to be classified. The domain of our classification includes the dynamic interaction between instruments and human behavior, developed in a manner which facilitates cross-cultural comparisons. The Hornbostel-Sachs Systematik is the most widely accepted classification scheme today. While aspects of the Systematik have often been criticized, most other classification systems have simply extended, refined, or modified the Hornbostel-Sachs scheme. Few have even attempted to reconsider its underlying logic: a hierarchical taxonomy. We propose an alternative approach to instrument classification based on the principle of grouping instruments according to shared observable characteristics. It is nonhierarchical and consistent, and reveals underlying similarities among seemingly unrelated instrument groups. It is also capable of incorporating categories employed in existing systems of classification. We employ a technique of data analysis-Multidimensional Scalogram Analysis (MSA)-to graphically represent relationships and affinities among instruments. Each instrument is graphically represented as a point in space and located within a configuration of points in such a way as to reflect the shared characteristics: the overall configuration is immediately interpretable by scholar and layperson alike. This paper contains preliminary results of research carried out in 1982. The first section of the paper will briefly review the major existing systems of instrument classification and sketch the principles and innovations of our alternative approach. In constructing this approach, we have borrowed from the existing classification schemes. Moreover, we propose not simply to add to, or refine, the Systematik or any other system, but rather to reassess the intellectual problem of instrument classification. The second section of this paper will illustrate the application of our approach through the 213