Resource Description II: Access and Authority Control

Previous post: Metadata structure (left side of a surrogate record)

• Signposts
• Element set modeling (metadata schemas)

This post will discuss: Metadata content (right side of a surrogate record)

• Professionally-generated: high quality rules-based data entry
• Authors-generated: variable quality of data entry

How do users ask for resources?

• Left side element modeling for various resources addresses this question. 

How do users verbalize their need?

• Content: Right side data entry for each metadata record ("access points"
• Problem: Normal variation in "natural language"
• Variance across individuals
• Variance over time
• Meta data content standards are needed and metadata/cataloging professionals are needed for implementation.

The "Dumb" Computer

When a search is performed using a computer, the computer matches query words (i.e. character strings) entered by user with words contained in the surrogate records of a file:

• Keyword search: search of all words in each record in the file
• Fielded search: structured search that is limited to certain metadata fields (e.g., title search or author search)

It is important to remember that the computer "matches" terms selected by human catalogers/indexers with terms selected by human users of bibliographic systems:

• Catalogers/indexers → terms → match ← terms  ← users
• Computers are getting "smarter" e.g. mapped misspellings [www.gogle.com]

Descriptive Cataloging

• Surrogate (i.e., catalog) record content created by professional and supervised paraprofessional catalogers. 

• Rules based (AACR2): Main and added entries are created for each catalog record representing the books, recordings, and other type of resources that are in a library collection.

• Sources are consulted for cataloging record content:
• Data taken from the resource in hand
• Data taken from authority (a.k.a. content standards)
• Name and authority files
• Title authority files

Main and Added Entries in Card Catalogs

• Main entry is also knows as the "primary access point."
• In the days of card catalogs, the main entry contained the full bibliographic record:
• Author main entry was the convention
• Title main entry was used for resources without authors
• Other "added entries," such as titles and subjects, had abbreviated bibliographic information on the cards in their card catalog drawers.

Current Main Entry Controversy

In an OPAC, there is only one "card" ie. the record in the database, so in theory, we no longer need to distinguish main entry and added entries. However, author main entry continues to be convivial with print bibliographies, and with the need for sorting and displaying retrieved records in an online catalog.

• Author
• Catalogers use the author as main entry for works by a single author. 
• For works with unknown authorship: title main entry.
• For works by multiple authors with synchronous (i.e. the same kind of) responsibility for work, the employ the "rule of three"
• Three of fewer authors: First author is main entry
• Four or more authors: Title is main entry
• For works by multiple authors with asynchronous (i.e. different kinds of) responsibility for work:
• Later editions of work may have different authors
• With certain types of resources there may be obvious differnces in author contributions. 
• Main entry choices in asynchronous cases:
• Use original author as main entry for new editions of works (except if new edition is changed significantly)
• Use original author as main entry for translated works
• Judgement call for artist/writer relationships:
• Art book with captions - Artist's name is main entry
• Art accompanies text - writer's name is the main entry
• Title
• For truly anonymous works
• For works with more than three responsible authors with none having primary responsibility.
• For works by multiple authors that are compiled by an editor.
• For works by corporate authors.
• For works accepted as sacred scripture by a religious body.

Authority Control

• Cataloger's decision concerning authorship: " I know who the author is (from the resource in hand), but what personal name data do I enter into the surrogate record?"
• Authority work involves the determination of authorized forms for entities known by variant forms, e.g.:
• Author name changes over time (e.g. pseudonyms)
• Variant spellings of personal names that proliferate over time
• Title changes of a work over time
• Authority records are maintained for personal names and some titles at the Library of Congress in various files.

Why Authority Control?

• To fulfill Cutter's 2nd objective (The Collocating Objective):
• Author criterion
• Title criterion
• To deal with the problem of natural language variation:
• Variation is how individuals "label" themselves (or have been labeled) over time (e.g. pseudonyms and variant spellings (see "Khadafi as an author search))
• Variation is how titles change over time (e.g. new editions or sacred works ("King James Bible" as keyword search and "Bible. English. Authorized" as an authorities search))
• To enable consistent data entry into cataloging and metadata records over time. 

Authority File

• An authority file consists of authority records.
• Catalogers and other metadata record creators use authority files for certain data entry tasks.
• Authority files are also an integral part of integrated library systems.
• Common authority tiles:
• LC Name Authority File - maintained collaboratively (NACO) according to AACR2.
• Getty Vocabularies - artist and geographical names
• International Standard Archival Authority Record - corporate bodies, persons, and families.

Basics of LC Name Authority File

• Available online
• Authority records are MaRC encoded, however, noth with MaRC for bibliographic data; rather with MaRC for authority data.
• Why are we using MaRC to encode authority records?

Functional Purposes of Authority Work

• To meet Cutter's first objective (the finding objective):
• To find a book when one of the following is known:
• Author (must distinguish between among persons or entities with the same name)
• Title (must normalize titles that have proliferated over time)
• Subject (must normalize "natural language" variation over time)
• To meet Cutter's second objective (the collocating objective)
• To show what the library has:
• By an author (collocation using author criterion)
• On a subject (collocation using subject criterion)

Authority Work - Record Creation

Current practice dictates the creation of an authority record for personal names (and some titles) containing the authorized "heading" that serve as the access point (i.e. the data to be entered into surrogate records). Records for each heading are maintained in authority files to track changes over time.

Authority Work - Heading Creation

When creating a new name authority name authority record, determination of the heading for a personal name is required:

• Which name to use?
• What form of that name to use?
• How to structure the name?

When creating new uniform title authority record, determination of the heading for a title is required:

• Which uniform title to use?
• How to format uniform title, including language and date

Authority Work - Which Name?

• Use the latest form of the lane has been changed:
• Chris Wallace to Notorious B.I.G. → use the latter
• Sarah Ferguson to Duchess of York → use the latter
• Use the predominant name of a person who is known by more than one name:
• Bill Clinton/William Jefferson Clinton →  use the former
• Mildred Zaharias/Babe Zaharias → use the latter
• Use different names for the same person if author name is identified with the subject are of that work:
• Charles Dodson (logic)/Lewis Carrol (children's stories) →  use both

Authority Work - What form of Name?

• Choice of form name:
• Variants in fullness (e.g. C.S. Lewis)
• Variants in spelling/transliteration (e.g. Khaddafi)
• Use the form used by the person in his or her own country of residence or activity and use predominant spelling (e.g. Qaddafi)

Authority Work - Structure of Name

Structure of personal name heading is usually family name (surname) followed by forenames usually followed by DOB/DOD. The heading should be structured as it would appear in the telephone directory of the person's home country.

• In Brazil, the last name of a compound surname is used.
• In Argentina, the first name of a compound surname is used.

Authority Work in Other Contexts

• Dublin Core - varies depending on who is entering the data - professionals or nonprofessionals:
• See TIMEA Metadata document for an example
• Dublin Core guidelines
• Online resource collections: Samuel Clemens versus Mark Twain
• Yahoo?
• Google?