The Library of Congress >> Especially for Librarians and Archivists >> Standards
HOME >> MARC Development >> Discussion Paper List
DATE: June 1, 2006
REVISED:
NAME: Recording set information for multipart cartographic materials
SOURCE: ALA/MAGERT Holdings Task Force
SUMMARY: This paper explores the historical methods of recording information for multipart cartographic materials, identifies relevant item specific information to record, and proposes several possible ways of recording the data using MARC 21 Bibliographic and/or Holdings Formats. This discussion is intended to lead to the establishment of standards for recording and sharing cartographic set data.
KEYWORDS: Cartographic material (BD, HD); Multipart cartographic materials (BD, HD)
RELATED:
STATUS/COMMENTS:
06/01/2006 - Made available to the MARC community for discussion.
06/24/06 - Results of the MARC Advisory Committee discussion - The participants felt that a new discussion paper should be written that explores other types of materials (For example, archival moving image materials, graphic and still image materials, etc.) to ascertain whether they, like cartographic materials, require more descriptive detail to be added to the MARC record. The participants also felt that field 034 should be added to the holdings format.
In addition to atlases and single sheet maps, most cartographic collections include a category of multipart items traditionally stored together and controlled by one catalog record. Two main types are: the multi-sheet single, a map printed on a number of sheets, usually published simultaneously, and intended to be thought of as a single map; and the map series or set, also printed on a number of sheets, often issued over time with no predictable seriality, with repeated and connective information on each sheet which enables the sheets to be used singly or in conjunction with all or part of the remaining sheets. Some map sets include thousands of sheets, each with unique sheet level data elements.
For convenience, the term "sets" will be used throughout the rest of the paper to refer to both types of multipart cartographic materials.
Traditionally, each institution's set holdings have been recorded manually on a graphic index map, with diagonal lines indicating the presence in the collection of the corresponding sheets. This method is not shareable, nor does it include often useful metadata, such as the names of individual sheets.
Some institutions have created digital graphic index maps to record their holdings (and some have made them available online). While this results in much wider dissemination of information, the ability to do text searching for sheet titles, numbers, etc., as is possible with MARC 21 Bibliographic and Holdings Formats, is also quite desirable and would complement the index map. An ultimate goal would be to have standards for both the digital index maps and the recording of textual data so that a user could access both simultaneously. An example of a digital index map is at: www.libraries.psu.edu/maps/15mintopos/15mintopo.html
This paper, however, will concern itself with the possible ways of recording the textual data, with the idea of leaving such records open to connection with a digital index map, or even providing information (e.g. coordinates) that would facilitate the creation of such a map.
There are a number of possible sheet level data elements that need to be accommodated by or at least considered in the development of any individual sheet map data standard. For some sets, individual sheet maps will have only one or two, others will have all of them.
A full level record can be created for each sheet, giving all data and access points, even if they are already on the record for the set. This is certainly desirable in terms of search and is easily shared but would, in some cases, result in much repeated data, and would make the collective aspect of the set less transparent to the user. It is also hard to see how most institutions could support the amount of cataloging time this would require, probably making this option a viable one primarily for locally important, valuable, or heavily used sets.
Another alternative is to create a record for each set plus brief records for each sheet, with links to the set bibliographic record through 773 fields. This could eliminate some repeated data, as the individual sheet records would have only access points specific to that sheet and some other data elements could be treated likewise. It may be difficult for displays to differentiate the set record from the sheet records, however, requiring more interpretation by the user. Some systems might not be able to accept these brief records, so the set records would be required to be in the database. Although it would probably require less cataloging time than a full level record for each sheet, this approach may not be viable in current systems.
When map sets have been scanned for display on a web site, many institutions have created separate bibliographic records, using one of the above models or another type of possibly abbreviated record.
Example of a full level record for a single sheet of a World War II situation map sets posted on a web site
050 00 $a G5701.S7 svar $b .A4 052 $a 5701 110 2 $a Allied Forces.$bArmy Group, 12th.$b Engineer Section. 245 10 $a [April 25, 1945], HQ Twelfth Army Group situation map. 255 $a Scale 1:1,400,000. 260 $a [England?] : $b Twelfth Army Group,$c [1945] 300 $a 1 map : $b col. ; $c 50 x 49 cm. 530 $a Available also through the Library of Congress web site as a raster image. 650 0 $a World War, 1939-1945 $x Campaigns $v Maps. 710 1 $a United States. $b Army. $b Army Group, 1st. $b Headquarters. 752 $a Germany. 773 0 $a Allied Forces. Army Group, 12th. Engineer Section.$t HQ Twelfth Army Group situation map $w (DLC) 2001628569 856 41 $d g5701s $f ict21015 $u http://hdl.loc.gov/loc.gmd/g5701s.ict21015 $q c
This has the advantage of being full text searchable and shared with the distribution of the bibliographic records. However, there are limits to the amount of data that can be recorded and/or easily read in the bibliographic record and some of these sets have hundreds of sheets. The field does not lend itself to parsing the separate data elements, and therefore coordinates would not be usefully recorded there. If the only sheet level identifiers are numeric, or alphanumeric, this display is also less than ideal.
Example of a contents note for a district set for a state of India without numbering but with sheet titles
505 1 $a Ahmadnagar (North half) -- Ahmadnagar (South half) -- Akola -- Amravati -- Auranga¯ba¯d -- Bhanda¯ra -- Buldana -- Dhule -- Garhchiroli -- Jalgaon -- Jalna -- Kolha¯pur -- La¯tu¯r -- Mumbai city and Mumbai suburban -- Na¯gpur -- Na¯nded (North half) -- Na¯nded (South half) -- Na¯shik -- Parbhani -- Pune -- Raigarh -- Ratna¯giri -- Sa¯ngli -- Sa¯ta¯ra -- Sindhudurg -- Sola¯pur (East half) -- Sola¯pur (West half) -- Tha¯ne -- Wardha -- Yavatmal
Example of a contents note from a Burkina Faso set with numbering and editions
505 0 $a NC-30-XXI, Houndé / 2ème ed., 2002 -- NC-30-XXII, Léo -- NC-30-XXIII, Po -- NC-30-XXIV, Tenkodogo / 2ème ed., 2003 -- NC-31-XIX, Pama -- ND-30-IV, Koudougou -- ND-30-V, Ouagadougou -- ND-30-VI, Boulsa -- ND-30-X, Ouahigouya -- ND-30-XI, Kaya -- ND-30-XII, Pissila -- ND-30-XVII, Djibo -- ND-31-VII, Sebba.
Field 774 (Constituent unit entry) allows for a more detailed type of contents note, parsing data elements into separate subfields. It would be possible to include more of the data elements discussed in section 2 in this field than would fit comfortably in a contents note. Again, the data would be easily searchable and shared with the distribution of the bibliographic record. However, both current system restrictions and readability would make this approach problematic for very large sets, e.g. with hundreds of sheet maps.
Each sheet map would be recorded in a separate field 774 with some or all of the following data elements:
Sheet title 774 $t Title Coordinates 774 $m Material specific details (the definition in the format even specifically mentions coordinates) Edition 774 $b Edition Sheet number 774 $g Relationship information Stock or publisher number 774 $o Other standard identifier ISBN 774 $z ISBN Issuing body 774 $d Publishing place, body, date?
If this approach is considered viable, some changes may need to be made to the subfields. For instance, subfield $t (title) may need to be repeatable in cases where there are parallel or other variant titles, although the impact of making this change needs to be considered.
Example of a linking entry for a sheet in a set of Ethiopia
774 $t Metu $b ed. 1 $g Sheet NC36 $m E (30°--E 36°/N 12°--N 8°)
Currently, most Holdings records are not shared or migrated. As this is in the process of changing for many utilities and institutions, it is worth considering whether or not the MARC Holdings format is appropriate for recording the particular data elements required to identify sheets of a map set. In any case, OPACs should be able to use the data to display holdings in an institution's online catalog.
Many of the elements in section B can be accommodated by using a combination of Holdings fields 844, 853, and 863. Some of the nonserial data elements in MARC 21 Holdings are appropriate for this data. Field 034 (Coded Cartographic Mathematical Data) could be added to the Holdings format to facilitate the use of MARC Holdings records for cartographic materials, since coordinates will be needed at the sheet map level. (Note that Proposal No. 2006-06 proposes adding field 034 to MARC authority records.) According to the MARC 21 Holdings Format documentation, using field 844 and possibly 034 will require a separate Holdings record to be made for each sheet map, in order to link multiple 844s to the appropriate 85X/86X pair. Since 844 field $a (Name of Unit) is not repeatable, it would be useful to be able to repeat the subfield in cases where there are parallel or other variant titles.
Another already defined data element that may be used is 86X subfield $z for Edition (Note that the mapping of Z39.71 data elements to MARC in Appendix D of the MARC 21 Holdings Format uses this subfield for Specific extent note and examples in Z39.71 show edition statements as specific extent note).
Data element mapping using this approach:
Sheet title 844 Sheet number 86X$a-f Coordinates 034 (to be defined) Edition or version 86X$z Issuing body ? ISBN 020 Stock or publisher number 024 Date 86X$i-m Supplements/separate indexes 855/865 Grid structure/number 85X/86X $a-f
There are several disadvantages to this approach:
Example of MARC separate Holdings record for a sheet in a set of Ethiopia
844 Metu 853 $a [grid] $b [no] 863 $a NC $b 36 $z ed. 1
Instead of requiring separate Holdings records for each sheet map, repeatable 85X/86X fields may be used, although some of the data elements are not available in these fields.
Data element mapping using this approach:
Sheet title 86X$o Sheet number 86X$a-f Coordinates Could use 86X$z note? Edition or version 86X$z Issuing body ? ISBN ? Stock or publisher number ? Date 86X$i-m Supplements/separate indexes 855/865 Grid structure/number 85X/86X $a-f
As in the other Holdings record option, there may be a need to develop guidelines for the display of compressed holdings so that they are comprehensible, giving only sheet numbers rather than sheet titles, somewhat analogous to serial volumes.
This option has many of the same disadvantages as the multiple Holdings record approach. In addition there are few subfields left for additional data elements. Another option would be to add a new field to holdings for these kinds of materials, but it also would have similar disadvantages as the other holdings options.
Example of a Holdings record with multiple 86X fields
Note that $o was defined in Jan. 2006 for Type of unit in fields 853/863.
Map with only chronology and title
853 01 $8 1$a(year) $t c. 863 41 $8 1.1 $a 1946 $t 1-4 $o Ashburnham, Mass-NH 863 41 $8 1.2 $a 1950 $t 1-2 $o Ashburnham, Mass-NH 863 41 $8 1.3 $a 1946 $t 1-3 $o Ashby, Mass-NH 863 41 $8 1.4 $a 1950 $t 1-4 $o Ashby, Mass-NH 863 41 $8 1.5 $a 1955 $t 1-4 $o Ashfield [etc.]
Map with numbering and date
853 01 $81 $a(sheet) $i (year) 863 41 $81.1 $a 3054 $i1999 $oLake Nyasa 863 41 $81.2 $a 3055 $i1999 $oBwanje River [etc.]
4.1. Is it desirable to have a number of options for the recording of map set holdings or should this be standardized in the MARC formats or best practices provided?
4.2. Is it preferable to use bibliographic or holdings records to accommodate the data for sheet maps?
4.3. If the bibliographic record is preferable, which approach is better (separate record or 774 field in map set record)? Or should there be both options?
4.4. What are institutions doing in regard to controlling scanned map set images mounted on web sites? Are individual bibliographic records for each sheet being created? Is this necessary for making them accessible?
4.5. Which sheet level data elements, many of which cannot be expressed easily in a bibliographic record for a map set, are essential or important? Are there others which would not provide sufficient added value to be included in a future standard?
4.6. If the MARC Holdings format is used, how should it be adapted for cartographic materials (adding repeatable subfields, adding the 034, etc.)? Is it preferable to create separate Holdings records for each or use one? If the latter, how do we accommodate the needed data elements?
4.7. Is there information that should be considered for inclusion in a bibliographic or holdings record exclusively for the purpose of generating an index map? Or should it be assumed that the software to produce index maps will pull information from other sources?
HOME >> MARC Development >> Discussion Paper List
The Library of Congress >> Especially
for Librarians and Archivists >> Standards ( 09/18/2006 ) |
Contact Us |