Scribe Database - Documentation


How I started to think about computer-aided scribe identification and why I think it is helpful for research.

I first got into contact with sources of Bach's music during an internship at the Bach-Archive, Leipzig, Germany, in February 2001. There I developed an overriding passion for music manuscripts and scribe identification. For one month, I was occupied with the musical handwriting of Anna Magdalena Bach, Johann Christoph Altnickol and Friedrich Agricola. It was in Leipzig that I read the articles of Dürr, Kobayashi, and Dadelsen on the subject of scribes and the chronology of manuscripts and works.

While making extensive forays into the field of scribe identification, I found myself faced with the following problems: Under which circumstances is it possible to assign a specific manuscript to one particular scribe? What are the criteria for being able to do so? Are these criteria of an analytical or intuitive nature? And last but not least: Is there a way to define a "scribe fingerprint" through his handwriting characteristics, or is the range of their variation too broad?

During my second internship at the Bach-Archive in August 2001, I worked intensively on the current database project to address these problems and questions. During this investigative process, I became increasingly optimistic that the problems can be considerably reduced, if not solved, by the database application. I also became increasingly aware that the graphic presentation of the handwriting images in the database significantly contributes to a better understanding of a scribe's development, as well as to the overall understanding of the scribe identification process itself.


As shown above, the objectives for the Scribe Database are:

  1. to define a scribe identification system
  2. to enable scholars to search for information on:
  3. to proof already existing scribe assignments

The scribe identification system is the heart of the Scribe Database. It is based on the concept of handwriting categories. These categories are abstract forms of handwriting characteristics taken from "Beiträge zur Chronologie der Werke Johann Sebastian Bachs" by Georg von Dadelsen in "Tübinger Bach-Studien, Ed.:Walter Gerstenberg. Trossingen, 1958." The Scribe Database uses these categories as a filter for its scribe identification system. Every single handwriting characteristic has to be assigned to one category. Thus you are able to query manuscripts for similar handwriting characteristics.

Apart from scribe identification, you can also use the Scribe Database to search for information on scribes, manuscripts, and handwriting characteristics:

Scribe Information

Manuscript Information

Handwriting Characteristics Information


Emphasis will be placed on adding new data andincreasing the ease of use.

After having programmed the database system, which includes tools to insert, query and check data, it is obviously important to add new data. To date, and for the purposes of testing, I have only included information on Johann Friedrich Hering: known manuscripts and some handwriting charateristics. Only categories that were needed for Hering are presently available; nevertheless the tools for adding more categories are already available. Once the data is available, it will then be simple to add more complex querying functions and throughly test them.

For the future, new, more user-friendly tools for category assignment will be implemented. This will take the form of aset of questions, the answers to which allow the computer to automatically assign the correct category.


For those interested in the technical implementation of the scribe database.

The Scribe Database was implemeted on a MySQL system. I used the following versions:

Ver 9.38 Distrib 3.22.32 for Linux

MySQL website:

Database access is handled by PHP. PHP is Pre-Hypertext-Processor

PHP website:

Internet presentation is done by apache web-server.

Apache website:

The Scribe Database was implemented with free software only. To learn more about free software click here:



I would like to thank Dr. Wollny and Dr. Leisinger from the Bach-Archive for their support during the first weeks of this project. Dr. Wollny gave me all the support I needed and mentioned the handwriting categories of Dadelsen.

Thanks to Michael Edwards for his critical comments concerning the database structure. Providing me with a web-server and mysql-php support made this project possible at all.

Francisco Domingues and Andreas Prlic of "Center of Applied Molecular Engineering" at University of Salzburg helped me all the way with their infinite knowledge on database, progamming and webdesign.

Dr. Rainerius Fuchs supported the first testing-version of this project by providing facsimiles of Anna Magdalena Bach which I could scan for testing purposes.

I would like to thank Alexander Grimalovsky from Velikiy Novgorodwho was very kind to formulate an important SQL-command for me. Many thanks!


Sources and Scribes:

For a detailed biography see the Bach-Bibliography.


Matthias Röder
November, 23rd 2001