2. The Starting Point

The origins of the Bibliography of German Grammar date from the year 1980, when Peter Eisenberg decided to gather information about scientific publications in the fields of German grammar. Starting with a small team of collaborators, located at the "Freie Universität" at Berlin and at the University of Hannover, he systematically tried to include every book and every article dedicated to German grammar that was published since the early sixties. This work was carried out by analysing various specialized publications like linguistic journals or catalogues from publishing houses. In the beginning, the entries were collected using simple index cards. The first print edition of the bibliography was published in 1985 (see [3]), followed by a second edition in 1988 (see [4]). In the late eighties, after the introduction of the personal computer, the index cards were replaced by a database system running under MS-DOS. This made it easier to maintain the increasing amount of bibliographic entries, which was already larger than 5,000 items. But this first database just mirrored the functionality of the index cards without adding any new relational structures. In 1992 the bibliographic work was continued at the University of Potsdam. Some years later, after the publication of the third print edition (see [5]), the project group decided to use the Standard Generalized Markup Language (SGML) for the logical formatting of the bibliographic data. The aim was to make the process of collecting and maintaining the entries fully independent from any underlying computer platform or operating system. Whereas the old, proprietary database system reduced the scope to a single platform, the use of SGML allowed the effective processing of the data on literally every computer system. Another aspect was the handling of language-specific characters (e.g., special characters in the names of east-european authors), for which SGML offers a simple and effective solution.

In 1999 the location and composition of the project group changed again. Since then, the Bibliography of German Grammar - with now more than 10,000 entries - is maintained at the Institute for German Language (Institut für deutsche Sprache) in Mannheim/Germany. The current goal is not only to keep the bibliography permanently up-to-date, but to make the administration more effective and secure. Besides, the plan is to broaden the potential readership by adding new value and offering the contents on the World Wide Web. To achieve this, the first step was to change the way the data is stored and edited. In order to extend the possibilities of SGML by the advantages of a database management system (DBMS), plans were worked up to migrate the data to the object-relational DBMS Oracle™. This decision was based on the following arguments (of course a lot more could be said about the general advantages of database management systems, just think of aspects like scaling, high availability, data analysis, etc.):

  1. Robustness: Aside from the stability of the hardware and the operating system itself, database management systems provide a robust environment for the development of data collections and applications as well as advanced backup and recovery strategies.
  2. Security: A DBMS can control access to every piece of data and prevents the unauthorized modification or deletion of existing database entries. This seems important especially for multi-user environments, where users with different responsibilities work on a complex collection of data (such as a bibliography).
  3. Accurate Data: Database management systems promise data accuracy and the avoidance of redundancy by offering automatic consistency and integrity checks.
  4. Retrieval: Stored in a database, the complex bibliographic data can be easily accessed using the mighty Structured Query Language (SQL). Furthermore the execution of queries can be offered to a global audience by developing a gateway to the World Wide Web (WWW).
  5. Platform independency: Since Oracle™ runs on nearly every computer platform, the migration to another machine or operating system at some future date should be no problem. Besides, it seems easy to generate well-formed SGML out of database tables.
The data used as starting point for the migration was organized in three separate SGML-files: One for the collected articles, one for the books (monographs as well as anthologies) and one for periodicals (journals, series, etc.). To get an impression of the internal structure, just take a look at the following examples.

Sample article entry: 

<aufsatz id="t6830" jahr="1992" bd3="bd3" utyp="11" relevant="ja">
  <autor typ="autor"><nachname>Strecker</nachname><vorname>Bruno</vorname></autor>
  <titel>Zum Begriff des Satzes</titel>
  <verweis idref="t6815" jahr="1992" von="408" bis="416">
  <unstruk>Hoffmann, Ludger</unstruk>
  <schlagwort>Satz, Definition</schlagwort>
</aufsatz>

Sample book entry: 

<buch id="t1657" jahr="1980" bd1="teil-b" typ="mo" utyp="50" relevant="ja">
  <autor typ="autor"><nachname>Heringer</nachname><vorname>Hans</vorname><vorname>Jürgen</vorname></autor>
  <autor typ="autor"><nachname>Wimmer</nachname><vorname>Rainer</vorname></autor>
  <autor typ="autor"><nachname>Strecker</nachname><vorname>Bruno</vorname></autor>
  <titel>Syntax. Fragen - Lösungen - Alternativen</titel>
  <pubdata>
    <ort>München</ort>
    <verlag>Fink</verlag>
    <reihe ref="b208" band="251">UTB</reihe>
  </pubdata>
</buch>

Sample periodical entry: 

<periodikum typ="RW" id="b43">
  <titel>Forschungsberichte des Instituts für deutsche Sprache</titel>
  <sigle>IdS</sigle>
  <ort>Tübingen</ort>
  <verlag>Narr</verlag>
</periodikum>

[Previous]   [Next]