DISCUSSION PAPER NO. 100

DATE: June 16, 1997
REVISED:

NAME: Recording additional characteristics in USMARC Authority records

SOURCE: Library of Congress

SUMMARY: This paper discusses several characteristics -- language, script, transliteration, country (nationality), and catalog rules -- that might be coded in Authority records, indicating possible uses and problems, along with techniques.

KEYWORDS: Language; Script; Transliteration; Nationality; Catalog rules (Authority Format)

RELATED: 97-10

STATUS/COMMENTS:

6/16/97 - Forwarded to USMARC Advisory Group for discussion at the June 1997 MARBI meetings.

6/28/97 - Results of USMARC Advisory Group discussion - The different characteristics were discussed with more concentration on the language aspects of headings. Several saw uses for language of heading, when one agency is trying to serve different language groups from one record. Some saw that as an application problem and not a communications format concern. Others felt the difficulty in making a determination of which language, increasing cataloging expense. One participant reported that a European to provide access to authority records from different countries considered combining all records for an entity into a master record, identifying languages, etc. They decided to take a more decentralized approach, leaving records in their "own environments" and providing access to records in those environments. That is a little more like Model B, except that the records are not necessarily linked. Discussion of script centered on the usefulness of the script code explicitly in the $6 subfield. OCLC, VTLS, and EOS all indicated they did not use that coded information.

Transliteration was seen as useful in some cases but problematic as there are so many if considered from a global viewpoint.

One participant spoke for the nationality at the record level, although the group pointed out many difficult examples. In the more clear cut cases, it would not take much time to code and the information could be somewhat useful. It was pointed out that if systems then allowed searches by the characteristic, the less obvious cases might be erratically retrieved.

There was little discussion of the 880 as the user is largely unaware of the field (most displays show duplicate fields). This item will need more review and consideration by systems staff.

The participants were encouraged to continue the discussion on the USMARC list in preparation for more discussion at Midwinter. There was a request for some subject examples in a revision of the DP.


DISCUSSION PAPER NO. 100:  Coding additional characteristics

1.   INTRODUCTION

The USMARC Authority record is multifaceted and flexible,
accommodating agencies that require records for established
headings, subdivisions of headings, references, and node labels. 
It supports various models for creating authority records, but in
particular the general model of established heading with associated
reference tracings.  There has been discussion over the last few
years about the need to identify various characteristics of
headings in authority records in order to support cataloging and
retrieval activities.  The characteristics include language of
catalog, language of heading, script, transliteration scheme,
nationality, and rules.  MARC already supports recording of this
information in some situations, but a look at the primary models
used for authority records may indicate more generalized approaches
would be useful for the future.

The following standard vocabulary is used below.  It is adapted
from an IFLA document that specifies displays of name authority
data, Guidelines of Authority and Reference Entries.  The headings
are defined with respect to their relationship to other headings:

Uniform heading -- A heading established in a form to be followed
without variation whenever the heading appears in a bibliographic
record.  (e.g., 1XX heading)

Parallel heading -- An alternative form of the uniform heading
based on another language form of the name or title.  (e.g., some
4XX or 7XX headings)

Related heading -- One of two or more uniform headings, each of
which is bibliographically related to the other(s).  (e.g., 5XX
heading)

Variant heading -- A heading in a form other than that established
as the uniform heading.  Generally such a heading is either based
on a variant name sometimes used by the person or corporate body
itself, or sometimes used by others to identify the person,
corporate body, or work; or constructed on a pattern different from
that used to establish the uniform heading.  (e.g., 4XX and 7XX
headings)

Two basic models for authority records are illustrated below,
labeled A and B.  Model A takes the approach of including a heading
and all related and variant reference tracings in one record,
regardless of language differences.  In Model B separate records
are made for parallel headings needed for different language
catalogs, and the records are linked via the 7XX fields.  

The models are presented below first vis-a-vis the indication of
language of catalog, as that is a fundamental characteristic of a
file of authority records that constitutes a catalog (all headings
interrelated), rather than a simple resource file (records
essentially independent of each other).  Although the discussion
applies equally to subject thesauri, to simplify it is presented in
terms of name authorities, with a note on subjects and
classification schedules at the end.

Note: Diacritics and special character characters have been
normalized to ASCII in the following examples.

Language of Catalog

The language(s) of the expected users of a catalog is fundamental
to an authority file.  Most rules for construction of catalogs
specify the language(s) to be used for additions to headings and
notes in the record.  For example, AACR2, when used in an English-
speaking country, specifies that uniform headings be established in
the "original language" of the entity's name (transliterated as
needed, and with some exceptions) and that qualifiers and notes be
in English.  The cross-reference structure for the uniform heading
is affected by the rules and the languages used to formulate the
heading.  The following are the models used for constructing
authority files.

Model A:

Each record is constructed around one heading for an entity (the
1XX uniform heading) and that heading is appropriate for a catalog
in the language designated at the record level by the 040$b field. 
The 4XX and 5XX traced reference headings constitute the reference
structure for the 1XX heading in that catalog.

All parallel forms of the 1XX heading in other languages are
treated as variant headings and traced in the 4XX fields.  Headings
that would belong in a reference structure of a parallel form may
also be among the 4XX variant headings.  

Ex.1 040 $beng                                          n79006935
     100  1#$aMarx, Karl,$d1818-1883
     400  1#$aMarukusu,$d1818-1883
     400  1#$aMarx, Carlos,$d1818-1883
     400  1#$aMarks, Karl,$d1818-1883
     400  1#$aMarks, Karol,$d1818-1883
     400  1#$aMarksas, Karolis$d1818-1883
     400  1#$aMac, Cac$d1818-1883

Ex.2 040 $beng                                          n83201056
     100  1#$6880-01$aZemtsovskii, I. I.$q(Izalii Iosifovich)
     400  1#$6880-02$aZemtsovskii, Izalii Iosifovich
     400  1#$aZemtsovskiy, I.
     880  1#$6400-02$a<name in Cyrillic>
     880  1#$6100-01/cyrillic$a<name in Cyrillic with
          initials>$q(<qualifier in Cyrillic>)

Ex.3 040 $beng                                          n82020426
     110  2#$aNational Library of Canada
     410  2#$aBibliotheque nationale du Canada

Ex.4 040 $beng                                          n50057296
     110  2#$aDeutsche Bucherei (Germany)
     410  2#$aGermany (East).$bDeutsche Bucherei
     410  2#$aGermany (East).$bGerman National Library
     410  2#$aLeipzig.$bDeutsche Bucherei
     410  2#$aGerman National Library
     410  2#$aBibliotheque allemande (Germany)
     410  2#$aNemetskaia biblioteka (Germany)

Ex.5 040 $beng                                          n80020373
     110  1#$aSoviet Union.$bPosolstvo (Egypt).$bMaktab al-sihafal
     410  1#$aRussia (1923-  U.S.S.R.).$bPosolstvo
          (Egypt).$bMaktab al-sihafal.

Ex.6 040 $beng                                          n79076156
     151  ##$aMoscow (Russia)
     451  ##$aMoskva (Russia)
     451  ##$aMoscou (Russia)
     451  ##$aMoskau (Russia)
     451  ##$aMoscova (Russia)
     451  ##$aMo-ssu-k'o (Russia)

Ex.7 040 $beng                                          n80125931
     151  ##$aGermany
     451  ##$aDeutschland
     451  ##$aFederal Republic of Germany
     451  ##$aBundesrepublik Deutschland
     451  ##$aGermanyah

Model B:

Each record is constructed around one heading for an entity (1XX
uniform heading) and that heading is appropriate for a catalog in
the language designated at the record level by the 040$b field. 
The 4XX and 5XX traced headings constitute the reference structure
for the 1XX heading in that catalog.

For specified languages, however, a cataloging agency constructs
parallel records that contain parallel 1XX uniform headings that
differ in language of the catalog into which they fit, e.g.,
English catalog and French catalog. (If the heading and its
reference structure would be the same for the different language
catalogs, then one record may serve, but that is an implementation
decision. For the model there are parallel records.)  These
parallel uniform headings have separate authority entry records in
which they are the 1XX heading and where their reference structures
and other information are recorded in the 4XX and 5XX reference
tracing and 6XX note fields.

The parallel language records may be linked through the 7XX record
linking fields.  In each authority entry record, the parallel 1XX
headings (and their associated record numbers (subfield $0)) may be
recorded in 7XX linking fields.

Ex.1 040 $beng$bger [Note: one record for both catalogs)
     100  1#$aMarx, Karl,$d1818-1883
     400  1#$aMarukusu,$d1818-1883
     400  1#$aMarx, Carlos,$d1818-1883
     400  1#$aMarks, Karl,$d1818-1883
     400  1#$aMarks, Karol,$d1818-1883
     400  1#$aMarksas, Karolis$d1818-1883
     400  1#$aMac, Cac$d1818-1883

Ex.2 040 $beng
     110  2#$aNational Library of Canada
          <Tracings and notes for an English language catalog>
     710  25$aBibliotheque nationale du Canada

     040 $bfre
     110  2#$aBibliotheque nationale du Canada
          <Tracings and notes for a French language catalog>
     710  26$aNational Library of Canada

Ex.3 040 $beng
     110  2#$aDeutsche Bucherei (Germany)
     410  2#$aGermany (East).$bDeutsche Bucherei
     410  2#$aGermany (East).$bGerman National Library
     410  2#$aGerman National Library
     410  2#$aBibliotheque allemande (Germany)
     410  2#$aNemetskaia biblioteka (Germany)
     710  24$aDeutsche Bucherei (Deutschland)

     040 $bger
     110  2#$aDeutsche Bucherei (Deutschland)
     410  2#$aLeipzig.$bDeutsche Bucherei
     410  2#$aDeutschland (Ost).$bDeutsche Bucherei
     410  2#$aBibliotheque allemande (Deutschland)
     410  2#$aNemetskaia biblioteka (Deutschland)
     710  20$aDeutsche Bucherei (Germany)

Ex.4 040 $beng
     151  ##$aMoscow (Russia)
     451  ##$aMoskva (Russia)
     451  ##$aMoscou (Russia)
     451  ##$aMoskau (Russia)
     451  ##$aMoscova (Russia)
     451  ##$aMo-ssu-k'o (Russia)
     751  #4$aMoscou (Russie)

     040 $bfre
     151  ##$aMoscou (Russie)
     451  ##$aMoskva (Russie)
     451  ##$aMoscow (Russie)
     451  ##$aMoskau (Russie)
     451  ##$aMoscova (Russie)
     451  ##$aMo-ssu-k'o (Russie)
     751  #0$aMoscow (Russia)

Ex.5 040 $beng
     151  ##$aGermany
     451  ##$aFederal Republic of Germany
     451  ##$aGermanyah
     751  #4$aDeutschland

     040 $bger
     151  ##$aDeutschland
     451  ##$aBundesrepublik Deutschland
     451  ##$aGermanyah
     751  #0$aGermany

Note that even when these parallel uniform headings are in a
different script, in addition to being in a different language of
catalog, the above models can be used.


Language of Heading

The above discusses language in terms of language of catalog,
rather than identifying the actual language or languages found in
the heading field.  The two may differ because of rules that
prescribe use of the original language for certain headings.  The
language of the catalog, not the heading, establishes the
foundation for the accompanying reference structure.

The language of the heading is not currently being specified in
MARC.  In 1985, LC's Office of Descriptive Cataloging Policy tried
to formulate guidelines for coding the language of heading
positions that then existed in the 008 field, but abandoned the
task with the following comment:
     At the time of the development of the format, even though
     this code was finally included, we wondered about its
     utility, particularly in view of non-uniform application. 
     In formulating the guidelines, we find the initial
     pessimism reinforced.  Given the difficulties with
     ambiguous cases that are bound to arise, no matter the
     guidelines, we cannot recommend an application of the
     code to anyone.
As a result the 008 positions were made obsolete in the format in
1986.   Some of the problems stem from the fact that many headings
are mixed language as the following examples from an English
catalog illustrate, and many others are either not language based
or language of heading is ambiguous.  Examples:
     Ludovicus IV, Emperor of Germany, 1287-1347   [Latin and
     English]
     Deutsche Bibliothek (Frankfurt am Main, Germany)   [German and
     English]
     Siege d'Orleans (Mystery play)   [French and English]
     Nabokov, Vladimir  [English or Russian?]
The significant information about these headings is the language of
the catalog into which they fit, not the languages that are used in
their formulation.

But difficulties aside, some librarians believe that it may be
useful to also indicate at the heading level the language(s) of a
heading.  A repeating subfield $7 could be used to indicate the
languages used to formulate each heading.



Model A:

Ex.1 040 $beng
     100  10$aMarx, Karl,$d1818-1883$7eng$7ger
     400  10$aMarukusu,$d1818-1883$7jpn
     400  10$aMarx, Carlos,$d1818-1883$7spa
     400  10$aMarks, Karl,$d1818-1883$7rus
     400  10$aMarks, Karol,$d1818-1883$7pol
     400  10$aMarksas, Karolis$d1818-1883$7lit
     400  10$aMac, Cac$d1818-1883$7vie

Ex.2 040 $beng
     110  2#$aNational Library of Canada$7eng
     410  20$aBibliotheque nationale du Canada$7fre

Ex.3 040 $beng
     110  2#$aDeutsche Bucherei (Germany)$7ger$7eng
     410  20$aGermany (East).$bDeutsche Bucherei$7ger$7eng
     410  20$aGermany (East).$bGerman National Library$7eng
     410  20$aLeipzig.$bDeutsche Bucherei$7ger$7eng
     410  20$aGerman National Library$7eng
     410  20$aBibliotheque allemande (Germany)$7fre$7eng
     410  20$aNemetskaia biblioteka (Germany)$7rus$7eng

Ex.4 040 $beng
     110  10$aSoviet Union.$bPosolstvo (Egypt).$bMaktab al-
          sihafal$7eng$7rus$7ara
     410  10$aRussia (1923-  U.S.S.R.).$bPosolstvo
          (Egypt).$bMaktab al-sihafal.$7eng$7rus$7ara

Ex.5 040 $beng
     151  ##$aMoscow (Russia)$7eng
     451  ##$aMoskva (Russia)$7eng
     451  ##$aMoscou (Russia)$7fre$7eng
     451  ##$aMoskau (Russia)$7ger$7eng
     451  ##$aMoscova (Russia)$7lit$7eng
     451  ##$aMo-ssu-k'o (Russia)$7chi$7eng

Ex.6 040 $beng
     151  ##$aGermany$7eng
     451  ##$aDeutschland$7ger
     451  ##$aFederal Republic of Germany$7eng
     451  ##$aBundesrepublik Deutschland$7ger
     451  ##$aGermanyah$7heb

Model B:
Model B records would similarly use $7 in heading fields to
indicate language of heading.

Script

Some agencies need to record headings, tracings, and notes in more
than one script (1) because of different orthographies used for a
language (e.g., kana and kanji scripts for Japanese; devanagari,
khmer, and lao scripts for Pali) and (2) because transliteration of
some or all data in a record is needed to support limitations on
automated systems or telecommunications systems and/or needs of
users.  

Alternative graphic representations of the headings, reference
tracings, and notes may be co-resident in an authority record
(Model A).  Logically Model B could also be used, with one script
for each linked record, although Model B is more appropriate when
the different script representations of the 1XX uniform heading
differ also in language of catalog.  In that case the data is
treated like the models for alternative language of catalog data.

Model A:

When the alternative script representations of headings are
co-resident in a record, the alternative graphic forms of the data
fields are currently recorded in 880 fields, linked to their
corresponding fields via the $6 subfield, which also identifies the
main script in the 880 field.

     $6<linking tag>-<occurrence number>/<script code>/<field
     orientation code>

For records produced in the United States, this means the 880
fields contain most of the non-Latin characters in a record.

     066  $alatin$bextended latin$ccyrillic
     100  1#$6880-01$aZemtsovskii, I. I.$q(Izalii Iosifovich)
     400  1#$6880-02$aZemtsovskii, Izalii Iosifovich
     400  1#$aZemtsovskiy, I.
     880  1#$6400-02$a<name in Cyrillic>
     880  1#$6100-01/cyrillic$a<name in Cyrillic with
          initials>$q(<qualifier in Cyrillic>)

[Note that in this document script is designated in field 066 and
subfield $6 by name rather than the proper codes that are specified
in MARC, e.g., in MARC Cyrillic would be indicated as "(N".]

The $6 subfield is inserted in the fields to link the corresponding
fields when the alternate graphic field is a pure transliteration
of an associated field.  If the non-Latin field in an 880 does not
have a transliteration relationship with another field, the
subfield $6 in field 880 carries the imputed tag for the data, but
the link number is 0, indicting it does not link to an associated
field.  In field 880, following the slash, the code for one of the
character sets used in the field is indicated.  If more than one
non-Latin set is used, the others are not indicated in the $6.  The
indication of script in subfield $6 is used for unspecified
purposes by the application programs.  The information is redundant
since the same code in a character set escape code sequence, e.g.,
esc)N, appears before every series of non-Latin characters in the
field and incomplete since only the first set is specified.  These
escape codes are used by the machine to process the characters in
the proper script, not the script code in the $6.  

In the future, with the Universal Character Set (USC) character
set, escape codes will not be needed for non-Latin characters. 
There may or may not be a need to mark a field at the subfield
level as containing other than the default script.  If there were,
there are several schemas for identifying repertoires of characters
that could be considered.

The above has been stated in terms of a record in an environment
that is based on the Latin script.  In that setting the regular
tagged fields are Latin and the 880 fields contain data in other
scripts.  The 880 has been useful for systems that could not handle
alternate script data as it allows quick elimination of non-Latin
data.  Since the format is increasingly being used worldwide, the
880s present problems for some format users with multiscript and
non-Latin script environments.  

Model A without field 880:

With the introduction and use of the UCS more systems are going to
be able to handle data in the original script and slowly the
transliteration of some parts of records will be phased out.  LC,
for example, provides much more transliterated data in online
records for non-Latin scripts than it did when GPO was printing
vernacular cards.  LC expects to increase inclusion of vernacular
in records when it has vernacular capability in its OPAC.  As a
result more unlinked data will be carried in the 880 fields instead
of the regular tagged fields.  Consideration should be given,
therefore, to normalizing the treatment of alternative graphic
information and using repeating regular tags rather than 880
fields.  With the establishment of the $8 subfield as the
"universal" field-to-field link, the use of $6 for linking
information will become increasingly confusing.  A more logical way
to handle the alternate script data in the same record might be to
use the $8 to link alternate graphic representations when needed
and the $6 to indicate explicitly the script(s) in the field, if
necessary.

     066  $alatin$bextended latin$ccyrillic
     100  1#$801\g$aZemtsovskii, I. I.$q(Izalii Iosifovich)
     100  1#$801\g$6/cyrillic$a<name in Cyrillic with
          initials>$q(<qualifier in Cyrillic>)
     400  1#$802\g$aZemtsovskii, Izalii Iosifovich
     400  1#$aZemtsovskiy, I.
     400  1#$802\g$6/cyrillic$a<name in Cyrillic>

An alternative would be to code the alternate graphic form of the
1XX as a 4XX:

     066  $alatin$bextended latin$ccyrillic
     100  1#$801\g$aZemtsovskii, I. I.$q(Izalii Iosifovich)
     400  1#$801\g$6/cyrillic$a<name in Cyrillic with
          initials>$q(<qualifier in Cyrillic>)
     400  1#$802\g$aZemtsovskii, Izalii Iosifovich
     400  1#$aZemtsovskiy, I.
     400  1#$802\g$6/cyrillic$a<name in Cyrillic>

Subfield $8 contains the link number and field link type code;
subfield $6 indicates Cyrillic script.  It is interesting to note
that virtually all systems that support input of multiscript MARC
records, use repeating regular tags rather than 880 fields for
tagged displays.

Model B:  

Technically, Model B could also be used for alternate scripts.  The
separate records for alternative script representations would
greatly limit use of the 880 field since non-Latin records would be
built around the non-Latin script, declared in the 066 (Character
Sets Present) field.  There would be special complexity with
parallel language of catalog records if separate records were used. 
In that case Model B might be used for the parallel heading data
and Model A used within each parallel language record for alternate
script data.

Model B without field 880:

If the 880 technique were abandoned, then Model B for alternative
graphics would follow the outline of Model B for languages. 
Separate records would be created for the scripts with the two
linked through the 7XX linking heading fields which would contain
the alternative script form of the 1XX field.

     066  $alatin$bextended latin$ccyrillic
     100  1#$aZemtsovskii, I. I.$q(Izalii Iosifovich)
     400  1#$aZemtsovskii, Izalii Iosifovich
     400  1#$aZemtsovskiy, I.
     700  1#$6/cyrillic$a<name in Cyrillic with
          initials>$q(<qualifier in Cyrillic>)

     066  $acyrillic$bextended cyrillic$clatin
     100  1#$a<name in Cyrillic with initials>$q(<qualifier in
          Cyrillic>)
     400  1#$a<name in Cyrillic>
     700  1#$6/latin$aZemtsovskii, I. I.$q(Izalii Iosifovich)


Transliteration

Some agencies need to provide transliterations for headings carried
in authority records, especially the 1XX uniform heading, if its
native script is not that of the catalog.  Some of the reasons for
inclusion of transliterated headings are that the transliterated
form of a name appears on an item being cataloged, or because the
cataloging practice is to provide transliterations of some or all
of the data in a record.  The latter may be needed in order to
provide a complete alphabetic sorting of entries for displays.  It
may also be needed for systems and communications clients and
servers in which only the local script is supported.  In countries
where the Latin script is the vernacular, for example, many systems
and programs and much equipment may not accommodate non-Latin
scripts, so romanization is used.

Model A:

Currently the transliterated forms of headings are recorded in the
regular data fields as 1XX, 4XX, and 5XX headings.  The
transliteration scheme for the 1XX uniform heading can be indicated
at a general level (international standard, national standard,
local standard, conventional, etc.) in field 008.  The particular
transliteration (e.g., Wade-Giles or Pinyin) cannot be indicated
for the 1XX, nor can transliteration be indicated for the 4XX and
5XX headings.  

Different cataloging agencies may prefer one transliteration over
another as the standard for their catalog or a decision might be
made to change from one scheme for the uniform headings to another. 
Both of these situations would benefit from designation of
transliteration scheme used at the heading level.  For example,
changing from Wade-Giles to Pinyin has been under discussion for
many years.  In the NACO file, for Chinese headings the 1XX uniform
heading is transliterated using Wade-Giles and one of the 4XX
fields contains the Pinyin transliteration of the heading, but the
Pinyin heading is not identified among the 4XX fields. 

One technique for identifying transliteration would be through the
$6 subfield.  The $6 subfield could be extended:

     $6<linking tag>-<occurrence number>/<script code>/<field
     orientation code>/<transliteration scheme>

In the absence of a need for the first, second, or third items, the
subfield would contain only the transliteration scheme:
          ///<transliteration scheme>

     008/7  b  [national standard]
     040 $beng
     110  10$aSoviet Union.$bPosolstvo (Egypt).$bMaktab al-sihafal
     410  10$aRussia (1923-  U.S.S.R.).$bPosolstvo (Egypt).$bMaktab
          al-sihafal.$6///<translit. scheme for
          Cyrillic>$6///<translit. scheme for Arabic>

     008/7  b? or g?     [national standard? conventional?]
     040 $beng
     151  ##$aMoscow (Russia)
     451  ##$aMoskva (Russia)$6///<translit. scheme for Cyrillic>
     451  ##$aMoscou (Russia)$6///<thanslit. scheme for Cyrillic>
     451  ##$aMoskau (Russia)$6///<translit. scheme for Cyrillic>
     451  ##$aMoscova (Russia)$6///<translit. scheme for Cyrillic>
     451  ##$aMo-ssu-k'o (Russia)$6///<translit. scheme for
          Cyrillic>

The level of scheme indication would needed to be worked out. 
There are many - perhaps hundreds - of  transliteration schemes,
besides the ALA tables standardly used in US libraries.  Going to
the specific level would be difficult, but the potential usefulness
of the general level as it is now specified in the 008, needs to be
examined.

Model B:

Using separate records for transliteration would be a variant of
using deparate records for script.


Country or Nationality Data

Some cataloging agencies, particularly outside the US, want to be
able to record the nationality of an entity so that a specific
country's authors can be identified.  The US practice has never
advocated this because the information would be difficult to
determine in many cases, increasing the cost of cataloging, and not
seen as particularly valuable for retrieval.  If nationality were
carried in the record it could be at the record level thus would
entail definition of 008 bytes and/or a variable field (if multiple
nationalities needed to be recorded).


Catalog Rules

A final characteristic, that is consistently coded in MARC already,
is the identification of the cataloging rules used to create the
authority record.  Field 008/10 and 11 identify the rules for names
and subjects, with the possibility to record the information in 040
$e or $f if the rules are not specified in the codes defined for
008/10 and 11.  


Subjects and Classification

Subject headings would have many of the same characteristics as
names.  Some are in a non-English language but intended for an
English language catalog (e.g., Singspiele, Ojo de Dios (Talisman),
and Commedia dell'arte).  Some are transliterations (e.g., Haiku,
Waka, and al-Jum'ah).  With subjects there is the need to link
different thesauri that are in the same language (e.g., LCSH and
MESH) and in different languages (e.g., Canadian Subject Headings,
LCSH, and Repertoire des vedettes-matiere).  There are also script
and transliteration needs (e.g., Savoskin, Anatolij
Nikolajev$vBibliography and the multiscript translations of Dewey).


SUMMARY

The above discussion has suggested that the format be enabled to
indicate language of catalog, language of heading, and
transliteration at the field level and to begin using direct
tagging of alternate script data.  The indication of script
explicitly at the field level also needs to be evaluated along with
the codes that would be used.  The new and augmented elements to
enable this would be the following:

   - Use subfield $6 for script, field orientation, and
     transliteration by adding an indication of transliteration
     scheme in the following upward-compatible manner:

          $6<linking tag>-<occurrence number>/<script code>
          /<field orientation code>/<transliteration scheme>

     Note that the linking tag and occurrence number would no
     longer be needed in $6 except when the 880 is being used.  If
     use of 880 is made obsolete, then the linking tag and
     occurrence number would also be, but the slash would probably
     need to be retained.

   - Define subfield $7 for language of heading at the field level

   - Use subfield $8 for linking alternate script in directly
     tagged fields, with code "g" defined for that linking use. 
     The $8 data would be:
          $8<linking number>\g

   - Make field 880 obsolete.

An IFLA Working Group on Authority data elements feel that adding
the above new data elements to records would result in reduced cost
of cataloging, since authority records could be shared and the same
data elements found in all.  They see field level information on
language of heading, script and transliteration scheme as
especially important but also support inclusion of nationality.


QUESTIONS

- Are systems likely to use the new data elements?  
- Would the NACO program endorse coding language of heading,
transliteration, etc.?
- What is the impact on utilities that support distribution using
the 880 field?  Would this benefit their international services? in
the long run?
- What would be the impact on local systems?  
- How many libraries are currently loading the vernacular that they
receive from RLIN and OCLC?
- Would authority records from other countries be more useful if
they had the new data encoded?
- Ceasing to use the 880 would be applied to other formats,
especially bibliographic.  Are there differences to note there?


Go to:


Library of Congress
Library of Congress Help Desk (09/03/98)