This project will provide the software infrastructure to support the establishment of a publicly available digital library of medieval music manuscripts.

Software already exists for transcribing modern musical notation and Western square-neume notation. None of this software, however, is appropriate to the dozens of styles of early notation. Much is still unknown about the meanings of some early music symbols, and many styles of notation contain little or no information about pitch and duration of notes. (Indeed, the concept of a 'note' may, itself, be anachronistic.) These are emergent forms of writing and cannot be subsumed under later, more developed types of writing. To do so would be prejudicial to rigorous scholarly study and would distort the public's appreciation of the artefacts on their own terms. No existing software is capable of retaining the native notational characteristics of medieval artefacts, while permitting automated comparison of melodies across notational styles.

Our project is seeking to create, for the first time, the software infrastructure necessary for stable, interoperable resources of neumed transcriptions. We intend the project to be self-sustaining after the infrastructure has been completed, so that scholars around the world will be able to add transcriptions freely to a distributed e-library on the World-Wide Web. This e-library will allow the general public to view manuscripts in symbolic notation (see Figure 1) as a supplement to the digital photographs that others are now putting online, or in conjunction with audio recordings. Since these transcriptions are in a self-describing file format, they can be translated automatically to databases or other, less comprehensive, data formats for scholarly study, comparative analysis, typesetting, etc. There are compelling reasons, however, to have 'canonical' transcriptions in a comprehensive data format so as to avoid duplication of effort and fragmentation of digital libraries.

Figure 1. View generated in a Web browser from transcription data of a neumed Greek text.

The main output from this project will be a group of computer programs to facilitate the production, access, and study of neumed manuscript transcriptions in 'canonical' form. They will be deployable as independent programs or as Internet utilities for the shared, virtual library. The principal parts are as follows:
  1. Improved data-entry and visualisation that builds upon earlier, 'beta-test' software. Transcription is currently labour-intensive; easier, faster methods are needed. Visualisation is used in Web browsing, comparative study of notations, and checking the accuracy of transcriptions; this software needs tighter integration with Web browsers and optional print-quality output for scholarly papers or critical editions.

  2. Parallel markup for users to annotate transcriptions as critical editions, teaching tools, etc., without changing the 'canonical' transcription.

  3. Melodic pattern-matching for finding particular or similar examples among many transcriptions. The criteria for a 'match' will be user-configurable by heuristic rules. The long-term value of this capability is substantial. It can ultimately be used by Web search engines for finding transcription files and it will present many new possibilities for musicological research.


Copyright © 2005 by the University of Oxford.
Contains software or other intellectual property protected by one or more of the following copyrights.

Copyright © 2003-2005 by Louis W.G. Barton.
Copyright © 2002-2003 by The President and Fellows of Harvard College.
Copyright © 1995-2001 by Louis W.G. Barton.
All rights reserved. The copyright holders provide this intellectual property online as reference material for educational, cultural, and charitable purposes. The material is provided "as is" without any warranty whatsoever.