Library of Congress>> Standards >> MARC >> Understanding MARC Authority Records


I t is impossible these days to read a library journal, attend a library conference, or even have an informal chat with other librarians without hearing the phrases "MARC format," "MARC records," or "MARC-compatible." Many library professionals have not had an opportunity to take formal courses explaining the important topics of library automation and the role of MARC, yet automated library systems remain an important part of their libraries.

This booklet will explain -- in the simplest terms possible -- what a MARC record is, and it will provide the basic information needed to understand and evaluate a MARC authority record. A companion document, Understanding MARC Bibliographic, focuses on the MARC bibliographic record.

Part I: What Does MARC Mean?
Part II: Why is a MARC Record Necessary?
Part III: MARC Terms and Their Definitions
Part IV: What is a MARC Authority Record?
Part V: Are MARC Authority Records Shared?
Part VI: MARC Data Issues
Part VII: In Conclusion

Part I

What Does MARC Mean?

What is a MARC record? A MARC record is a MAchine-Readable Cataloging record.

And what is a machine-readable cataloging record?

Machine-readable: "Machine-readable" means that a computer, can read and interpret the data in a record. The following pages will explain why this is important and how it is made possible.

Cataloging record: Cataloging records come in two types: 1) Bibliographic records, which contain information about a book, serial, sound recording, videorecording, etc., and 2) Authority records, which contain standardized forms for names, titles, and subjects that are used on bibliographic records and provide cross references in catalogs. Understanding MARC Bibliographic describes the first type, this booklet describes the second.

MARC Authority Record: MARC authority records contain the standardized forms of names for people, corporate bodies (for example, societies, businesses, institutions, etc.), meetings, titles, and subjects. In doing this, authority records provide authority control. Authority control means establishing a recognized form for an entity name and using that form whenever the name is needed as an access point in a bibliographic record.

For example, if you searched in your library's online catalog for the popular book, "The Adventures of Tom Sawyer," by Mark Twain, you could retrieve the following bibliographic record display:

TITLE: The adventures of Tom Sawyer / Mark Twain ; with an introduction by Robert S. Tilton.
AUTHOR: Twain, Mark, 1835-1910.
PUBLISHED: New York : Signet Classic, [1997]
MATERIAL: xxi, 216 p. ; 18 cm.
NOTE: Includes bibliographical reference (p. 213-216)
SUBJECTS: Sawyer, Tom (Fictitious character) -- Fiction.
Boys -- Missouri -- Fiction.
Mississippi River -- Fiction.
Missouri -- Fiction.
Adventure stories.

The AUTHOR and SUBJECT entries (highlighted above) are controlled, recognized forms of name and subject headings that are contained in separate authority records and used as access points in the bibliographic record.

Without authority control, finding particular books in a large library catalog would be like finding a needle in a haystack!

It is important to note that the formulation of a name, or subject heading in an authority record is based on generally accepted cataloging and thesaurus-building conventions. The MARC 21 authority record simply holds or carries these headings for use in library systems. Thus, it does not itself set the rules for formulating name or subject headings.

It is also important to note that unlike bibliographic records, authority records do not represent materials in a library's collection. Rather, they are tools used by librarians to achieve consistency among bibliographic records and provide a linking framework for related names and subjects in a catalog -- thus organizing the catalog to assist users in finding resources.

An authority record includes three basic components:

1) Headings, 2) Cross references, and 3) Notes. (MARC authority records often also contain additional information.) Although these components are described briefly below, they are also discussed throughout the rest of this booklet.

1) Heading: The standardized "authoritative" form of a name, subject, or title that is used for access points on bibliographic records. The purpose of using standardized names and subjects on bibliographic records is to help related records be retrieved together.

2) Cross references: References that direct a user from a variant form of a name or subject to the authoritative form (this is called a see reference) or from one authoritative form to another authoritative form because they are related to one another (this is called a see also reference). For MARC authority records, the references are carried or "traced" on the record for the authoritative heading.

3) Notes: Notes that contain general information about standardized headings or more specialized information, such as citations for a consulted source in which information is either found or not found about a heading.

An example of the use of an authority-type cataloging record in the catalog is one for the well-known French Romantic writer, George Sand. Users may approach the catalog with knowledge of either her real name, Amandine Aurore Lucie Dupin, or her pen name, George Sand. The authority record assures that all bibliographic records use her most commonly known name in its most frequently used form so that all of her works can be efficiently retrieved together -- and that there is a cross reference in the catalog from the other names and forms not used to the one used. The cross references enable either an end user to adjust a search to the correct form or the system to automatically adjust the search.

Part II

Why Is a MARC Record Necessary?

Why can't a computer just read an authority card? The information from an authority card cannot simply be typed into a computer to become part of an automated catalog. The computer needs a means to interpret the information found on an authority card. The MARC record contains a guide to its data, or little "signposts," before each piece of authority information.

The place provided for each of the pieces of authority records (headings, cross references and notes) is called a "field." The MARC file structure allows for records with an unlimited number of fields and unlimited field lengths. This flexibility is necessary because, for example, not all headings are the same length ("Plato" versus "Kennedy, Albert J. (Albert Joseph), 1879-1968").

The computer cannot assume a certain type of information begins and ends at the same position in every authority record. For example, a see also reference tracing will not always begin with the 220th character of the record and end at the 248th position. Therefore, each MARC record contains a little "table of contents" formulated according to a predefined standard.

Data "signposts": The computer must have assistance if it is to read and interpret the authority record correctly. The box charts below illustrate the information that these "signposts" need to convey.

If an authority record has been marked correctly and saved in a computer data file, computer programs can then be written to format the information correctly for displaying the information on a computer screen. Programs can also be written to search for and retrieve certain types of information within specific fields. Computer programs can also display lists of headings and references meeting specific search criteria.

Record with textual "signposts"

Heading area
King, Stephen,
See from reference tracing area
King, Stiven,
See also from reference tracing area
Bachman, Richard,
Note area
Source citation:
Source citation:
Information found:
His Carrie, 1974.
Washington post, 4/9/85
(Stephen King has written 5 novels using the pseudonym Richard Bachman)

Same record with MARC tags

100 1#  $a
400 1#  $a
500 1#  $a
670 ##  $a
670 ##  $a
King, Stephen,
King, Stiven,
Bachman, Richard,
His Carrie, 1974.
Washington post, 4/9/85
(Stephen King has written 5 novels using
the pseudonym Richard Bachman)

Why one standard? You could devise your own method of organizing authority information, but you would be isolating your library, limiting its options, and creating much more work for yourself. Using the MARC standard prevents duplication of work and allows libraries to better share authority information. Choosing to use MARC enables libraries to acquire authority data that is predictable and reliable. If a library were to develop a "homegrown" system that did not use MARC records, it would not be taking advantage of an industry-wide standard whose primary purpose is to foster communication of information.

Using the MARC standard also enables libraries to make use of commercially available library automation systems to manage library operations. Many systems are available for libraries of all sizes and are designed to work with the MARC formats. Systems are maintained and improved upon by the vendor so that libraries can benefit from the latest advances in computer technology. The MARC standard also allows libraries to replace one system with another with the assurance that their data will still be compatible. See Part V of this booklet for more information about the importance and benefits of using the MARC authority standard.

MARC 21: Because of the Library of Congress' large collections and its cooperative programs with other libraries, it is a primary source of authority records for U.S. and international catalogs. When the Library of Congress began to use computers in the 1960s, it devised the LC MARC bibliographic format, a system of using brief numbers, letters, and symbols within the bibliographic record itself to mark different types of information. The Library of Congress developed the companion authority format in the 1970's. The original MARC formats evolved into MARC 21 and have become the standards used by most libraries for bibliographic and authority records. The MARC 21 authority format, as well as all official MARC 21 documentation, is maintained by the Library of Congress. It is published as the MARC 21 Format for Authority Data.

A comparison of the same record with textual information and with MARC tags illustrates the compactness of the MARC 21 format. It's a matter of storage space. Look at the boxes above. The MARC 21 format uses "670" "$a" and "$b" to mark fields that contain source data found notes instead of storing the words "source," "citation," "information" and "found" in each authority record. This convention makes more efficient use of computer storage space.

Part III

MARC Terms and Their Definitions

This section contains information on how to read, understand, and use a MARC authority record. It deals with what librarians using a library automation system will see and need to understand on their computer screens when adding, editing, or examining records. The emphasis will be on the areas commonly used in authority records.

A box chart in Part II showed a MARC record labeled with MARC tag "signposts." The proper names of these "signposts" are field, tag, indicator, subfield, subfield code, and content designator. These MARC 21 terms are covered in this section.

1. FIELDS are marked by TAGS.

A field: Each authority record is divided logically into fields. There is a field for the heading, fields for reference tracings, and so on. These fields are subdivided into one or more "subfields."

As previously noted, the textual names of the fields are too lengthy to be reproduced within each MARC record. Instead they are represented by 3-digit tags. (Although online catalogs may display the names of the fields, the names are supplied by the system software, not by the MARC record.)

A tag: Each field is associated with a 3-digit number called a "tag." A tag identifies the field -- the kind of data -- that follows. Even though a printout or screen display may show the tag immediately followed by indicators (making it appear to be a 4- or 5-digit number), the tag is always the first 3 digits.

A sample of tags used in an authority record for a personal name are:

100 tag marks a personal name heading (author)
400 tag marks a personal name see from reference tracing
500 tag marks a personal name see also from reference tracing
670 tag marks s source data found note

Here is an example of a field. The number 100 is the tag, defining it as a personal name heading field.

100 1#   $a Woolf, Virginia, $d 1882-1941

The Cataloging Distribution Service (CDS, for short) of the Library of Congress distributes a detailed listing of all tags in both the full MARC 21 Format for Authority Data and a summarized version entitled MARC 21 Concise Formats. For continued work with MARC authority records, these specifications are highly recommended. They are detailed documents containing many examples. Information is also available on the MARC 21 web site at:

Even after a short exposure to the MARC 21 authority format, it is common to hear librarians speaking in /MARCese." Librarians who work with MARC records soon memorize the numbers for the fields common to the authority records that they create.

An outline of many of the most frequently-used tags is listed in Part VIII of this booklet. A brief list of some of the other tags is included Part IX.

2. Some fields are further defined by INDICATORS.

Indicators: Two character positions follow each tag (with the exception of Fields 001 through 009). One or both of these character positions may be used for "indicators." In some fields, only the first or second position is used; in some fields, both are used; and in some fields, like the 010 field, neither is used. When an indicator position is not used, that indicator is referred to as "undefined" and the position is left blank. In this booklet, it is the convention to represent a blank, or undefined indicator position by the character "#".

Each indicator value is a number from 0 to 9. (Although the rules say it can be a letter, letters are uncommon.) Even though two indicators together may look like a 2-digit number, they really are two single-digit numbers. The allowable indicator values and their meanings are spelled out in the MARC 21 documentation. In the example that follows, the 2 digits following the 100 tag (a 3 and a #) are indicator values. The 3 is the first indicator. The # is the second indicator.

A first indicator value of 3 in the personal name heading field indicates that the name is a family name, rather than an individual person's name. The second indicator value in field 100 is currently undefined. It thus contains a "#."

100 3#  $a Guelf, House of


A subfield: Most fields contain several related pieces of data. Each type of data within the field is called a subfield, and each subfield is preceded by a delimiter-subfield code combination. Fields 001 through 009 have no subfields.

For example, the field for a personal name heading includes a subfield for a personal name, numeration, titles and dates associated with the name, among several others.

100 0#  $a Gustaf
        $b V, 
        $c King of Sweden, 
        $d 1858-1950

A subfield code: Subfield codes are one lowercase letter (occasionally a number) preceded by a delimiter. A delimiter is a character used to separate subfields. Each subfield code indicates what type of data follows it. (For each field in the MARC 21 authority format, the MARC 21 documentation lists and describes the valid subfield codes.)

In the above example, the subfield codes are $a for personal name, $b for numeration, $c for titles and other words associated with a name, and $d for dates associated with a name.

A delimiter: Different software programs use different characters to represent the delimiter on the screen or on printouts. Examples are a double dagger (double dagger), an "at sign" (@), a dollar sign ($), an underline ( _ ), or the graphic symbol "delimiter symbol - 855 Bytes". In this publication the dollar sign ($) is used as the delimiter portion of the subfield code.

4. CONTENT DESIGNATORS is an inclusive term used to refer to tags, indicators, and subfield codes.

The three kinds of content designators -- tags, indicators, and subfield codes -- are the keys to the MARC 21 notation system. In his book, MARC for Library Use (2nd ed. (Boston: G.K. Hall & Co., 1989), p. 5), Walt Crawford calls the MARC system a "shorthand notation" system. The three types of content designators are the shorthand symbols that label and explain the authority record.

Part IV

What is a MARC Authority Record

1. The authority record content.


An authority record is made for each authorized form of a heading. For names, one of the most widely used authority files is the Library of Congress Name Authority File (or LCNAF). That file is used in this document for examples. The whole file is available from the Cataloging Distribution Service (or CDS) or may be searched for free online at: For topics or geographic names, the most used subject authority file is the LCSH. It is available from CDS in print and electronic form. There are many other subject heading lists, such as the Sears List of Subject Headings and the Art and Architecture Thesaurus. The form of a subject heading should match one on these lists or follow their rules for construction.

All headings are one of the following types: names, name/title combinations, uniform titles, or subjects. Examples of name headings include:

A name heading may be a personal, corporate, meeting, or jurisdiction (including geographic) name.

100 1#  $a Woolf, Virginia, $d 1882-1941
             (personal name heading)
110 2#  $a Association for Childhood Education International
             (corporate name heading)
111 2#  $a La Crosse Health and Sports Science Symposium
             (meeting name heading)
151 ##  $a Mexico
             (geographic name heading)

A name/title heading consists of both a name and a title of a work. The name portion contains a personal, corporate, meeting, or jurisdiction name. The title portion contains the title by which an item or a series is to be identified for cataloging purposes. It is especially used for classic works that have been published under different titles over time. An example of a name/title heading is:

100 1#  $a Dostoyevsky, Fyodor, 
        $d 1821-1881.  
        $t Crime and punishment

A uniform title heading consists of a title not associated with a particular author. A familiar example is:
130 #0  $a Bible. $l Latin. $s Vulgate

A subject heading may be a single term, phrase, or a group of terms. The tag for the subject is based on the first term or phrase in the heading -- topic, genre/form, geographic, chronological, personal name, corporate name, meeting name, or title. The other terms in the heading are in specific subfields that are called subdivisions.

100 1#  $a Shakespeare, William, $d 1564-1616  
        $x Criticism and interpretation $x History 
        $y 18th century
150 ##  $a Books and reading $z Argentina

Tracings and references

MARC 21 authority records contain two types of cross references that either lead a user from an unauthorized heading to an authorized heading (these are called "see references" and use 4XX tags) or from one authorized heading to another (these are called "see also references" and use 5XX tags). Actual cross references are generally not carried explicitly in authority records, however. Instead, variant form and related headings are "traced" in the authority record for the authorized form. Tracing cross references rather than making separate authority records for references enables efficient and accurate display of authority information by catalogs -- and the users see the information in the form that is appropriate for them. Cross reference displays may then be generated by systems which combine the contents of a tracing reference field and the 1XX heading field of a record.

Authority record information:
100 1#  $a Twain, Mark, $d 1835-1910
             [Standardized heading]
400 1#  $a Conte, Louis de, $d 1835-1910
             [See from tracing]
500 1#  $a Clemens, Samuel Langhorne, $d 1835-1910
             [See also from tracing]
Display generated from the above record:
 Twain, Mark, 1835-1910
      see also: Clements, Samuel Langhorne, 1835-1910
   Clemens, Samuel Langhorne, 1835-1910
      see also: Twain, Mark, 1835-1910
   Conte, Louis de, 1835-1910
      see: Twain, Mark, 1835-1910


The MARC 21 authority record may contain notes. These may be intended for display in the public catalog, or only for viewing by librarians using the records in their work.

Authority record information:
670 ##  $a Phone call to National Register of Historic Places
             [Note for cataloging use]
680 ##  $i Surgery performed on an outpatient basis. May be
           hospital-based or performed in an office or 
             [Note for public catalog use]

2. Structural components of authority records.

Like MARC 21 bibliographic records, MARC 21 authority records consist of three main components: the Leader, the Directory, and the Variable Fields.

The Leader provides information required for the processing of a record. The data elements contain numbers or coded values and are identified by relative character positions. The Leader is 24 character positions long and is the first field in a MARC 21 authority record. Much of the information in the Leader is for computer use in reading and processing the record and is computer generated. See Part X for more information about the Leader.

The Directory is a series of entries that contain the tag, length, and starting location of each variable field within a record. Each Directory entry is 12 character positions in length. The Directory is always generated by the computer. See Part XII for a more in-depth discussion about the Directory.

Variable fields are identified by a three-character numeric tag that is stored in the Directory entry for the field. Each field ends with a field terminator character. Part VIII describes the most common variable fields.

It is important to note that there are two types of variable fields. Variable control fields are the 00X fields. Although these fields are identified by a field tag in the Directory, they contain no indicator positions or subfield codes. Instead, they may contain either a single data element or a series of fixed-length data elements identified by relative character positions. The 008 field, for example, is referred to as Fixed-Length Data Elements, or Fixed Field Codes. Its 40 characters contain important information, but in an abbreviated form. Its data is often used to identify and retrieve records matching specific criteria. Part XI discusses the 008 field in greater length. The following is an example of the 008 variable control field:

008  860107in#acannaaan##########sa#ana#####u

Variable data fields, on the other hand, contain two indicator positions stored at the beginning of each field and a two-character subfield code preceding each data element within the field. The following is an example of the 100 variable data field (Personal name main entry field):
100 1#  $a Cameron, Simon, $d 1799-1889

The variable data fields are grouped into blocks according to the first character of the tag, which identifies the function of the data within the field. The type of information in the field is identified by the remainder of the tag. The blocks are:

0XX Standard numbers, classification numbers, codes
1XX Headings (authoritative and reference)
2XX Complex see references
3XX Complex see also references
4XX See from tracings
5XX See also from tracings
6XX Notes
7XX Linking entries
8XX Alternative graphics
9XX Reserved for local implementation

There are also parallels of content designation within the 1XX, 4XX, and 7XX blocks that may help you in identifying the content within authority fields. The following meanings, with some exceptions, are given to the final two characters of the tag of fields:

X00 Personal names
X10 Corporate names
X11 Meeting names
X30 Uniform titles
X48 Chronological terms
X50 Topical terms
X51 Geographic names
X55 Genre/form terms

The 1XX, 4XX, 5XX, and 7XX blocks and the parallel content designations within them are very important to remember when working with authority records. They allow you to "predict" the content of a field, even when you do not know precisely what is in it.

Part V

Are MARC Authority Records Shared?

As in the case of bibliographic records, libraries have been sharing authority information for many years. Although libraries may (and many do) maintain their own name and subject authority files, many institutions have found that sharing the responsibility of authority file maintenance is both cost effective and quality enhancing. Doing so prevents different authority librarians from reinventing the wheel by creating the same authority record over and over again. The long established NACO program (Name Authority Cooperative) allows institutions to help build the LCNAF authority file. More information about the NACO project may be found online at:

A corresponding but "younger" cooperation is SACO (Subject Authority Cooperative Program). This program was established to provide a means for libraries to submit subject headings and classification numbers to the Library of Congress. More information about the SACO project may be found online at:

MARC 21 allows libraries to participate in these and other cooperative cataloging programs by providing standard means for representing and communicating authority information in machine-readable form. Without MARC's stability, libraries would not be able to share their authority data as readily because each library's system could not "talk to" other systems. MARC prevents this virtual Tower of Babel by providing technical specifications that allow librarians, library vendors -- anyone at all -- who distributes authority data to communicate with each other. Likewise, MARC allows library automation venders to design their systems to receive and process standard MARC records.

You may now wonder about how MARC provides standard technical specifications to libraries throughout the world. The Library of Congress issues and keeps up-to-date the document, MARC 21 Specifications for Record Structure, Character Sets, and Exchange Media, to aid libraries and other organizations who create or acquire MARC 21 records. This document provides technical information on the structure of MARC records, the character sets used in MARC records, and the format for distribution media for MARC 21 records. It is intended for the use of personnel involved in the design and maintenance of systems for the exchange and processing of MARC records.

The Library of Congress serves as the maintenance agency for the MARC 21 formats for bibliographic and authority data. As part of that responsibility, the Library of Congress maintains the MARC Forum, an electronic discussion list for the formats ( that provides a conduit for broad, open discussion of proposed changes and other issues for MARC 21 users from around the world. The Library of Congress and the National Library of Canada also hold open meetings for discussion of changes to the MARC 21 formats.

In the United States, there are actually two groups that are responsible for reviewing changes to the MARC 21 format documentation specification: The MAchine-Readable Bibliographic Information Committee (MARBI) and the MARC Advisory Committee. MARBI is a committee of the American Library Association (ALA) and is composed of three representatives from each of the three function-oriented divisions of ALA: ALCTS (technical services); LITA (automation); and RUSA (reference). ALA tries to assure that all types of expertise are represented on MARBI. The broader MARC Advisory Committee is composed of representatives from the institutions and organizations that have adopted MARC 21 and that participate in its development, such as national libraries, the bibliographic networks, vendor groups, and other library and scholarly associations. MARBI meets in conjunction with the MARC Advisory Committee at each ALA conference (both annual and midwinter). The major activity of the Committee is the review of discussion papers and proposals submitted by and through the Library of Congress for changes or additions to existing MARC formats. When needed, the MARC Advisory Committee also develops new MARC format support in emerging areas. Decisions on changes are made by consensus of all review parties with responsibility resting with the adopting national libraries, such as the Library of Congress, the National Library of Canada and the British Library.

As you can see, individual libraries are the direct beneficiaries of all of this MARC 21 standards work. Without it, libraries would not be able to share authority records with ease and accuracy.

Part VI

MARC Data Issues

As computers and technology continue to advance, important issues arise about the quality of library data and computer-based library systems. You need to be aware of these issues and their importance to your library.

A. Data questions: When authority records are evaluated, some important questions which need to be addressed are: What is the quality of the MARC data? Which MARC 21 fields are present? Is the data based on Library of Congress MARC records? If so, is the full information that is available evident in the records? Is anything added to the records?

Some librarians ask several vendors of authority data for printed examples of their MARC records in order to compare their fullness and compliance with MARC 21 standards.

Clearly, the contents of the authority records will determine the success of a library's automated authority control efforts. It is very important to ensure receipt of the highest quality records available.

B. Software questions: The next logical questions to ask are: Does the library automation system being evaluated make full use of the cataloging information? Does it retain all of the data and the MARC 21 content designators?

Does the system load in the full record, with no unreasonable or false limitations on such things as the references it will index or the length of the note fields? Even though the disk contained excellent, full records, any library automation system will be able to use only what was transferred to the hard disk.

Are the MARC tags, indicators, and subfield codes still present on the librarian's data entry screen? Are the indicators used correctly? Are the subfield codes used correctly? (The information in Part VIII will help in evaluating correct use of the content designators.)

After records are added to a database, it is important to store them in the MARC 21 authority format. Future projects may require their use. The MARC 21 format is an industry-wide standard. As additional programs become available, a library's ability to participate in them could hinge on the quality of its authority records.

That brings up another point. Does the system allow for downloading, or writing the records back out to a disk to be used in other projects, such as participation in NACO? Will a program be available to write them in the MARC communications format?

Part VII

In Conclusion

In these pages you've learned what the acronym MARC means. You have seen why a standard format for identifying elements of authority data is important. You have learned to define and identify the three types of MARC content designators: tags, subfield codes, and indicators. You have also found out about what kind of information is recorded within the various content designator blocks and the parallel content designation within them. Most importantly, you have learned what a MARC 21 authority record actually is, how access points are recorded in authority records, and the difference between subject and name authority records.

To build on this basic introduction to MARC 21 authority records, additional reading on the subject or courses in online cataloging can be beneficial. Further readings are suggested in the bibliography. MARC 21 authority records may appear difficult at first, but with knowledge and use, it will begin to make sense. As you become more familiar with the MARC 21 authority format, the more simple it will become.

Library of Congress
Library of Congress Help Desk ( 10/28/2005 )