The Structure of an Oral Tradition ? Mapping of Hungarian Folk Music to a Metric Space

Abstract

To find the essential musical relations characterising 2323 typical Hungarian folk melodies as a musical system, we developed a method relating melodies to points of a multidimensional Euclidean melody space. The study of the resulting point systems is based on the method of principal component analysis, and shows the inherent relations of a rich oral musical tradition as a very regular, unbroken clustered structure. The majority of the points are grouped in parallel as well as perpendicular clusters. The musical interpretation of parallelism and orthogonality highlights certain basic construction and variation principles in Hungarian folk music.

Generation of Folk Song Melodies using Bayes Transforms

“…In the approach of Juhász (Juhász, 2002;Juhász, 2004;Juhász, 2009), maplike representations are derived for 'motive contours'. Uncertainties calculated in terms of locations in the map then become the means of predicting boundaries.…”

A systematic comparison of different European folk music traditions using self-organizing maps

“…Therefore, we represented the original melody contours and those calculated by the self-organizing map as ''types'' in a multidimensional ''melody space''. We have shown in a previous work that the spatial characteristic of the point system representing melody contours in the above-mentioned space can be well interpreted from a musical point of view (Juha´sz, 2002).…”

“…We have shown in a previous work that 32-dimensional point systems representing melody sections of Hungarian folk songs can be approximated in a lowdimensional subspace, and a very clear musical meaning can be attributed to the orthogonal basis vectors determining this subspace (Juha´sz, 2002). (The basis is constructed by certain principal vector components of the main clusters that are shown in Figure 7.…”

“…The point system is built up by large clusters situated along parallel and orthogonal ''planes'', which are perpendicular to the plane of the map. Typical section contours are also represented in the cuttings of the ''planes'', showing that (1) vertical planes consist of sections beginning at the tonic, fifth and octave, (2) horizontal planes consist of melodies ending at the tonic and fifth, (3) melodies along the lines of intersections of the planes have definite beginning and ending, and their central part varies according to the third coordinate in the subspace (Juha´sz, 2002).…”

“…A method based on dynamic programming has been successfully applied to characterize the diversity of the contours (Sachira & Milan, 1997). A systematic description of Hungarian folk music in a multidimensional space was also based on the study of the contour, using principal component analysis (Juha´sz, 2002). Principal component analysis has also been applied for timbre characterization (Sandell & Martens, 1995).…”