The Library of Congress >> Especially for Librarians and Archivists >> Standards

MARC Standards

HOME >> MARC Development >> Discussion Paper List


MARC DISCUSSION PAPER NO. 2022-DP03

DATE: December 21, 2021
REVISED:

NAME: Recording Concrete Action Interval Dates in Field 583 of the MARC 21 Bibliographic and Holdings Formats

SOURCE: OCLC Shared Print Metadata Advisory Group

SUMMARY: This paper explores options within field 583 (Action Note) of the Bibliographic and Holdings formats to allow the recording of concrete end dates, in alignment with the convention in subfield $c (Time/date of action). Additionally, the paper proposes refining and possibly relocating the date formatting instructions presently located under 583 subfield $c.

KEYWORDS: Field 583 (BD, HD); Action Note (BD, HD); Time/date of action (BD, HD); Subfield $c, in field 583 (BD, HD); Action interval (BD, HD); Subfield $d, in field 583 (BD, HD); Shared print metadata guidelines (BD, HD)

RELATED:

STATUS/COMMENTS:
12/21/2021 – Made available to the MARC community for discussion.

01/27/22 – Results of MARC Advisory Committee discussion: The committee was broadly split into two groups – those who favored an easier to implement approach with respect to the existing data and those who favored a cleaner implementation approach (that would involve significant data migration). A straw poll was held concerning whether to keep the data in $d (and adjust its definition to accommodate it) or to move the data to a new subfield, with the majority voting in favor of a new subfield. The possibility of using a range of dates, as supported by EDTF, in $c was raised. There was some point/counter-point regarding the need to specify EDTF via $2, and the evolving nature of IS0 8601 for dates. A second straw poll concerning whether to use a new subfield or expand the use of $c to accommodate a range of dates yielded a slight preference for adjusting $c. The paper may return as a proposal or follow-up dicussion paper. The authors will explore implementing the EDTF standard within subfield 583 $c to represent date/time intervals; coding end dates in a new subfield may form the basis of a second option. 


Discussion Paper No. 2022-DP03: Recording Concrete Action Interval Dates in Field 583

Preliminary note about the OCLC Shared Print Metadata Advisory Group: This is a group of shared print practitioners, including both program staff and librarians at institutions participating in a shared print program. Membership includes staff from the Eastern Academic Scholars’ Trust (EAST), the Big Ten Academic Alliance (BTAA), the Western Regional Storage Trust (WEST) the Center for Research Libraries (CRL), COPPUL SPAN, and the HathiTrust shared print programs. This group serves as an advisory group to OCLC on Shared Print uses of the 583, and has been meeting with representatives from OCLC over the past several years to discuss Shared Print needs.

1. BACKGROUND

Field 583 is currently defined in both the Bibliographic and the Holdings formats as follows:

Subfields $c and $d are currently defined in both the Bibliographic and the Holdings formats as follows:

$c - Time/date of action
Concrete time or date of a future or past action.

In conjunction with the appropriate action this may indicate date of accessioning, acquisition, classification, transfer, or description. The date and time are recorded according to Representations of Dates and Times (ISO 8601). The date requires 8 numeric characters in the pattern yyyymmdd (4 for the year, 2 for the month, and 2 for the day). The time requires 8 numeric characters in the pattern hhmmss.f (2 for the hour, 2 for the minute, 2 for the second, and 2 for a decimal fraction of the second, including the decimal point). The 24-hour clock (00-23) is used.

$d - Action interval
Time period which cannot be expressed as a specific date (e.g., at end of academic term or every six months).

2. DISCUSSION

Starting around 2010, the practice of libraries working together to commit to retain print materials (typically books or serials) began to emerge. Practitioners saw these activities as extensions of the already well-defined set of actions used by the Preservation community. The Preservation community had already formalized an extensive vocabulary for its actions, the Preservation and Digitization Actions terminology, and was using Field 583 to record information about those actions. The new Shared Print community decided to record commitment retention information in the same field used by the Preservation community.

The Preservation community conceived of actions as occurring either on specific dates or within time periods that cannot be precisely specified. They recorded the "concrete time or date of a future or past action" in subfield $c and any "time period which cannot be expressed as a specific date" in subfield $d.

The Shared Print community needed to express a range of specific dates between which retention commitments would be considered valid. In order to distinctly code for start and end dates, the community of practice has co-opted $c for a start date of a commitment and $d for the purpose of recording end dates, regardless of whether this date was a concrete date (e.g., March 15, 2035) or a time period (e.g., "25 years" or "permanent"). This allowed them to determine whether a particular institution was still under an obligation to retain a particular title or item. While this was sometimes not in accordance with the current definition of the subfield $d, no attempt was made at that time to distinguish the coding of concrete dates from time periods. As the shared print community grew, this practice evolved into the recommendation that concrete dates in both subfields $c and $d be recorded in the format "yyyymmdd". This permits shared print practitioners to do date calculations and also to systematically compare the end dates of retentions.

Currently, the vast majority of Shared Print Field 583s (i.e., those with a subfield $a of "committed to retain") found in local holdings records, both in local catalogs and in OCLC, use the "yyyymmdd" format in the subfields $c and $d. There are currently over 25 million records in OCLC using Field 583 in this manner. A brief investigation of 583 $d fields outside of the Shared Print community has revealed that the subfield $d is neither widely used (only ~1.7% of non shared print 583s have a $d) nor consistently formatted, as concrete dates are often found in the $d of records that are not part of Shared Print commitments.

Note that in developing the paper, it was found that the existing guidance under 583 $c seems to mandate full development of a date and time stamp, where neither the examples nor the ISO 8601 standard require it. We recommend aligning the guidance language to conform with that found elsewhere in the formats, such as at Authority field 046: "The date and time are recorded according to Representations of Dates and Times (ISO 8601) in the pattern yyyy, yyyy-mm, or yyyymmdd (4 for the year, 2 for the month, and 2 for the day) unless subfield $2 (Source of date) specifies another date scheme."

Also in the development of the paper, the possibility of using EDTF (Extended Date Time Format) formatted dates in $c to support the pairing of start and end dates was raised. Since the default date specification is ISO 8601, the deployment of EDTF dates would require specifying that in an additional subfield for Source. The typical mechanism is $2. However, $2 is already explicitly defined in field 583 for identifying "the source of the term used to record the action information." It would not be practicable to divide the data into multiple instances of field 583 that would split the action from the date, in order to distinctly record the data with its source code in $2. It is conceivably possible to develop another subfield to record Source of date, but this feels simultaneously unwieldy with respect to the resulting complexity of the solution and the divergence from other cases where the EDTF standard is specified.

For discussion purposes in Section 3 below, we include four options for addressing the need for concrete start and end dates for the shared print community, with the first option being our preferred solution.

3. OPTIONS FOR ENCODING SHARED PRINT END DATES

3.1 OPTION 1: Change in the definition of subfield $d

Change the definition of $d (Action interval) from the current "Time period which cannot be expressed as a specific date" to "Time period over which an action can occur. This can be expressed either as a time period or a concrete date formatted according to ISO 8601."

This represents the smallest change to the format, would accommodate all current usages of this subfield and would require no action on the existing records. The disadvantage of this alternative is that it includes mixed types of data in a single subfield.

3.2. OPTION 2: Define a new subfield $g to record a specific, formatted end date

A new subfield could be defined to record concrete end dates:

$g - Action interval concrete ending date
Ending date of action described in $a, with the start date defined in $c. The date and time are recorded in the same format as $c.

Existing $d data in the "yyyymmdd" format would need to be moved to the new $g. This would retain the existing definition of $d and preserve any free-text interval data recorded there, with the implication that moving a significant amount of specific date data for these retention commitments would be involved both in local systems and national utilities.

3.3. OPTION 3: Revision of subfield $d and creation of subfield $g

Revise subfield $d to record "Action interval concrete ending date" and create subfield $g to record "Action interval":

Current definition of $d:

$d - Action interval
Time period which cannot be expressed as a specific date (e.g., at end of academic term or every six months).

Proposed revision of $d:

$d - Action interval concrete ending date
Ending date of action described in $a, with the start date defined in $c. The date and time are recorded in the same format as $c.

Proposed new subfield:

$g - Action interval
Time period which cannot be expressed as a specific date (e.g., at end of academic term or every six months).

This separates out the distinct usages of the current subfield $d. While this option also requires the moving of data, the subfield $d is not currently widely used within the Preservation community, thus this change would have a lower impact on existing records.

3.4. OPTION 4: Create a new pair of subfields ($s and $t) for start and end dates

Define new subfields $s and $t to record the specific start and stop endpoints of an interval.

Subfield $c could be retained as the date of an action, in this case the effective date of agreeing to a shared print retention, even if it is duplicative with the data in $s. The $d would be retained for the existing free-text interval data. There would be significant data movement involved for the formatted date data that has given rise to the discussion paper, and even more so, as not only would the existing $d data need to move to $t, but the existing $c date would need to be duplicated in a $s. This option would result in a clean data structure.

Proposed new subfields:

$s - Action start date
Concrete start date of action interval for action described in $a.

$t - Action end date
Concrete end date of action interval for action described in $a.

4. EXAMPLES

Since the nature of the retention data recorded in field 583 is not specific to any particular resource, the examples are confined solely to the articulation of field 583 under the various options.

4.1. OPTION 1

Examples with concrete end date for the retention commitment in subfield $d as well as a time period which cannot be expressed as a specific date:

583 1# $a committed to retain $c 20110103 $d 20351231 $f WEST $f WEST Bronze $u https://www.cdlib.org/wp-content/uploads/2018/11/WESTProgramStatement.pdf

583 1# $3 v.1-v.195 (1912-2010) $a committed to retain $c 20110101 $d permanent $f CRLJSTOR $j CRLAA $u http://www.crl.edu/archiving-preservation/print-archives/crl-administered/jstor $5 ICRL

4.2. OPTION 2

Examples moving concrete date into subfield $g:

583 1# $a committed to retain $c 20110103 $g 20351231 $f WEST $f WEST Bronze $u https://www.cdlib.org/wp-content/uploads/2018/11/WESTProgramStatement.pdf

583 1# $3 v.1-v.195 (1912-2010) $a committed to retain $c 20110101 $d permanent $f CRLJSTOR $j CRLAA $u http://www.crl.edu/archiving-preservation/print-archives/crl-administered/jstor $5 ICRL

4.3. OPTION 3

Examples moving time period which cannot be expressed as a specific date into subfield $g:

583 1# $a committed to retain $c 20110103 $d 20351231 $f WEST $f WEST Bronze $u https://www.cdlib.org/wp-content/uploads/2018/11/WESTProgramStatement.pdf

583 1# $3 v.1-v.195 (1912-2010) $a committed to retain $c 20110101 $g permanent $f CRLJSTOR $j CRLAA $u http://www.crl.edu/archiving-preservation/print-archives/crl-administered/jstor $5 ICRL

4.4. OPTION 4

Examples using paired subfields $s and $t to record concrete start and end dates:

583 1# $a committed to retain $c 20110103 $f WEST $f WEST Bronze $s 20110103 $t 20351231 $u https://www.cdlib.org/wp-content/uploads/2018/11/WESTProgramStatement.pdf

583 1# $3 v.1-v.195 (1912-2010) $a committed to retain $c 20110101 $d 25 years $f CRLJSTOR $j CRLAA $s 20110101 $t 20460101 $u http://www.crl.edu/archiving-preservation/print-archives/crl-administered/jstor $5 ICRL

583 1# $3 v.1-v.195 (1912-2010) $a committed to retain $c 20110101 $d permanent $f CRLJSTOR $j CRLAA $s 20110101 $u http://www.crl.edu/archiving-preservation/print-archives/crl-administered/jstor $5 ICRL

5. BIBFRAME DISCUSSION

Field 583 information is part of the BIBFRAME Item. There is currently no attempt to convert Field 583 subfield $d to BIBFRAME. Field 583 subfield $c is converted as a date. That could also apply to Field 583 subfield $d or whatever subfield carries this information when the yyyymmdd format is used within the subfield. The option to redefine 583 subfield $d would mix data formats in the subfield, with potential implications for linked data conversion in any prospective BibFrame usage.

6. QUESTIONS FOR DISCUSSION

6.1. Should the 583 $c date definition be updated to align with the language found elsewhere in the formats?

6.2. Has the case for allowing concrete ending dates in field 583 been adequately stated?

6.3. Do the proposed solutions meet the needs discussed?

6.4. Is a new subfield or subfields merited, considering that (a) the majority of existing data is well formatted and (b) it is unlikely that libraries would retrospectively change their existing data.

6.5. Should a subfield $g solution be pursued (options 2 and 3) if there is a known reason why subfield $g is open and if another alternative among the open options (subfields $m, $p, $q, $r, $s, $t, $v, $y) would be preferable?

6.6. Is there another solution to support recording a specific end date for an action such as a retention commitment that spans a period of time?

6.7. Are there other potential issues that have been overlooked but need to be considered?

6.8. Assuming the use case is made, of the possible options, which is preferred for development?


HOME >> MARC Development >> Discussion Paper List

The Library of Congress >> Especially for Librarians and Archivists >> Standards
(05/17/2022)
Legal | External Link Disclaimer Contact Us