DISCUSSION PAPER NO. 92

DATE: December 1, 1995
REVISED:

NAME: Change in Definition of Computer File in Leader/06 (Type of record) in the USMARC Bibliographic Format

SOURCE: Library of Congress

SUMMARY: This paper explores issues on coding MARC records for computer files in Leader/06. Currently code "m" is defined broadly as any body of information encoded in a manner which allows it to be processed by a computer. It is suggested that the definition be changed so that code "m" is used only for executable software. If this were implemented, when an item has two or more attributes, one of which is computer file, the record would be coded for the characteristic other than the computer file aspect. Questions are raised concerning the use of the Leader/06 after format integration with the availability of 006.

KEYWORDS: Leader/06 (Bibliographic); Type of record; Computer files

RELATED: 95-9 (June 1995)

STATUS/COMMENTS:

12/1/95 - Forwarded to USMARC Advisory Group for discussion at the Midwinter 1995 MARBI meetings.

1/20/96 - Results of USMARC Advisory Group discussion - Participants listed the various uses of the Leader/06 code: to separate databases based on form of material; to determine workform displays for keying the 008; sorting records; validation of fixed fields; support boolean searching; matching for duplicate detection; display of labels to identify fields; selecting subsets for products; generating icons to show format when searching multiple databases. There was general agreement that in cases where the content of the material is clear, that identifying the primary record type in the Leader/06 by its content rather than carrier better served users. These cases include electronic text, music CDs, digital maps, digital photographs, etc. However, the general feeling was that to define code "m" as only executable software was too restrictive for various reasons: the growing existence of hybrids, which include pictures, graphics, text, software; files that don't fit into a category, e.g. survey data; defining "m" only as executable software would not allow input of an 006 for computer file characteristics for electronic text, since its secondary characteristic is not an executable. It is likely that each constituency will need to issue guidelines. A new discussion paper will be prepared for the annual conference MARBI meetings to further explore this issue.


DISCUSSION PAPER NO. 92: Change in Definition of Computer File 
in Leader/06 (Type of record) 

1.    INTRODUCTION

      The last phase of Format Integration will include all changes
to the Leader through 008 elements in the MARC bibliographic
format.  This includes field 006, which enables a user to supply
coding for characteristics of a secondary form of material. 
Leader/06 (Type of record) contains a code that is used to
determine what type of 008 (Fixed-Length Data Elements) is included
in the record; the 008 character positions vary in 008/18-34
depending upon the type of material as coded in the Leader.  It
also may be used in some systems for record retrieval.  Field 006
will include applicable codes that would otherwise be coded in
008/18-34, so that additional information may be given for other
additional aspects of the item.

      The document Format Integration and its Effect on the USMARC
Bibliographic Format included a model for coding records after the
implementation of format integration.  It divided the cases into an
item that is basically textual and an item that is not basically
textual (although it did not specifically define what "basically
textual" means).  For the former, it instructed that the Leader/06
be coded as language material or mixed material.  For nontextual
items, the record can be coded for any of the other types of
material, but no preference was given when there are
characteristics of two nontextual forms.

      The definition of computer file in Leader/06 of the USMARC
bibliographic format currently reads as follows:

      m - Computer file
      Code m indicates that the content of the record is for a body
      of information encoded in a manner which allows it to be
      processed by a computer.  The information in the computer file
      may be numeric or textual data, computer software, or a
      combination of these types.  Although a file may be stored on
      a variety of media (such as magnetic tape or disk, punched
      cards, or optical character recognition font documents), the
      file itself is independent of the medium on which it is
      stored. 

      Planning for Format Integration began in the late 1980s when
the ability to digitize was not what it is now.  Thus, it is
necessary to reconsider how useful it is to designate an item as a
computer file when it might be desirable to emphasize its content
type.    The question arises as to whether it serves users well to
consider all digitized items to be computer files and assign them
code "m" in Leader/06, or whether the definition should be limited
in some way.  In addition, the record for a computerized version
of, for example, an original photograph will be coded differently
than the record for the original (if separate records are created). 
This may cause problems for retrieval, particularly in systems that
separate records by form of material.  Also, because of economic
considerations, many users are adding information about the digital
item on the MARC record for the original, rather than creating a
separate record.

      Proposal No. 95-9 (Encoding of Digital Maps in the USMARC
Bibliographic Format) was considered by the USMARC Advisory Group
in June 1995.  It proposed renaming code "e" in Leader/06 from
"Printed map" to "Cartographic material" so that all maps, whether
digital or print, could be coded the same (there is also a code for
"manuscript map").  Because of the increasing number of digital map
images becoming available (resulting partly from digital library
projects and the _Content Standards for Geospatial Metadata_), this
change was considered necessary for the map community.  In many
cases the bibliographic record for the paper copy will contain
information about the location of the digital image.  This paper
brought up the issue of coding for content rather than for physical
carrier.  Although the portion of the proposal concerning Leader/06
was approved, it was suggested that a broader discussion paper be
presented.


2.    DEFINITION OF COMPUTER FILE

      Because almost anything can be digitized and called a computer
file, it may be necessary to limit the definition given in the
Leader/06.  If only executable software were included in the
definition of computer file, it might read as follows:

      m - Computer file
      Code m indicates that the content of the record is computer
      software. Any other type of computer file is coded for the
      form of material it is when processed for display.  Databases
      are coded for their intent, in most cases as language
      material.  

If the definition were changed, any electronic textual material
(e.g. textual electronic serials) would be coded as "a" for
language material.

      Coding for the content of the item for digitized materials is
consistent with the method used for handling microforms.  The
USMARC bibliographic format in Leader/06 says the following:

      "Microforms, whether original or reproductions, are not
      identified by a distinctive Type of record code.  The type of
      material characteristics described by the codes take
      precedence over the microform characteristics of the item."

      Digitized materials can also be either originals or
reproductions of originals.  This method of handling them as
language, cartographic, music, etc. would ensure that records for
digital reproductions are not separated from the originals, and
allows flexibility for record creation (i.e., using one record and
adding a field 856 with location information of the digital
reproduction or creating a separate record).  The statement above
about the treatment of microforms could be revised to include
digitized materials.

      Additional information about the computer file aspects can be
given in an 006 for computer file fixed field data and in 007 for
physical description.  A general material designator (GMD) would be
given in 245$h to indicate that the physical format is computer
file; it is not necessary for the 008 to agree with the GMD, so the
008 for the content of the item would be given.

      After format integration, since all variable fields are
available for all types of material, the Leader/06 code no longer
determines field validity.  Once the electronic aspects are moved
from the Leader/06, then the format issues can be divorced from the
cataloging rules.  The cataloger still needs to choose which
chapter of AACR2 to use, which will determine, among other things,
which fields are needed in the record.  The cataloger is no longer
bound by the format, since any USMARC defined fields will be valid.


3.    QUESTIONS

Consideration needs to be given to the following questions.

1.  How is the Leader/06 byte currently being used (e.g.,
filtering, file sorting, etc.)?

2.  Do users need a dominant form specified (as it is in choosing
the Leader/06 code?  Or will the availability of 006 after format
integration obviate this need?  Since more than one set of 008
elements can now be given, does it matter what the Leader/06 code
is?

3.  If the change in the definition of computer file were
implemented, how would this affect retrieval of records?  If the
change is not approved, how would using the Leader/06 for computer
file for all digital items affect retrieval?

4.  How many users split their files according to form of material? 
How would this change affect record selection for different files?

5.  How important is it that the choice of Leader/06 be mandated,
or could users emphasize the aspect they want to depending upon
individual needs? Should there be options in determining the code
in Leader/06?

6.  Is further clarification needed for nontextual materials having
two or more attributes when one is not computer file (e.g. music
for America, the Beautiful on a wall chart)?

7.  Does the GMD and the existence of a computer file 006 suffice
to determine what is the actual physical fors of the item?  For
instance, if the Leader/06 were coded "e" for cartographic material
and there is a field 006 for computer file, could one determine
whether this record represents a paper map with accompanying
computer disk or a computer file that displays maps? Note that the
GMD is not a required data element.  Does an additional data
element need to be defined (perhaps in the Leader) for carrier?


See Attachment A for LC examples using one record for the original 
and the digitized reproduction.  Leader/06 represents the content 
type.

-----------------------------------------------------------------

                                ATTACHMENT A
EXAMPLE 1: LC bibliographic record (visual materials) using one
record.  Description for original, 856 for digital reproduction.

Leader  *****nkc##22*****#a#4500 
001   $a94837685 $PP
008   950619i18601960xxunnn#|||||||###|kneng||
035   $a(CStRLIN)DCLC94837685-F
050   00$aGuide Record 
245   00$aSelected Civil War Photographs, 1861-1865 (Library of
      Congress)$h[graphic].
246   33$aCivil War Photograph Collection (Library of Congress)
260   ##$cca. 1860-1960, bulk 1861-1865.
300   ##$a1,114 photographs +$e4 illustrations.
520   0$aImages depict military personnel and facilities, primarily
      from a Union perspective. Includes the main Eastern theater,
      the federal navy and seab orne expeditions against the
      Atlantic Coast of the Confederacy, the war in the West,
      Washington, D.C., African Americans, fortifications,
      battlefields, preparations for battle and the aftermath of
      battle. Also, many portraits of officers and enlisted men, and
      of federal and confederate government members, including
      Abraham Lincoln and his assassins.
500   ##$aCollection title devised.
500   ##$aMost photographs were taken during the American Civil War
      under the supervision of Mathew B. Brady. Photographers
      represented by more than 20 images include George N. Barnard,
      Alexander Gardner, James Gibson, Timothy H. O'Sullivan, and
      William Morris Smith.
500   ##$aIncludes 1,048 wet collodion glass negatives, 8 x 10 in.
      or smaller, and 66 modern black-and-white photographic prints,
      8 x 10 in. or smaller.  The prints are copies made by the
      Library ca. 1960 of ambrotypes, tintypes, and other portraits
      of enlisted men in private collections.
510   3#$3Selections from Anthony-Taylor-Rand-Ordway-Eaton
      Collection$aCivil War photographs, 1861-1865 / Library of
      Congress. Prints and Photographs Division. Washington, D.C. :
      Library of Congress, 1961, reprinted 1977.
530   ##$aAvailable on the Internet;$bLibrary of Congress World Wide
      Web site.
530   ##$3Selections from Anthony-Taylor-Ordway-Eaton
      Collection$aAvailable on 2 reels of microfilm;$bPublished as
      Civil War Photographs, 1861-1865 (Library of Congress
      Photoduplication Service, 1961)
530   ##$aAvailable on videodisc;$bLibrary of Congress, American
      Memory Program, 1994, unpublished.
530   ##$aCD-ROM;$bAustin, Tex. : Stokes Imaging Services, 1994.
540   ##$aNo known restrictions on reproduction. For information see
      LC P&P Restrictions Notebook.
545   ##$aThe bulk of the photographs were selected from the
      Anthony-Taylor-Rand-Ordway-Eaton Collection by the Library of
      Congress for inclusion in a microfilm in 1961. Additional
      images of enlisted men were selected from various sources by
      the Library's American Memory project.
555   8#$aIn the electronic finding aid, each image is described
      with full caption information and subject indexing. Electronic
      reproductions of each image and a narrative introduction
      accompany the records in the Library's Internet (World Wide
      Web) catalog.
555   8#$aIn the printed finding aid, 1,047 of the images have
      captions in a booklet that accompanies a published microfilm
      with reproductions of negatives.
650   #7$aBattlefields$zUnited States$y1860-1870.$2lctgm
650   #7$aForts & fortifications$zUnited States$y1860-1870.$2lctgm
650   #7$aMilitary officers$zUnited States$y1860-1870.$2lctgm
651   #0$aUnited States$xHistory$yCivil War, 1861-1865.
651   #0$aWashington (D.C.)$xHistory$yCivil War, 1861-1865.
655   #7$aPortrait photographs$y1850-1890.$2gmgpc
700   1#$aBarnard, George N.,$d1819-1902,$ephotographer.
700   1#$aBrady, Mathew B.,$d1823 (ca.)-1896,$ephotographer.
700   1#$aGardner, Alexander,$d1821-1882,$ephotographer.
700   1#$aGibson, James,$db. 1828,$ephotographer.
700   1#$aO'Sullivan, Timothy H.,$d1840-1882,$ephotographer.
700   1#$aSmith, William Morris,$ephotographer.
755   ##$aWet plate negatives$y1860-1870.$2gmgpc
755   ##$aPhotographic prints$y1860-1870.$2gmgpc
852   ##$aLibrary of Congress$bPrints and Photographs
      Division$eWashington, D.C. 20540-4840 USA$ndcu
856   7#$3finding aid$dcwp$nLibrary of Congress, Washington,
      D.C.$2file
040   ##$aDLC$cDLC$dDLC$egihc$dCStRLIN
043   ##$an-us---
952   ##$aDCLC94837685F$bLibrary of Congress
005   $a19951018074651.2
[Note that this record is coded in Leader/06 as "k" for Two-
dimensional nonprojectable graphic. Field 856 is used to indicate
electronic location.  In this case, LC has chosen to use a local
implementation (indicated by "file" as the access method in the
URL).  The data in subfield $d and $f provide a unique name
(similar to the function of a URN), which allows LC's local
software to use a locator table to determine the host, software to
display the file, and other necessary information. "CWP" (data in
subfield $d) is the aggregate name for Civil war photographs.]

-------------------------------------------------------------------

EXAMPLE 2:  LC bibliographic record (books record)
Separate record not created for digitized version.

Leader *****nam##22*****#a#4500
001   $a92000507$cr95
008   920427s1993####dcua#####b###f##x###eng##
020   ##$a0844407399
040   ##$aDLC$cDLC$dDLC
043   ##$af-et---
050   00$aDT373$b.E83 1993
082   00$a963$220
245   00$aEthiopia, a country study /$cedited by Thomas P. Ofcansky
      and LaVerle Berry.
250   ##$a4th ed.
260   ##$AWashington, D.C. :$bFederal Research Division, Library of
      Congress :$bFor sale by the Supt. of Docs., U.S.
      G.P.O.,$c1993.
300   ##$Axxxvi, 412 p. :$bill. ;$c24 cm.
440   #0$aArea handbook series,$x1057-5294
440   #0$aDA pam ;$x550-28
500   ##$aSupersedes the 1981 edition of Ethiopia: a country study,
      edited by Harold D. Nelson and Irving Kaplan.
500   ##$a"Research completed July 1991".
504   ##$aIncludes bibliographical references (p. 347- 376) and
      index.
651   #0$aEthiopia.
700   10$aOfcansky, Thomas P.,$d1947-
700   10$aBerry, LaVerle Bennette,$d1942-
710   20a$Library of Congress.$bFederal Research Division.
856   7#$3Full text$uhttp://rs6.loc.gov/ethiopia.html$2http


Go to:


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