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

HOME >> MARC Development >> Discussion Paper List

MARC DISCUSSION PAPER NO. 2018-DP09

DATE: May 25, 2018
REVISED:

NAME: Improving Subfield Structure of Field 245 in the MARC 21 Bibliographic Format

SOURCE: OCLC EMEA, Leiden (Netherlands) and OCLC, Dublin, Ohio (US)

SUMMARY: This paper describes two options for a more granular subfield structure of Bibliographic field 245 (Title Statement) to improve data exchange with other more granular formats, matching (and merging), displays, and indexes.

KEYWORDS: Field 245 (BD); Title Statement (BD)

RELATED: 2010-07; 2010-DP01

STATUS/COMMENTS:
05/25/18 – Made available to the MARC community for discussion.

06/24/18 – Results of MARC Advisory Committee discussion: Despite the historical appeal of what is explored in the paper, the general consensus from MAC was that it introduces more complexity and confusion as well as moving in a direction away from RDA’s "manifestation statement" of a single, undifferentiated string of characters derived from the preferred source. Additional comments observed that attempting to make a single field serve double duty (description and access) has proven problematic. A straw poll on whether it would be fruitful or not to move forward was 12 to 8 against. The authors, OCLC’s Leiden office, were not present and the OCLC liaison did not feel comfortable burdening them or speaking for them with respect to pursuing the idea further. Though the discussion paper in its present form will not return as a proposal, it was credited for initiating a conversation about the merits/drawbacks of more granularity in field 245.

Discussion Paper No. 2018-DP09: Improving Subfield Structure of Field 245

1. BACKGROUND

In 2010, MARBI considered Discussion Paper No. 2010-DP01: ISBD punctuation in the MARC 21 Bibliographic Format, leading to the passage of MARC Proposal No. 2010-07: ISBD Punctuation, in which codes ‘c’ and ‘I’ were added to Leader/18. As part of the discussion at the Midwinter Meeting, the committee looked at issues surrounding the internal punctuation in the 245. One proposed solution for this issue was the addition of more granular subfield coding. We believe it is time to revisit that discussion.

The structure of the 245 field has many consequences for things that are related to or dependent on it, like data exchange with other formats, displays (e.g. labeled display), indexes and matching (and merging).

In this document we will present two different options to improve the structure of field 245 (Title Statement).

• Option 1 uses existing subfields, but making them repeatable.
• Option 2 uses not only existing subfields, but also adds several new subfields.

From the perspective of data management, we think that Option 2 is preferable to Option 1.

The starting point is that subfields ought to be as specific as possible and that the different elements of the title statement ought to be put in subfields that are very precise.

Note: The “Revised Final Report of the PCC ISBD and MARC Task Group”, online at https://www.loc.gov/aba/pcc/documents/isbdmarc2016.pdf, examines the use of both pre-ISBD and ISBD punctuation in MARC 21 bibliographic and authority records. In this report, section 4.5 Parallel Data, and section 4.6 Bibliographic Field 245 – Title Statement analyze field 245 specifically. Although the approaches of this report and this Discussion Paper are somewhat different, we think that the approaches are not contradictory, but complementary, especially option 2. As said, option 1 uses only repeatable existing subfields. Because option 2 introduces several new subfields, this option comes closest to the approach of the “Revised Final Report of the PCC ISBD and MARC Task Group”.

2. DISCUSSION

2.1. Problems That Will be Solved

By implementing Option 1 or Option 2, many current problems will be solved.

2.1.1. Structure of field 245 causes problems in data exchange with other more granular formats

The data exchange from non-MARC 21 formats to the MARC 21 format, and vice versa is very problematic for many records of (European) libraries. Especially for libraries that use UNIMARC data exchange is problematic, as UNIMARC has a more elaborate title statement structure too. A more granular subfield structure in field 245 for MARC 21 will improve data exchange globally.

2.1.2. Current structure of field 245 makes it impossible or at least difficult for machines to distinguish key elements in bibliographic descriptions

When you have defined separate subfields for special elements, it is very logical that the use of it is not limited to a single instance, rather than defining that a subfield is not repeatable. When a subfield has a certain meaning or purpose, the corresponding data should be put in the most specific subfield. And when a kind of element occurs more than once, the corresponding most specific subfield should be repeatable. Now there exist inconsistencies in field 245, because on the one hand a kind of element is put in a specific subfield and on the other hand next occurrences of that same element are not.

Examples of non-repeatable subfields: subfields $a (Title), $b (Remainder of title), $c (Statement of responsibility).

When subfields $a, $b and $c will be made repeatable, it will have consequences for many systems which have been designed around those being non-repeatable.

In Option 1, subfield $a is repeatable when there is/are (an)other title(s) of the same author, (a) title(s) of (an)other author(s) and (a) parallel title(s) proper; in Option 2, subfield $a is only repeatable when there is/are (a) parallel title(s) proper, because subfield $t is proposed as a new subfield.

2.1.3. Precise indexing is not possible with current structures

Indexes on subfields contain on the one hand a lot of “wrong” data, and on the other hand lack much data that could be part of a specific index.

Example: subfield $c Statement of responsibility, etc.
Subfield $c is used for the statement of responsibility, so for personal authors, corporate bodies, meeting names. In addition to statements of responsibility, $c could also contain titles of additional works, parallel titles, other title information, or weak or dependent titles of multivolume publications. This can negatively impact search results for (end) users.

Conclusion: The current use of subfield $c has a negative impact on search results for (end) users.

2.1.4. Labeling of data displays is difficult

When you use e.g. the label “authors / corporate bodies / meetings” in a labeled display, and you display subfield $c behind this label, you often get much more than only a statement of responsibility. For instance, other titles, subtitles, parallel titles, etc. But also, in case of multivolume publications with weak titles, you get the number and the title of the part.

2.1.5. Problems in matching data

The matching (and merging) of records in which subfield $c contains other titles, volume numbers and titles of weak or dependent titles, etc. is problematic. By distinguishing all title elements and statement of responsibility elements by putting these in separate subfields, matching and merging problems for these kinds of records can be improved in shared catalogs.

2.1.6. No clear distinction between “bibliographic” and “literal” punctuation

With the proposed allowed repeatability of (new) subfields in this discussion paper, it will be clearer whether a certain punctuation is “bibliographic” or “literal”. Bibliographic (ISBD) punctuation is e.g. a space-slash-space between the title proper and the first statement of responsibility; non-bibliographic punctuation is e.g. the periods used for abbreviations and initials.
With the proposed allowed repeatability of (new) subfields, bibliographic punctuation can only exist in combination with a subfield (indicator). However, this punctuation will not always be the same, because in the English language of cataloging records alone, the transitions from AACR1 to AACR2 to RDA have resulted in varying practices over the decades.

2.2. Overview of the Current Situation and the Two Options

The following table offers an overview of the current situation of field 245 and the two options for improving the structure of it. The proposed changes are marked with bold. We have left out the indicators.

2.3. Benefits and Expected Results

2.3.1. Improved data structure to distinguish key data elements in bibliographic data

a) Clear distinction between “real” titles and “remainders” of title

In both options, a clear distinction is made between on one hand the “real” titles of an item, like the main title, the parallel title proper, the other title(s) of the same author and the title(s) of (an)other author(s), and on the other hand the “real remainders” of title, i.e. subtitles and other additional title information.

These “real” titles are put in separate subfields $a in Option 1.

In Option 2 the main title and the parallel title proper are put in separate subfields $a; and for the other title(s) of the same author and the title(s) of (an)other author(s) separate subfields $t need to be used.

In both options, subfield $b is used for the “real remainders” of title, i.e. subtitles and other additional title information. A consequence of this is that a distinction is made between a parallel title proper (=$a) (Option 1 and 2) and a parallel remainder of title (=$b) (Option 1) or (=$j) (Option 2). So, subfield =$b or =$j is used when the subtitle of the main title itself has one or more parallel titles.

b) Repeatability of statement of responsibility, distinction between 1st and next statement(s) of responsibility, and further subfielding, once a subfield $c has been recorded

Under current practice, once there is a subfield $c in field 245, no further subfielding is possible at all. Subsequent titles and statements of responsibility are therefore “lost,” having been subsumed into the first statement of responsibility in a single subfield $c.

Now subfield $c is not repeatable. So next categories of publications are a problem:

• Publications with titles of several authors (without a collective title)
  Both the titles and the authors of the second and next titles “disappear” in subfield $c behind the author of the main title.
• Multivolume publications with weak or dependent titles with authors on different levels
  The different weak titles (and remainders of title) including the authors belonging to the specific level(s) also “disappear” in subfield $c behind the author of the main title. Weak or dependent titles are put in the subfields $n (Number of part/section of a work) and/or $p (Name of part/section of a work).

Not only for indexing (and searching), but also for matching (and merging), it is important that the different 1st and next statements of responsibility, but also the different weak titles will be taken in account properly. But in the current format these statements of responsibility and weak titles can't be recognized as such, because these all can be lumped together (including the main statement of responsibility) in the same single subfield $c.

From the perspective of data management, it is preferable to make also a distinction between the first and next statements of responsibility.

In Option 1 subfield $c is used in two “flavors”: /$c First statement of responsibility and ;$c Next statement of responsibility.

In Option 2 subfield /$c is used for First statement of responsibility and ;$e for Next statement of responsibility. A separate subfield ;$e is used for every next statement of responsibility.

The terms “first statement of responsibility” and “next statement of responsibility” are “relative” and refer to the title proper or to (a) subsequent title(s).

Example
245 [Title 1] / [First statement of responsibility] ; [Next statement of responsibility]. [Title 2] / [First statement of responsibility] ; [Next statement of responsibility] ; [Next statement of responsibility]

c) Other subfields also repeatable

Since subfields will be made repeatable, other subfields that depend on these should also be made repeatable:
$f         Inclusive date
$g        Bulk dates
$h        Medium
$s         Version

2.3.2. Improved data exchange with other more granular formats

(See above: 2. Discussion)

2.3.3. Improved matching mechanisms

(See above: 2. Discussion)

2.3.4. Improved displays

(See above: 2. Discussion)

2.3.5. The more granular the MARC 21 format is, the better it can be crosswalked to a linked data structure

When you have defined separate subfields for special elements, it is very easy to convert field 245 to another format, like linked data, and vice versa. When subfields can contain a combination of different elements, like subfield $c, you will have to solve many problems during the conversion process when you want to distinguish the different elements.

So, improving the structure of field 245 in the MARC format is not only good for linked data, but also for the data exchange with other bibliographic formats.

3. EXAMPLES

The following table shows three situations. First the current structure (C), followed by Option 1 (O1) and Option 2 (O2). We have marked the differences with the current situation.

Example 1

Example 2

Example 3

Example 4

Example 5

Example 6

Example 7

Example 8

Example 9

Example 10

Example 11

Example 12

4. BIBFRAME DISCUSSION

No direct impact on BIBFRAME but for any conversion of data it supports more “data picking” from statements.

5. QUESTIONS FOR DISCUSSION

5.1. In this document two different options to improve the structure of field 245 (Title Statement) are presented.

• Option 1 uses existing subfields, but makes them repeatable.
• Option 2 uses not only existing subfields, but also adds several new subfields.

Do you agree that, from the perspective of data management, Option 2 is preferable to Option 1?

5.2.  Do you agree that data exchange between the MARC 21 format and other more granular formats will be improved by implementing option 1 or option 2?

5.3. Do you agree that improving matching (and merging) of bibliographic records is important and will be accomplished by implementing option 1 or option 2?

5.4. Do you agree that displays and indexes will be improved by implementing option 1 or option 2?

5.5. Do you agree that the more granular the MARC 21 format is, the better and more precisely it can be crosswalked to a linked data structure?

HOME >> MARC Development >> Discussion Paper List