The Library of Congress >> Especially
for Librarians and Archivists >> Standards
The Library of Congress >> Especially
for Librarians and Archivists >> Standards
HOME >> MARC Development >> Discussion Paper List
DATE: June 6, 2007
REVISED:
NAME: Representation of the Dewey Decimal Classification (DDC) System in MARC 21 Formats
SOURCE: Dewey editorial team, Deutsche Nationalbibliothek, LC, and OCLC
SUMMARY: This paper discusses the provisions for the DDC across the MARC 21 format, and proposes changes in order to improve representation of classification information for the purposes of data communication, application, and retrieval.
KEYWORDS: DDC; Dewey Decimal Classification; Subfield $z, in field 082 (BD); Field 082 (BD); Dewey Decimal Classification Number (AD, BD); Subfield $y, in field 153 (CD); Subfield $y, in field 253 (CD); Subfield $y, in field 353 (CD); Subfield $y, in field 453 (CD); Subfield $y, in field 553 (CD); Subfield $y, in field 680 (CD); Subfield $y, in field 683 (CD); Subfield $y, in field 684 (CD); Subfield $y, in field 685 (CD); Subfield $y, in field 761 (CD); Subfield $y, in field 765 (CD); Subfield $y, in field 082 (BD); Subfield $y, in field 083 (AD); Internal add table identification (CD); Field 153 (CD); Field 253 (CD); Field 353 (CD); Field 453 (CD); Field 553 (CD); Field 680 (CD); Field 683 (CD); Field 684 (CD); Field 685 (CD); Field 761 (CD); Field 765 (CD); Field 083 (AD); Classification Number (CD); Complex See Reference (CD); Complex See Also Reference (CD); Invalid Number Tracing (CD); Valid Number Tracing (CD); Scope Note (CD); Application Instruction Note (CD); Auxiliary Instruction Note (CD); History Note (CD); Add Or Divide Like Instructions (CD); Synthesized Number Components (CD); Field 085 (BD); Synthesized DDC Number Components (BD); Subfield $m, in field 085 (BD); Subfield $2, in field 085 (BD); Subfield $5, in field 085 (BD); Subfield $5, in field 082 (BD); Subfield $5, in field 084 (CD); Assigning agency of a Dewey number (BD, CD); Field 084 (CD); Classification Scheme and Edition (CD); Subfield $m, in field 082 (BD); Subfield $o, in field 082 (BD); Standard or optional designation (BD); Primary or access designation (BD); Field 653 (CD); Segmented Classification Number (CD); Field 654 (CD); Segmentation Instruction (CD); Subfield $t, in field 680 (CD); Subfield $t, in field 683 (CD); Subfield $t, in field 761 (CD); Topic information (CD); Subfield $e, in field 153 (CD); Classification number hierarchy--single number or beginning number of span (CD)
RELATED:
STATUS/COMMENTS:
6/6/2007 - Made available to the MARC community for discussion.
06/23/2007 - Results of the MARC Advisory Committee discussion
1 BACKGROUND
1.1 - Introduction
The Dewey editorial team is converting from a proprietary representation of the DDC into one based on the MARC 21 formats for Classification and Authority data in conjunction with the development of a new editorial support system at OCLC. The new system needs to support new models of representation, efficient editorial operations, interaction with translations, seamless delivery of DDC data to demonstration and research projects, version control, and derivative products.
The record format used for entries in the schedules, tables, and Manual will be based on an enhanced version of the MARC 21 Format for Classification Data. The record format used for Relative Index (RI) heading records and records for mapped headings from other vocabularies will be based on an enhanced version of the MARC 21 Format for Authority Data.
We are making this conversion because we want the data in the new Dewey Editorial Support System (ESS) to be available in a format supported by an international standard, and one that supports flexible data representation. We also want the data to be more fully integrated with data in OCLC’s WorldCat and in the related authority files, all of which use MARC 21. In addition, the system is being developed by people familiar with the MARC 21 Bibliographic and Authority Formats and using tools based on the MARC 21 formats.
The development team has identified several changes to the Classification and Authority Formats to improve the representation of DDC data. Some of these changes are being proposed below in section 2 of this discussion paper. Others are likely to be used primarily by DDC editors and translators, and thus will be local additions to the Classification and Authority Formats.
The German and Austrian communities are converting to MARC 21 from the MAB format and have proposed several changes to the MARC 21 Bibliographic Format related to the representation of DDC numbers. The Dewey editorial team and the team from Deutsche Nationalbibliothek (DNB) have agreed to work together on a proposal to address improvements to the representation of Dewey information across the MARC 21 formats. Changes are needed along a number of dimensions to address specificity and consistency in the representation of data.
1.2 - MARC 21 and DDC
MARC 21 contains provisions for Dewey numbers in all the formats, with the Classification Format having the most detailed provision and the holdings format having the least. Outside of the Classification Format and field 083 in the Authority Format, the representation is limited to full classification numbers and call numbers. There is little or no support for representation of notation from the external Dewey tables, notation from internal tables or elements of synthesized numbers, and limited information about the type and source of the number. The Bibliographic and Community Information Formats represent Dewey numbers in the 082 field. The Authority Format represents Dewey numbers in field 082 which contains a Dewey call number for a series and field 083 which contains a Dewey classification number associated with the 1XX heading. The Holdings Format represents Dewey numbers in the $h subfield (Classification part) of field 852 field.
The Classification format contains far more details for the representation of Dewey numbers. In particular, it has conventions for the representation of notation in external tables and for ranges of Dewey numbers. Notation in internal tables is represented within the record for the number or span of numbers under which the table appears. Elements of synthesized numbers are documented within the record for the number. The format also provides detailed information about the type and source of numbers.
2 DISCUSSION
This section includes issues that the Dewey team and DNB have found in the representation of DDC in the MARC 21 formats, and suggested changes to accommodate these needs.
2.1 - Identification of external table numbers
The full edition of the DDC contains six auxiliary tables (Table 1-6), referred to in this document as "external tables." The notation from the six tables is applied across the DDC to represent forms, genres, geographic areas, languages, ethnic groups, etc. At present, only the Classification and Authority Formats allow for indicating the external table that has been used to construct a Dewey number. The Dewey editorial team recommends adding the Classification Format representation of external table notation to the Bibliographic Format to support the identification of and access to external table information, as will as the recording of component parts of numbers.
The Classification format uses subfield $z for the external table number. For example,
153 ## $z 2 $a 73 [Dewey notation 73 in Table 2]
153 ## $z 2 $a 74 $c 79 [Dewey notation 74-79 in Table 2]
Proposed Change:
2.2 - Identification of internal add table numbers
The DDC also includes tables of notation within schedule and table records. These are referred to in this document as "internal add tables." In the current Dewey editorial support system and MARC Classification Format, internal add tables are represented using a single record for the internal add table. In the new ESS, the Dewey editorial team plans to use separate records for each entry in internal add tables. In addition to making it easier to support references to and from notation within add tables, the provision of separate records will support external identification of and access to the notation in internal add tables.
The Dewey editorial team recommends an extension to the Classification format to provide a separate MARC record for each entry of an internal add table in a 153 field. The Dewey editorial team proposes representing the internal add table notation within the 153 field using the existing subfields ($a and $c) both for the number or span at which the internal add table is located and for the table notation itself.
The Dewey editorial team recommends using a new subfield $y with the internal add table number immediately before the add table notation. Subfield $y will have the value 1 for the first internal add table at that location and values 2, 3, etc. for additional internal add tables at the same location.
153 ## $a 290 $y 1 $a 071 [Dewey notation 071 in the internal add table located at 290]
153 ## $a 362 $c 363 $y 1 $a 01 [Dewey notation 01 in the internal add table located under 362-363]
There can also be a span of notation within an internal add table in addition to a span for the table designation.
153 ## $a 333.7 $c 333.9 $y 1 $a 01 $c 09 [Dewey notation 01-09 in the internal add table located under 333.7-333.9]
Internal add tables are also located under Dewey notation in external Tables.
153 ## $z 1 $a 08 $y 1 $a 01 [Dewey notation 01 in the internal add table located under Dewey notation 08 in Table 1]
Number spans can be found as notation within internal add tables, and at the location of internal add tables.
153 ## $z 3B $a 1 $c 8 $y 1 $a 901 $c 907 [Dewey notation 901-907 in the internal add table 1 under Dewey notation 1-8 in Table 3B]
If there were multiple internal add tables at the same location, the value of subfield $y changes:
153 ## $a 810.1 $c818 $y 2 $a 3 [Dewey notation 3 in the second internal add table under 810.1-818]
Proposed change:
2.3 - Inclusion of component parts of numbers in bibliographic records
In the MARC Classification Format, the component parts of numbers are represented in field 765 (Synthesized Number Components). The information in the 765 field is useful to show how a number is built, to show the meaning of the component parts of a built number, and to enable searching of particular component parts in classification number records. For example:
153 ## $a 346.0469516 $h Social sciences $h Law $k Branches of law; laws, regulations, cases; law of specific jurisdictions, areas, socioeconomic regions $k Branches of law $h Private law $h Property $h Government control and regulation of specific kinds of land and natural resources $h Other natural resources $h Biological resources $h General topics $k Management and control $j Conservation and protection
765 0# $b 346.046 $a 346.046 $r 333 $s 95
765 0# $b 346.04695 $a 333.7 $c 333.9 $w 333.7 $c 333.9 $t 16
The first 765 field indicates that Dewey notation 95 comes from 333.95 Biological resources. The second 765 field indicates that Dewey notation 16 comes from notation 16 Conservation and protection in the internal add table at 333.7-333.9.
Dewey numbers are usually associated with bibliographic records in a precoordinated form, but the components parts of those precoordinated numbers contain a rich array of information. For retrieval, DNB wishes to provide easy access to the component parts of Dewey numbers in bibliographic records, in association with the 082 fields containing Dewey numbers. There would be two ways of providing this:
We do not see this as a choice between one method and another-if we adopt a consistent representation of classification number components, then data can be easily exported or imported between bibliographic files and classification files. We recommend continuing to represent the component parts of Dewey numbers using the 765 field in the MARC Classification Format, and propose using the structure of field 765 as the basis for a new field in the Bibliographic Format that can be used when importing or exporting data in a bibliographic record. The Bibliographic Format field would be specified as follows:
The Dewey editorial team suggests:
For example, for the number illustrated above, there could be the following 082 and 085 fields:
082 0# $8 1 $a 346.0469516 $2 22
085 ## $8 1.1 $b 346.046 $a 346.046 $r 333 $s 95
085 ## $8 1.2 $b 346.04695 $a 333.7 $c 333.9 $w 333.7 $c 333.9 $t 16
A more complex example would be the following successive 082 fields with LC-assigned numbers from the full and abridged editions, respectively, together with a DNB assigned number.
If 085 fields were added for the two built numbers, the record would contain the following:
082 01 $8 1 $a 346.0469516 $2 22
085 ## $8 1.1 $b 346.046 $a 346.046 $r 333 $s 95
085 ## $8 1.2 $b 346.04695 $a 333.7 $c 333.9 $w 333.7 $c 333.9 $t 16
082 11 $a 346.04 $ 2 14
082 04 $8 2 $a 346.0469514 $2 22/ger $5 DE-101b
085 ## $8 2.1 $b 346.046 $a 346.046 $r 333 $s 95
085 ## $8 2.2 $b 346.04695 $a 333.7 $c 333.9 $w 333.7 $c 333.9 $t 14
Proposed change:
2.4 - Classification number edition and source information in the Bibliographic Format
Currently, in field 082 in the Bibliographic Format, the edition number is carried in subfield $2. For the expansion of Dewey, the edition information needs to be expanded to include language of edition. In addition, the DNB wishes to document the agency assigning the number in the same field as the number itself. The second indicator already reports LC/non-LC as the assigning agency. In the case where the assigning agency is another agency, the subfield $5 (Institution to which field applies) is proposed, as in field 083 (Dewey Decimal Classification Number) in the Authority Format.
For edition information, The Dewey editorial team proposes to develop and maintain a registry of edition identifiers in which each edition would have a standard representation in subfield $2:
Examples:
$2 22 [latest version of DDC 22 file (currently available in WebDewey)]
$2 22/ger [latest version of German DDC 22 file (currently available in Melvil)]
The assigning agency information in the new $5 subfield would be coded:
082 04 $a 004 $2 22/ger $5 DE-101b
Proposed change:
2.5 - Designations for optional numbers in the Bibliographic Format
The Dewey editorial team would like to investigate a provision for the designation of optional numbers in the Bibliographic, Authority, and Community Information formats. Many optional numbers are already coded in existing bibliographic records as if they were standard Dewey numbers. One example is optional numbers in 340 Law, which used to be assigned by the Library of Congress:
082 00 $a 345.730772 $a 347.305772 $2 20
In the example above, 345.730772 is the standard Dewey number for sentencing in US criminal law, and 347.305772 is an optional number. Another example is optional numbers for Australian literature and Canadian literature in records coming from the National Library of Australia and from Library and Archives Canada:
082 04 $a A823.30108 $2 21
082 04 $a C811.008005 $2 21
In the two examples above, an optional letter denoting the literature ("A" for Australian English literature and "C" for Canadian-American literature in English, respectively) has been inserted before the literature number. The rest of the Canadian number is a standard Dewey number, but the Australian number has been extended by notation from an optional period table.
Optional numbers may be represented by a MARC Classification Format record where they are encoded with value b (Optional) in 008/09 (Standard or optional designation).
There are many optional numbers in bibliographic records in WorldCat, but no special designation exists to label the optional numbers as such.
The Dewey editorial team considered coding optional numbers in a separate subfield in the Bibliographic Format, but rejected this approach because of the large number of optional numbers already coded in subfield $a in legacy data. In section 2.6 discusses approaches to addressing "type of number" issues.
2.6 - Multiple numbers in the Bibliographic Format
The MARC 21 Bibliographic Format already supports the addition of multiple Dewey numbers to bibliographic records. Field 082 is repeatable, and subfield $a within field 082 is also repeatable. The current convention is to use a single 082 field for numbers assigned from the same DDC edition cited in subfield $2. For example, a number assigned from DDC 22 is put in a different 082 field from a number assigned from Abridged Edition 14:
082 01 $a 346.0469516 $2 22
082 11 $a 346.04 $2 14
LC also routinely adds optional Dewey "numbers" (e.g., notation "B" for biographical works) in addition to the standard Dewey number in successive $a subfields in the same 082 field:
082 10 $a 780.92 $a B $2 22
And until recently, the Library of Congress had routinely added a number for a series as a whole to records for specific works in a series; these two numbers were distinguished by the addition of an "s" after the series number:
082 10 $a 081 s $a 327.7 $2 21
In addition the DNB is also interested in adding table notation in a separate subfield where further addition to the full DDC number is not permitted by the rules; German virtual libraries are interested in assigning partial numbers instead of undertaking the process of number building; and there is general interest in providing multiple numbers for works with aspects in more than one Dewey category.
The Dewey editorial team welcomes these additional access points. At the same time, in order to support interoperability and choice of shelf location for physical items, it remains desirable to be able to identify the single DDC number that would be assigned following the instructions in the edition cited in subfield $2. The Dewey editorial team recommends the following actions:
2.6.1 - Use of 082
The Dewey editorial team proposes the addition of subfields $m and $o in the 082 field in the Bibliographic Format that can be used to indicate the nature of the number in subfield $a in relation to the edition cited in subfield $2. They would be specified as follows:
| $m | Standard or optional designation | |
| a - standard | ||
| b - optional | ||
| $o | Primary or access designation | |
| a - primary | ||
| b - access |
Subfields $m and $o are not required, and the default is for them to be absent.
In the following example, the first 082 is the "primary" number; the second is optional; and the third and fourth represent the German virtual library approach of just entering the elements in successive 082 fields without building the number. Without the proposed coding, one cannot distinguish primary numbers from access numbers (except in the case of the third number, with its $z), and one cannot distinguish standard numbers from optional numbers.
245 00 $a Criminal sentencing / $c edited by Robert Emmet Long.
082 00 $a 345.730772 $m a $o a $2 20
082 00 $a 347.305772 $m b $o a $2 20
082 04 $z 2 $a73 $m a $o b $2 20
082 04 $a 345.0772 $m a $o b $2 20
Proposed change:
2.6.2 - Use of 085 for direct assignment of component parts of numbers
Another option for this information would be to use new field 085 (Number components) to represent the direct assignment of component parts of numbers (external table numbers, internal table notation, partial notation derived from Dewey numbers) in addition to its proposed use for the explanation of component parts of numbers found in the 082 field. The reason would be to limit the use of the 082 field to the representation of full Dewey numbers to avoid making this widely used field confusing and complicated.
The absence of $8 subfields linking field 085 to a specific 082 field could be interpreted as a sign that field 085 is being used for direct assignment of component parts of numbers rather than to explain how a number in field 082 was built. If used to represent the direct assignment of component parts of numbers, the 085 field would also need to include aspects of 082 information not previously specified for the 085 field--first and second indicators for edition and assigning agency, respectively; subfield $2 for edition; subfield $5 for assigning agency; subfield $m for standard or optional designation (subfield $o is not needed, since it would be by nature "access only"). While this approach removes some complexity from the 082 field, it does so at the expense of adding complexity to the 085 field.
Proposed change:
If the 085 field is used both for direct assignment of component parts of numbers as described in section 2.6.2 above and for number building information, the indicators will need to be defined as in field 082 and additional subfields ($m, $2, and $5) will be needed.
2.7 - Segmentation information in the Classification Format
The current ESS encodes segmentation information, and this is shared with users of DDC in WebDewey. Currently the segmentation mark indicates the end of the number used in the abridged edition. (Formerly a second segmentation mark could be used to indicate the start of standard subdivision (Table 1) notation.) Segmentation marks are used in the 082 fields in bibliographic records created by the Library of Congress.
In the ESS, seg and set fields are used for segmentation information. The seg field is used for records for a single Dewey number or number span to show where the segmentation mark appears. The set field is used for Dewey number records that contain number-building instructions to give segmentation information that cannot be expressed using a single number or number span. New fields 653 and 654 in the Classification Format are proposed for segmentation information, to encode seg and set fields, respectively.
Examples from the old ESS of seg and set fields are:
In the record for 025.431:
| seg | 025.4/31 [025.4 is the abridged edition number][old ESS] |
653 ## $a 025.4/31
In the record for 636.9:
| set | Segment as shown in 599.2-599.8, e.g., 636.9205, 636.92/9, 636.935/6 [There is an add instruction at 636.9 to add notation from 599.2-599.8][old ESS] |
654 ## $i Segment as shown in$a599.2$c 599.8$i, e.g.,$a 636.9205 $i,$a 636.92/9$i,$a 636.935/6 [MARC proposal]
Proposed change:
2.8 - Encoding topic information in some Classification format fields
The Dewey editorial team wishes to use subfield $t for topics named in examples in the 680, 683, and 761 fields. The $t subfield is already used for this purpose in fields 453, 553, and 768, e.g.:
768 0# $i Unless other instructions are given, class a subject with aspects in two or more subdivisions of $a 641.563 $i in the number coming first, e.g., $t low-carbohydrate, low-calorie cooking for persons with diabetes $e 641.56314 $i (not $n 641.5635 $i or $n 641.5638)
Subfield $t is needed to distinguish topical information from "canned text" (canned text being text that remains the same for particular kinds of fields, regardless of topic), and to enable automated identification of the meaning of DDC numbers given in examples. Examples of the proposed use of the $t subfield are:
680 1# $i An area is classed in its present number even if it had a different affiliation at the time under consideration, e.g., $t Arizona under Mexican sovereignty $z 2 $a 791 $i (not $z 2 $a 72)
683 0# $i In building numbers, do not add by use of 0 or 1 (alone or in combination) more than twice, e.g., $t history of rock protest songs $a 782.421661592 $i (not $a 782.42166159209)
761 #0 $i Add to base number $b 025.29 $i notation $z 2 $d 1 $c 9 $i from Table 2, e.g., $t acquisition of materials from Latin America $e 025.298
Proposed change:
2.9 - Number hierarchy
Currently in field 153 of the Classification format only the upward caption hierarchy is given, in subfields $h and $k. The Classification numbers associated with those captions are not given. For example:
153 ## $a 616.8001 $c 616.8009 $h Technology $h Medicine and health $h Diseases $k Specific diseases $k Diseases of specific systems and organs $h Diseases of nervous system and mental disorders $j Standard subdivisions
DNB recommends using a new subfield $e (Classification number hierarchy -- single number or beginning of a span) in field 153 to link to the next number or number span upwards in the Classification number hierarchy. This will provide the link to the appropriate headings and the appropriate notes with hierarchical force. It will also indirectly provide for links to the downward hierarchy.
Examples:
$a 616.8001 $c 616.8009 $e 616.8
$a 616.8 $e 616.1 $c 616.8
$a 616.1 $c 616.8 $e 616.1 $c 616.9
$a 616.1 $c 616.9 $e 616
$a 616 $e 610
$a 610 $e 600
Proposed change:
3 Summary of Changes to Bibliographic 082
The following changes would be needed for field 082 - Dewey Decimal Classification Number
New subfield codes:
$c - Classification number--ending number of span (R)
$m - Standard or optional designation (NR) (see section 2.6)
$o - Primary or access designation (NR) (see section 2.6)
$y - Internal add table identification (R) (see section 2.2)
$z - Table identification (R) (see section 2.1)
$5 - Institution to which field applies (R) (see section 2.4)
Appendix A
Reference fields: current key fields for use in reviewing this Discussion Paper.
Bibliographic Format:
082 - DEWEY DECIMAL CLASSIFICATION NUMBER (R)
Indicators
First - Type of edition
0 - Full edition
1 - Abridged edition
Second - Source of classification number
# - No information provided
0 - Assigned by LC
4 - Assigned by agency other than LC
Subfield Codes
$a - Classification number (R)
$b - Item number (NR)
$2 - Edition number (NR)
$6 - Linkage (NR)
$8 - Field link and sequence number (R)
Classification Format:
153 - CLASSIFICATION NUMBER (NR)
765 - SYNTHESIZED NUMBER COMPONENTS (R)
Indicators
First - Field of number analyzed
0 - 153 field
1 - Other field
Second - Undefined
# - Undefined
Subfield Codes
$a - Number where instructions are found--single number or beginning
number of span (R)
$b - Base number (R)
$c - Number where instructions are found--Ending number of span (R)
$f - Facet designator (R)
$r - Root number (R)
$s - Digits added from classification number in schedule or external table (R)
$t - Digits added from internal subarrangement or add table (R)
$u - Number being analyzed (R)
$v - Number in internal subarrangement or add table where instructions are found (R)
$w - Table identification--internal subarrangement or add table (R)
$z - Table identification (R)
$6 - Linkage (NR)
$8 - Field link and sequence number (R)
HOME >> MARC Development >> Discussion Paper List
| The Library of Congress >> Especially
for Librarians and Archivists >> Standards ( 10/16/2007 ) |
Contact Us |