Ryan Seamons and Dr. Jerry L. Laccard, Music Education

Introduction

In the last century, a new discipline developed around the scientific analysis of musical microstructures found in authentic or “deep-layer” folksong. Ilmari Krohn of Finland was the first to publish in this field (Suomen kansan sävelmiä—Volumes I–IV, 1904–1928). This publication came into the hands of two young Hungarian university students who later became world-class composers, Béla Bartók and Zoltán Kodály, who had already been collecting peasant tunes in remote rural villages. Krohn’s analysis and classification paradigm was just what Bartók and Kodály needed to make sense of the rich musical structures they were finding in the ancient layers of the Hungarian oral tradition.

The fieldwork of Bartók and Kodály eventually grew into a large division of the Hungarian Academy of Sciences known as the Institute for Musicology and which now employs dozens of musicologists, ethnomusicologists, composers and technicians and contains an academic press. In 1946, Kodály was elected president of the Hungarian Academy of Sciences.

Through analyzing and classifying over two hundred thousand ancient Hungarian folk tunes and variants, Kodály and Bartók and those who have succeeded them have created and continue to create extensive catalogs of tune-types that have provided a variety of viable structural and procedural models for composers, musicologists and pedagogues. The folksong centered musical education devised by Zoltán Kodály has now spread to 35 countries on 5 continents and continues to expand and deepen. Our project centered around the idea that as Kodály’s methods continue to spread, technology can hasten and coordinate the work being done by composers, musicologists, and pedagogues around the world. By creating an online database framework, we could increase diffusion of methods and reduce duplication of work, providing an immense technological leap forward in this kind of folksong research.

Project to Date

We have created a web-based research input prototype. Using php, mysql, javascript, and html, a solid framework has been constructed that enables further development of annotation and collaboration tools that will take folksong research from a pen, paper, and snail-mail project and empower it by providing simple tools to view, annotate, search, and archive folk songs online, available world-wide.

Our database has the ability to add, view, annotate, and search for folk songs. While the actual work of annotation is virtually the same, with an online audience, the MELOS-V Database allows researchers worldwide to submit tune data and participate in comparative decision-making about its nature, value and applicability.
We set out to create a prototype of a database that would:
1. assist in all aspects of current folk song research, especially transcription and pattern study,
2. provide a platform for collaboration of folk song transcriptions and mp3s among experts,
3. and be easy to use, ensuring rapid acceptance and use among possible users.
The prototype created and soon to be housed at melosv.byu.edu is a groundbreaking step toward these objectives.

Impact

The impact of this project is the merging of two formally separate fields of study. Just like family history research has been fundamentally enhanced by the use of the internet to find users who can help annotate records at speeds unimaginable before, so this database, when fully complete, will do the same for folksong music and teaching methods.

Future Research and Production

A number of future projects have been discovered through the creation of this database prototype.
The obvious next project is to take the framework we have built, and change it into a fully-functional, aesthetically-pleasing online tool, with ability for users in the international forum to have access to the resources and ability to add folk song music from their own repositories. The issue with this is not only development into the final product, but decisions and resources about who will maintain the database and how it will be managed. Preferably a solution which utilizes the users for most of the administrative functions would be preferable.

The other fascinating future project is the incorporation of neural networking into the database. Neural networking is the use of a computer program to find patterns in known data that can then be used to predict the outcomes of unknown data. Simply put, a neural network trained using the annotation protocols we intend to employ would greatly reduce the objective aspect of human annotation tasks so that scholars can focus on the more subjective issues of developing classification categories. Users would change their function from annotation to data validation – simply looking at what the computer predicted the annotation to be and verifying or correcting the output. This would enable the masses of un-annotated music to be quickly input and stored, providing vast amounts of data for analysis. Again, this is a future project requiring not just funding, but experts who understand both the world of computer programming and comparative folksong musicology.

The work we have done on the MELOS-V Database is just the start of a grand technology revolution for the world of scholarly folksong analysis and classification.