DATE: May 20, 2002
NAME: Coding for Publication Pattern at the First Level of Enumeration in MARC 21 Holdings Records
SOURCE: CONSER Task Force on Publication Patterns and Holdings
SUMMARY: This proposal considers the need to code publication pattern subfields $v (Numbering continuity), $u (Bibliographic units per next higher level) and $x (Calendar change) in fields 853-855 when only one level of enumeration is present. This is necessary for accurate predictive check-in and to eliminate inconsistencies in coding. The format currently specifies that these subfields are not used at the first level of enumeration and that they only apply to secondary levels. It would result in a coding practice change and revision of description of these subfields, but not a change to the content designators.
KEYWORDS: Bibliographic units per next higher level (HD); Calendar change (HD); Numbering continuity (HD); Subfield $u, in fields 853-855 (HD); Subfield $v, in fields 853-855 (HD); Subfield x, in fields 853-855 (HD).
RELATED: 2001-DP08 (June 2001)
05/20/2002-- Made available to the MARC 21 community for discussion.
06/17/02 - Results of the MARC Advisory Committee discussion - Rejected. There were objections from some that allowing restart at the first level had implications for display according to the NISO standard Z39.71, for legacy records, and for consistency with cataloging conventions. Others felt that the display should be accommodated by systems and that the change would enhance prediction capabilities. Since the user group in opposition to the proposal was some serials catalogers, it was suggested that the question be brought to the Committee to Study Serials Cataloging and the Serials Standards Committee for comments before further pursuing the change.
07/11/02 - Results of LC/NLC review - Agreed with the MARBI decisions.
PROPOSAL No. 2002-12: Coding for Publication Pattern at the First Level of Enumeration in MARC 21 Holdings Records
Users of the MARC 21 Holding format have encountered difficulties in producing 85X/86X publication pattern information for titles containing only one level of enumeration when the first level of enumeration restarts. For example, a title numbers from 1 to 4 every year of its quarterly life cycle and uses the change of the calendar year as the impetus to restart its numbering pattern. However, the instruction for subfield $v (Numbering continuity) in fields 853-855 state that restart cannot be used with the first level of enumeration. This instruction has resulted in a practice of putting the chronological unit (year) in the first level of enumeration so that the restart value (in subfield $v) can be associated with it. Reevaluation of the application of $u, $v, and $x should be made, since the practice of restarting the numbering on an issue with the turn of the calendar year is fairly commonplace, especially for titles with quarterly frequencies. No acceptable solution is currently available to aid prediction in these cases.
This issue was discussed at the MARC Advisory Committee meeting in June 2001 with Discussion Paper No. 2001-DP08. Participants in the discussion recognized the need for the change in the format, but wanted clear guidance on the situations in which coding at the first level would be used to allow for accurate coding. In addition it was felt that some consideration needed to be given to the impact on existing records.
2.1 Restart of first level of enumeration (Fields 853-855, subfields $u and $v)
The current practice of forcing chronology data into first level enumeration fields to enable restart capabilities at the second level is not consistent with the guidance provided in the format. The format now states, "When only chronology captions are used on an item (that is, the item carries NO enumeration), the chronology captions are contained in the relevant enumeration caption subfields..." A similar instruction follows regarding chronology values in the 86X fields. Libraries are now using chronology in subfield $a to force a restart because of the prohibition in the format of using $u and $v (with the value, restart) at the first level of enumeration. While some continue to question the advisability of transferring chronology data into enumeration fields, the inability to provide for titles that restart at the first level of enumeration encourages further variation of mixing both enumeration and chronology in the same series of subfields. From a data processing point of view this is a highly undesirable practice since we may need to evaluate coded chronology values in a different way than we evaluate enumeration values in generating our OPAC displays.
The practice of putting chronology data into enumeration fields to allow for restart of a then lower level of enumeration has been viewed favorably since the enumeration alone in such instances is not sufficiently unique to stand on its own as an identifier of the serial issue. However, titles such as Etudes Kurdes, a semiannual with serial issues descriptors that include a restarting enumeration in addition to year and month, have further complicated the prediction formula. Many systems force the year field to be recognized twice (e.g. 853 $a [year] $b no. $u 2 $v r $w f $i year $j month), yielding a display in one system of:
[year] 2001, no. 1 [2001 May]
The repeating year was made necessary since it was the only way to enable the enumeration to restart. Consideration
should be given to:
1. Enabling subfield $v (Numbering continuity) at the highest level of enumeration so that we can restart the enumeration at that level.
2. Allowing subfield $u (Bibliographic units per next higher level) to contain a limit for that level that would only be used in the event that there were no lower levels of enumeration. In other words, subfield $u could only be used at the first level of enumeration where there was only a single level of enumeration that restarts. In such instances, the value in subfield $u would tell the maximum number of issues that could be reached before the data in the first level of enumeration was re-initialized.
2.2 Allow fields 853-855, subfield $x (Calendar change) at the first level of enumeration
Another situation occurs when a serial with a single, higher level of enumeration is held static and only changes with the turn of the calendar year. In the case of ABA Journal, the volume number increases yearly while the monthly issues are distinguished by both year and month chronology. Using the format currently, a prediction system cannot hold the volume to its current value (volume 86). With an 853 field (e.g., 853 22 $8 1 $a v $i (year) $j (month) $w m), the automated systems force the volume number to increment every month. The pattern could only increment to a higher level when the calendar year changes if the limitation on the use of subfield $x at the first level of enumeration were to change. If the title was described as 853 $a v. $i (year) $j (month) $w m $x 01, the system should know that it can only increment the first level of enumeration every January. If we can say that calendar change is the chronological point at which a level of enumeration increments, we can apply it to a situation exactly like this. Again, this new interpretation of applying subfield $x would only occur in situations where the serial contained a single level of enumeration.
The following example illustrates where a single level of enumeration increments with reference to a chronological
point defined in subfield $x. (This assumes the approval of $u and $v at the first level.)
853 22 $8 1 $a v. $u 12 $v c $i (year) $j (month) $w m $x 01
[A monthly publication which increments or increases its enumeration value every January]
Here subfield $u indicates that there are 12 issues per calendar change and subfield $x shows that there is a single calendar change every year in January. The volume numbering is continuous in that it increments once per year. It covers ABA Journal which provides:
v. 87 (2000, Jan.) v. 87 (2000, Jul.) Next issue: v.88 (2001, Jan.) v. 87 (2000, Feb.) v. 87 (2000, Aug.) v. 87 (2000, Mar.) v. 87 (2000, Sep.) v. 87 (2000, Apr.) v. 87 (2000, Oct.) v. 87 (2000, May) v. 87 (2000, Nov.) v. 87 (2000, June) v. 87 (2000, Dec.)
Note that allowing subfields $u, $v, and $x at the highest level does not require a format change, only a change to the description in the subfields.
3.1. Implications for cataloging.
In previous discussions about allowing for publication pattern information at the first level of enumeration, questions have arisen about the impact on cataloging for the bibliographic item. However, the use of subfields $v, $u, and $x for publication pattern prediction should be independent of any decisions as to how to represent the enumerative and chronological designations for descriptive bibliographic purposes.
Current practice, which is codified in the revised Chapter 12 of AACR2, is to give the date as the first level of enumeration when it is determined the numbering is a subdivision of the date.
While cataloging description and predictive check-in may be seen as two separate processes, more catalogers have become involved with pattern creation with the implementation of MARC holdings. Furthermore, the macro developed for use in the CONSER publication pattern initiative uses the descriptive cataloging data in the record in order to create the pattern. The proposal to be able to use restart numbering at the first level of enumeration, thus, will create a disconnect between cataloging and check-in practices.
3.2. Implications for existing holdings records
Because the practice of coding at the first level of enumeration has been unavailable, existing holdings records may reflect a variety of coding practices. If this proposal is approved, institutions will need to work with their library system vendors to determine whether and how to change existing 853-855 fields in holdings records after implementations. If change is desirable, they could be revised as they are encountered to take advantage of the ability to more accurately predict the next issue in these situations.
The increasing use of the holdings format among libraries and the ability to share publication pattern information brought about by the CONSER publication pattern project suggests that a change in coding practice sooner rather than later will result in fewer existing records to change in the future.
3.3. Implications for compression
Some have questioned whether compression can adequately summarize enumeration if this change is introduced. Others have indicated that the ability to generate a display from the data in the 85X-86X fields may be addressed by display programs that are system dependent. These questions need to be considered.
In fields 853-855 of the MARC 21 Holdings format:
See Appendix A for suggested revisions to the format definitions.
Subfield $u - Bibliographic units per next highest level
The current definition of subfield $u (Bibliographic units per next highest level) in fields 853-855 defines it as a value describing the number of lower level units needed before the value in the next higher level increments as follows:
Subfield $u may be used with each level of enumeration except the first level (subfield $a or $g) because there is no higher level.
Based on the discussion offered here, this definition of subfield $u could be rewritten as follows:
When subfield $u is used in enumeration levels after the first (either subfields $b-$f or $h), it is defined as the number of lower level units at that level needed before the value at the next higher level increments. If there is only one level of enumeration in combination with chronological units (subfields $i-$l), subfield $u may be associated with subfield $a. In this context, subfield $u describes the maximum number of serial units issued (per calendar change) before the values in subfield $a increment. The calendar change dictates the point at which subfield $a increments. The interval of calendar change with reference to subfield $u applied at the highest level of enumeration is defined in subfield $x.
Subfield $v - Numbering continuity
The current definition of subfield $v (Numbering continuity) in fields 853-855 limits its application to lower levels of enumeration (subfields $b-$f and subfield $h) as follows:
Subfield $v may be used with each level of enumeration except the first level (subfield $a or $g).
Based on the discussion offered here, this definition of subfield $v could be rewritten as follows:
When subfield $v is used with enumeration levels after the first, it defines the numbering behavior as either "c," number increments continuously or "r," numbers restart at the completion of a specific unit. When it is used in conjunction with subfield $u to characterize the incrementing pattern of the first, and only, level of enumeration, it increments the numbering at the first level with reference to the values defined in subfield $x (Calendar change).
Subfield $x - Calendar Change
The current definition of subfield $x (Calendar change) in fields 853-855 limits its application to lower levels of enumeration (subfields $b-$f and subfield $h)as follows:
Subfield $x contains one or more two- or four-character codes that indicate the chronological point at which the next higher level increments or changes.
Based on the discussion offered here, this definition of subfield $x could be rewritten as follows:
Subfield $x contains one or more two-or-four-character numeric codes that indicate the chronological point at which enumeration increments or changes. In cases where a single level of enumeration describes a serial's sequential issues, subfield $x describes the chronological point at which the value in the first level enumeration (subfield $a) increments. In cases where multiple levels of enumeration exist, subfield $x describes the chronological point at which the next higher level of enumeration increments. It may only be used with the first level of enumeration ($a) when there are no other levels present.