Authority Control in the 21st Century: An Invitational Conference
The concept of authority control and the need for it has been under discussion in Germany only since the eighties - for barely ten years. As a result there is little or no technical literature on the subject.
The need for control really became apparent as library networks and OPACs continued to grow. The authority lists in use in most of the bigger libraries were no longer effective. The prerequisites for the construction of authority files were electronic data processing and cooperative cataloging projects. In the mid-eighties these conditions were realised.
A common concern began to be felt:
Authority files
In spite of this they require a lot of effort and expense, and they demand coordination. It is sometimes doubted whether the benefits of an authority file for modern names would outweigh the costs.
I am still not absolutely sure if there is a common understanding in Germany that authority files are an indispensible aid for authorized access to OPACs, at least to huge OPACs. This can be seen in particular in discussions in expert groups and especially among a considerable number of the most influential people in the German library world with reference to the revision of RAK § 320 (the most common heading for names of persons instead of abbreviation of the second forename).
Authority files in Germany are separately created files for names of persons, names of corporate bodies and titles. As far as names of persons are concerned many separate files have been created in recent years; they will be combined in the near future.
The first project of this kind was the Gemeinsame Körperschaftsdatei (GKD = authority file for corporate names) set up in the seventies. Interest increased with the introduction of computerized cataloging. An important impetus was provided by the recon projects of Göttingen and Munich: the names of persons who were active before 1850 especially proved in need of authority control.
Name authority files are usually built as a by-product of cataloging, with a few exceptions, which will be discussed below.
2.1.1 Authority File of Personal Names : Personennamendatei (PND)
PND indeed started as a by-product of recon projects which were sponsored by the Deutsche Forschungsgemeinschaft (DFG) in the eighties. Acting on a proposal from the DFG the Deutsches Bibliotheksinstitut (DBI - German Library Institute) built up a database of personal names linked to short titles.
The following recon projects were involved:
The following names of persons were identified and entered as part of special projects - not as by- products of cataloging:
All authority files mentioned above were combined in DBI-PND.
As PND gained in importance and acceptance, a national database including modern names was felt to be needed. A new concept (including data format, scope, validity for descriptive and subject cataloging, update and editing functions) was worked out and resulted in a new project from the DFG: the construction of a new PND by the Deutsche Bibliothek in cooperation with the Bavarian State Library, the regional network of Nordrhein-Westfalen in Cologne and the Central Archive of Autographs in Berlin. The project started in October 94 with the transfer of the DBI-PND to the PICA-Iltis-System of the German Library with the records mentioned above. It was opened to the public in September 1995.
The following records will be included in the PND - in addition to DBI-PND:
(loading from March to September 96)
Start of loading in October 96:
The database will comprise about 1.5 million records.
More dates and facts:
The data structure of PND takes account of the difficult rule (or rather political) situation: PND provides records with names of persons which are individualized and records with names which are not individualized (so called tp and tn records).
2.1.2 Corporate Names : Gemeinsame Körperschaftsdatei (GKD)
As already mentioned, the GKD was the first authority file in Germany. With the introduction of RAK and the accompanying introduction of corporate bodies cataloging required a good deal of additional effort in order to provide entries under corporate names. The two state libraries in Berlin and Munich and the Deutsche Bibliothek in Frankfurt collected corporate names. These files were combined to create the first authority file in a common project involving the three founding libraries and the German Library Institute (DBI). In October 95 the GKD comprised 550,000 headings with about the same number of cross-references. All the regional library cooperatives use GKD. The GKD record is directly linked with the network holdings via code number. The records are updated by these code numbers.
Bodies involved:
Like almost all major authority projects GKD is also sponsored by the German Research Society.
To conclude the chapter on name authorities: I am convinced that headings are no longer a real problem as long as
And that would make national and international authority work much easier.
2.2.1 German National Serials Database : Zeitschriftendatenbank (ZDB)
Zeitschriftendatenbank is the official name of the network, but the term is also used for the database itself and for its major output, the national serials union catalog.
The ZDB continues the tradition of earlier conventional serials files such as GAZ, GZS and GAZS, which were edited from 1914 to 1939 by the former State Library and from 1950 - again by the Staatsbibliothek Preussischer Kulturbesitz. The idea of an automated system had arisen in the late sixties in order to produce cumulated files. Cooperation between editorial staff in the Staatsbibliothek and systems staff in the Arbeitsstelle für Bibliothekstechnik", later the Deutsches Bibliotheksinstitut, began at that time and still continues today.
Three major developments occurred during the eighties:
The scope of ZDB in October '95 was 820,000 records of serials with over three million holdings from 3,500 German libraries. The input is decentralized and comprises about 90 active participants. According to German university tradition the main library takes care of the serial holdings of the departmental and institute libraries (sometimes over 100!). All title inputs have to undergo editorial authority control by staff in Berlin, while local holdings can be added or removed by any active participant.
As regards GKD there are steering bodies for authority control and cooperation and a users' council. Format and rules are defined by ZETA and RAK.
ZDB produces quarterly outputs on microfiche, a twice-yearly CD-ROM-version, and countless offline updates for regional networks and local libraries. A WWW OPAC is being prepared.
2.2.2 Authority File of Uniform Titles from Antiquity
TITAN was the first joint national authority project for descriptive and subject cataloging (sponsored by the German Research Society). It will show (has already shown) the need for coordination of the two aspects of cataloging - formal and subject - which are traditionally separate from each other in Germany. There needs to be a harmonization of rules between RAK and RSWK (the subject cataloging rules). Many adaptations have already been made - in order to make a common OPAC possible.
On the basis of PAN (Personal name file for antiquity) the correct uniform title has to be defined for single editions of works (RAK says the most common in the academic tradition) including cross references from different German, Greek and Latin title versions. The scope is defined by PMA (authority file of medieval names) which declares all persons who died before 500 to be persons of antiquity.
The University Library of Tübingen is responsible for authority work on nonchristian titles, the Deutsche Bibliothek for the early Christian titles (except for the Bible; here again title headings vary between RAK and RSWK!).
2.3 Subject Authority File : Schlagwortnormdatei (SWD)
Stages in the development of the subject authority file were:
It very soon became clear that an authority list demands continuing control. A cooperative model for authority control - sponsored by the German Research Society - was conceived instead of centralised authority control. This cooperative model was considerably enlarged by several regional networks (first additional cooperation partners were DDB, the Bavarian Library Network and the Northrhine-Westphalian Library Network, now joined by the Southwestern Library Network, the Göttingen Network and Salzburg for the Austrian Network).
SWD comprises headings for all subjects and categories: names of persons, geographic and subject terms as well as period and form headings including cross-references (about 300,000 records). SWD is an open file; new headings are based on practical subject cataloging.
The subject heading is chosen according to RSWK rules. The commonly known designation is preferred: usually names and terms in the German language - contrary to RAK. The heading may consist of one or more words: noun, compound, adjective-noun combination, or phrase. Combinations of words and letters with figures and special characters are permitted. Homonyms and words with multiple meanings are differentiated by additions in angle brackets. Compounds are preferred to adjective-noun- and other combinations. Multi-element subject headings for complex concepts can be constructed to form subject heading chains. These consist of a main subject heading and one or more subdivisions.
All subject headings are classified according to the Systematic List for all subjects" (Systematische Übersicht aller Fachgebiete), with a country code for geography and ethnography as defined by the German Library.
SWD is an online-file in the PICA-ILTIS-System. There is twice-yearly output on microfiche and a weekly update on magnetic tape and diskettes. The subject headings are used in the German National Bibliography, and increasingly in public and academic libraries as well as in regional networks.
To conclude the chapter on German authority files: when German regional networks use the national authority files almost all of them have a link to title records, that means: if an authority record has been changed the title record will be updated automatically.
Ann.: This is no systematic study
Very often in AACR the established English form is entered as the heading - especially for ancient, medieval, royal names etc. - , whereas in RAK the principle of citizenship is used in most cases with a few exceptions (Latin form for Greek authors, German for - absolutely - unknown vernacular form etc.)
Prefixes for modern names are used in a similar way with the following exception: if the prefix (article and preposition) is entered preceding a surname this name consists of one filing entry (word) without any spaces.
Additions to a name (so-called Ordnungshilfen - filing qualifiers - ) are entered in angle brackets in a given form (e.g.: Walther <von der Vogelweide> or a standardized form (Johannes Paulus <Papa, II>, Elizabeth <Great Britain, Queen, II>).
One of the major differences - the abbreviation of the second forename - has now been revised: the commonly used form is entered as heading. Still more convergence is needed: Except for ancient, medieval names, names of royalty etc. and of course names that are used for subject cataloging, names are not individualized": no additions are provided for identical names.
The definition of a corporate body differs between RAK and AACR. AACR defines more corporate bodies, e.g. projects and buildings, which are not corporate bodies in RAK.
The REUSE project (OCLC and LC with the State and University Library of Göttingen) will carry out a statistical evaluation of the differences.
The German priority list"- a list of reference sources that are to be consulted in prescribed order to ascertain the most commonly used name heading - has accepted NAF as second or third source (for German and nonroman names). The first source to be consulted is PND. Name entries of NAF of course have to be changed into RAK forms. This may indicate the acceptance of NAF.
The problem of varying headings between different nations could easily be solved if authorized national forms were combined into an international NAF. As far as Germany is concerned we will have to individualize personal names, that means to differentiate identical names as AACR does. This is a sine qua non in order to participate in international authority work.
As for corporate names, variant authorized entries should be permissible as well. The first step to adaptation will be to define the same designation for a corporate name in order to match (and combine) records.
One of the major problems of titles has been differing romanization. If original scripts could be entered in an online catalog, varying romanization would no longer be a problem. The REUSE project will examine whether an algorithm for variant romanization forms can be found.
The differences in uniform titles - except for collective uniform titles - are not of overriding importance. The designation of collective uniform titles (in AACR) is in the best traditions of the Prussian Instructions. I think a realignment of AACR and RAK towards an online environment will remove the differences. In this case collective uniform titles (in AACR) and designation of collections (Sammlung in RAK) should result in codes that can easily be exchanged.
We do hope for common rules for online catalogs.
But to start with the German isle has to be made a peninsula and hopefully later a part of the continent of rules and formats.
| Advanced Search | Careers at OCLC | Feedback | Privacy Policy | ISO 9001 Certificate | ©2003 OCLC Online Computer Library Center, Inc. |