Durance defined authority control as the function of providing access to the holdings of a repository through names and subjects, noting that authority records are being constructed in such a way that their relationships, or nonrelationships, to each other are apparent.1 She wrote, “In effect, authority control in an archival setting introduces order into, and a road-map through, the complexities of corporate organizations and their predecessors, personal names, functions, and subjects of documents, and their interrelationships.”² One of the fundamental steps in creating an authority record is for the cataloger to determine whether an intellectual relationship exists between the access points. Establishing and linking all the possible access points will collocate all relevant headings, including cross references, and display them in a meaningful way.

Harper and Tillett discussed the importance and the value of authority control in their “Library of Congress Controlled Vocabularies and their Application to the Semantic Web.”³ Among the issues discussed in their paper were the value of authority control and its important role in searching, navigation, collocation, and linking between various tools and resources. This is accomplished by establishing connections between related headings using cross-references, such as “see” and “see also.”

Vassie provided an overview of many of the major issues confronting multiscript cataloging and display.⁴ His article opened with the history of non-Latin script collections and access and went on to describe the situation at the Zayed Central Library of the United Arab Emirates University. The author correctly identified a lack of multiscript authority control as the major issue in non-Latin based language research. He went on to describe two options “for tackling authority control for names of authors represented in both available scripts in a bilingual, bi-script catalogue.”⁵ The article ended with a call for an international, controlled Arabic thesaurus that would be created and maintained according to international standards.

In her article on Arabic name authority, Speirs Plettner provided an overview of the Arabic name authority file from its origins to her time of writing (2003).⁶ She started with an overview of the problem of transliteration of Arabic scripts and the decision to maintain two files, one for original Arabic scripts and anther for their transliterated form. She pointed out the limitations of the Anglo-American Authority File and called for including institutions outside the English language sphere to participate in a global multiscript name authority file. Speirs Plettner continued with an overview of the discussion on how to enhance the structure of authority records to accommodate the non-Latin scripts. As with any idea that calls for cooperation between institutions, this one also will require commitment from the global library community.

Kuntz addressed the issue of transliteration, or romanization of Arabic script, and compared the Library of Congress/American Library Association transliteration scheme with that developed by the International Organization for Standardization (ISO).⁷ As he discussed the pros and cons of each system, Kuntz maintained that some form of transliteration must be used and that the libraries should inform the users which system is being employed. Kuntz indicted that both the Library of Congress (LC) and the ISO Arabic Romanization schemes have proven to be problematic, but he is not in favor of abandoning transliteration in favor of an exclusive use of vernacular.

Khairy took a different approach in her presentation at the 27th MELCOM (Middle East Libraries Committee) International Conference in 2005.⁸ She addressed some of the issues facing creators of Arabic online catalogs in Arabic speaking countries. She focused her attention specifically on the use of different reference sources when an authorized access point for a classical Arabic name is created and how variations in the form of the name could be a barrier to cooperation in constructing an Arabic script union catalog and authority file. She pointed to the lack of specific cataloging tools designed for cataloging Arabic materials as a major issue.

The Bibliotheca Alexandrina (BA) was created in association with the United Nations Educational, Scientific, and Cultural Organization (UNESCO). Snohetta Arkitektur Landskap of Oslo, Norway, won the international competition to design the library.⁹ Since its completion in 2002, the BA functions as a major library and cultural center serving the city of Alexandria and the region. In the words of Ismail Serageldin, Librarian of Alexandria, “The New Library of Alexandria, the New Bibliotheca Alexandrina is dedicated to recapture the spirit of openness and scholarship of the original Bibliotheca Alexandrina.”¹⁰

The ancient Library of Alexandria was situated on the shores of the Mediterranean Sea and dates to around 300 BCE, when the Ptolemy dynasty ruled Egypt. Although it was subsequently destroyed, its story provided inspiration for the creation of the new BA some two thousand years later. Today the library houses collections of rare books and manuscripts, a children’s library, a youth library, a library for the blind, and many others. It is open to scholars, researchers, and the public.

The BA’s online catalog is tri-lingual, providing access to its collections of materials in Arabic, English, and French. In 2010, the library received a half a million French books as a donation from the Bibliothèque nationale de France. This gift made the BA the sixth-largest Francophone library in the world. The BA is also the largest depository of French books in the Arab world, surpassing Tunisia, Algeria, and Morocco, in addition to being the main French library in Asia and Africa.¹¹

The organizational structure of the library consists of the General Directorate and six sectors: Academic and Cultural Affairs Sector, the Director’s Office, the External Relations Sector, the Finance and Administration Sector, the Information Communication Section, and the Library Sector. The Library Sector is headed by the chief librarian, who supervises six departments: Arabic Bibliographic Standards, Francophone Library Department, Information Services Department, Main Library Department, Special Libraries Department, and Technical Services Department.

The Technical Services Department consists of two sections: Acquisitions and Bibliographic Access. The latter includes three units: Cataloging, Materials Processing, and Periodicals. The BA online public access catalog (OPAC) was designed to manage the multilingual bibliographic and authority files (Arabic, English, and French) and provide users with multilingual access to its materials. It allows the user to search the catalog in all three languages. The BA is currently using the Virginia Tech Library System (VTLS) integrated library system (ILS) known as Virtua. Virtua’s cataloging system is compatible with the Unicode standard and stores data in bibliographic records in the Unicode encoding scheme. This allows the catalogers to access and create records in their natural languages (e.g., Arabic, Hebrew, Turkish, etc.). Virtua also supports direct input, indexing, and display of characters in Unicode from a single, standard workstation. For example, catalogers or patrons can change the interface language and search for bibliographic records in their natural language without affecting others using the system.¹²

At the BA, Arabic, English and French language materials are cataloged in their original language of publication. The same also is true of the authority file. The BA built a local authority file for personal names, corporate bodies, and subject headings where the authority record is presented in Arabic, English, and French languages.¹³

One of the most powerful features of the BA catalog is its linking feature, which links records in multilingual scripts. This means that when an author writes in Arabic and in French, the library creates an Arabic language record for the Arabic books and a French language record for the French language books. Two authority records also are created for this author. One is based on the Arabic script bibliographic records and the other is based on the name of the author in the French language records. The BA staff link the two authority records, which allows the user to search the database in either language and retrieve all the works by the same author regardless of the language of publication.

The Authority Control Section is a major component of the Cataloging Unit at the BA. The section is responsible for all of the authority work, including personal names, corporate bodies, and subject headings. Authority processing is divided into Arabic and Latin. A single staff member has responsibility for the Latin authority processing. This includes names and subjects in English and French. Three staff members work on creating Arabic authority records. A supervisor and an assistant supervisor also work with authority records. Each staff person creates between seventy and eighty headings a day.

The BA started to build its local Arabic authority control file in 2004 and the subject authority control file in 2006. The goal was to build a local Arabic authority file in Arabic and Latin scripts for the materials that are acquired and housed at the BA. The main reasons for creating the BA Authority File were the following:

• Specific tools designed for use with the Arabic language are lacking.
• Most tools used today in cataloging Arabic materials are translations of the Western tools.
• The needs of Arabic users are different from those in Western countries. An Arabic language catalog helps to resolve some of the difficulties with non-Latin scripts and offers search options that are not always available elsewhere.

According to Khairy, the intention of creating this file was to

implement an Arabic Script Union Catalog and Authority file by establishing something similar to NACO in Arabic script, which may be called ANACO (Arabic Name Authority Cooperation). This will allow an option for exchanging Arabic script name authority headings where Arabic script online libraries will have the advantage of keeping their local methodologies and practices, meanwhile conforming to one methodology under cooperation of the ANACO.²⁰

In building the BA Arabic Name Authority file, the BA followed the MARC 21’s Model B, which states that for simple multiscript records, “All data is contained in regular fields and script varies depending on the requirements of the data. Repeatability specifications of all fields should be followed.”²¹ The BA creates two different Name Authority Records (NARs): one for Arabic and one for Latin script. The two records are linked by using the MARC 21 field 7xx (Heading linking entry) in the record. In this field, the BA uses the romanized (Latin-script equivalent) form of the classical name from the LC authority file. If the name is not in the LC authority file, the library romanizes the name according to the ALA-LC Romanization Table for Arabic.²²

To build the authority file, the BA staff identified several issues related to the creation of authority records for classical Arabic names and corporate bodies. These included the choice of element of the classical name that could be used to establish the authorized access point, the issue of transliteration, misspellings and various ways of writing the Arabic names, and the use of different reference sources to establish authorized forms of the name. Modern personal names and corporate bodies are not as much a problem as the classical names. As a first step in the process, the BA decided to record all the forms of names found in the main reference sources commonly used by libraries that collect Arabic script materials.

The intention of the BA is to create an authority file that is based on the LC authority file structure. The BA is using the same reference sources for Arabic names used by the LC and follows the NACO rules for establishing Arabic names. According to Khairy,

The BA practice is to conform to one main reference source and methodology for entering the established Arabic script name heading, to record the rest of the forms in Arabic script within the other three reference sources as unused forms, and recording the transliterated form into Latin script according to the LC and OCLC in MARC 21 field 7xx (Linking Entry fields).²³

Currently, the BA is not using the MARC 21 field 7xx field. Instead, the BA uses the MARC 21 field 500 (see also from tracing) to add the transliterated form of the Arabic name. This is a temporary situation until the BA system is configured properly.

When the author is writing in Latin and Arabic scripts, the catalogers will create two authority records and link them to each other to do the following:

• Identify and collocate all the works by the same author.
• Establish consistency in the language of the bibliographic record and a cross reference to records in other scripts.
• Maintain authority records in the language of the work. In this case, the three main languages the BA catalog uses are Arabic, English, and French.
• Create Arabic language records for the primary user groups at the BA and have those records linked to records created in other languages. Access to the catalog for Arabic script users is of primary importance as it reflects both the emphasis of the collections and the preferences of the user community.

The authority record for the Latin script work is derived from the LC authority file. The BA staff create the authority record for Arabic script. For the Latin script record, the BA staff add the Arabic script as “see also” in MARC 21 field 500; see figure 7. In the Arabic script authority record, the BA staff adds the Latin script in MARC 21 field 500 as “see also”; see figure 8. The BA uses the MARC 21 field 500 temporarily because of issues with system configuration. In the future, the linkage will be through the MARC 21 field 700. Notice that in the Arabic script record, no additional MARC 21 field 670 (Source of Data) is present for the sources of the Latin heading.

Searching the BA OPAC under “Mustafa, Saleh Lamei” will retrieve a heading that is not used; see figure 9. The heading displayed with the minus sign in front is a cross reference that comes from the MARC 21 field 400 (see from tracing). Clicking on the minus sign heading “Mustafa, Saleh Lamei” will display the cross reference for the romanized form of the name. The number “1” in front of the related term shows that that the library has another title in a different script by the same author. Clicking on the plus sign in front of the author’s name will collocate the title under the romanized and the Arabic forms of the name.

Searching by the Arabic script will retrieve all the titles by this author. The plus sign in front of the hit count (figure 10) will show that this author has another title in different script.

A wide range of bibliographic agencies on the national and international level are creating and maintaining their own authority files. Significant overlap exists between the contents of these files, and the responsible agencies are expanding significant resources to maintain them. The concept of creating a Virtual International Authority file that would link these various international files was introduced in the early 1970s at the International Federation of Library Associations and Institutions (IFLA).³²

The LC, Deutsche Nationalbibliothek, and Bibliothèque nationale de France, in cooperation with fifteen other national libraries and other agencies, jointly administer the VIAF project.³³ The VIAF explores virtually combining the name authority files of participating institutions into a single name authority gateway service. As Tillett pointed out, the VIAF “is virtual because it is not really a file itself, but a linked system that connects existing Authority files.” The VIAF contains names of people and corporate bodies.³⁴ Other access points such as geographical names are planned for the future.

Objectives of the VIAF as defined by Bourdon are to do the following:

• Share authority data around the world
• Reduce cataloguing costs
• Simplify authority control internationally
• Provide authority data in form, language and script users want
• Provide linked data services³⁵

VIAF has the potential to solve many name authority problems that result from romanization of names written in non-Latin scripts. By providing links between records in numerous languages and scripts, it offers searching options in all those languages and produces multiscript cross references. This localization of bibliographic data by making local versions of names (e.g., in different scripts, spellings, or other variations) available for searching and display makes bibliographic searching easy and more effective.

The BA is participating in VIAF by submitting its authority records. Every month, the BA staff collect bibliographic and authority records and send them to OCLC for matching and linking the records to other authority files. In case of problems, OCLC returns the records to BA for correction and these are resubmitted the following month.

BA is creating an Arabic union catalog that can provide services to native Arabic speakers, researchers, and scholars. Early in the process, the BA decided to the use Arabic script instead of romanization when creating bibliographic and authority records. Despite that decision, the structure of the authority file still conforms to LC and PCC NACO practices. To make sure that the conformity is maintained, the BA uses the Arabic references sources identified earlier to establish preferred access points and MARC 21 for authority control.

Khairy stated that

the goal of the BA is to implement an Arabic Script Union Catalogue and Authority file by establishing something similar to NACO in Arabic script, which may be called ANACO, Arabic Name Authority Cooperation. This will allow an option for exchanging Arabic script name authority headings where Arabic script online libraries will have the advantage of keeping their local methodologies and practices, meanwhile conforming to one methodology under cooperation of the ANACO.³⁶

This approach opens the way for future cooperation with other libraries in the area. Thanks to their association with VIAF, the BA could become a conduit for including other Egyptian libraries in a larger project. More coordination between the libraries in the region as well as in Egypt itself is needed to achieve this goal.

The cataloging unit at the BA has undertaken an ambitious, multifaceted project to create an Arabic language authority file, similar to the LC PCC NACO project. Their vision is to create an authority center that would serve the Arab world through cooperative arrangements. The project is still in its initial phases, but expectations are that it will develop and grow as circumstances permit.

A number of important steps already have been accomplished. First among these is the effort to resolve the issues with the classical Arabic names. The BA has established procedures for creating standard forms for classical Arabic names. These procedures follow those established by the international library community, including NACO. However, the BA has selected and is using a set of standard reference sources in Arabic for establishing authorized forms of names. These are published sources that are generally available, although not all are being used by the international community. This is a very important development that can establish a foundation for future cooperative work between libraries in the Arab world and beyond.

Indirectly, the BA initiative has helped in resolving the problem of romanization of non-Latin script. Until now, most libraries in the Western world have transliterated non-Latin scripts because most of the databases they use do not have the capacity to display non-Latin scripts. By creating catalog records in Latin and non-Latin scripts and linking them through their authority records, the BA has found a solution to the transliteration dilemma. Because the BA creates records in multiple scripts and the BA OPAC has the capacity to display all the scripts, transliterating records is no longer necessary.

The BA has now established an in-house authority database for its own collections. This database is growing exponentially, with nearly one hundred records added each day. The vast majority of the authority records the BA creates are for Arabic language authors and corporate bodies. The BA creates records for these works in Arabic that are searchable in that language. These records provide links to bibliographic records in languages other than Arabic when the author writes in multiple languages.

The BA authority database is currently unavailable to the broader library community. However, the BA authority records are available to the international community through VIAF. These records are integrated with and linked to the records that are being contributed to VIAF by other national and international libraries. This is an important step toward future collaborative efforts between the libraries in Egypt and the rest of the world.

Some of the solutions developed at the BA can lead to future developments in resource-sharing and cooperation on a global scale. Their system of creating authority records in three languages and cross referencing them to provide multilingual displays is one major step toward overcoming the transliteration dilemma. This could serve as a model to many institutions that are currently transliterating and transcribing foreign-language records and creating name headings in a single language. This could be an opportunity for other libraries to rethink their current process of authority control. BA’s decision to contribute their records to the VIAF program may be an encouragement for other libraries in the region to follow in their footsteps.