DATE: November 15, 1997

NAME: New Type of Date code for incorrect dates

SOURCE: ALA Rare Books and Manuscripts Section, Bibliographic Standards Committee

SUMMARY: This discussion paper concerns the establishment of a new type of date code in the Bibliographic format to provide access to incorrect publication dates appearing on items.

KEYWORDS: 008/06; 008/07-14; Type of Date code; Date 1/Date 2



11/15/97 - Forwarded to USMARC Advisory Group for discussion at the January 1998 MARBI meetings.

1/10/98 - Result of USMARC Advisory Group discussion - The reaction to the proposal topic was mixed. Several speakers supported the inclusion of a new type of date code but recognized that the solution for situations of the multiple dates was problematic. There was a suggestion that the 046 be investigated to accommodate these dates. The type of date preference chart needs also to factored into the solution. The 4th example under section 2 was withdrawn by the ALA Rare Books Committee since it was not the intention to include the issue of non-Gregorian dates. There was consensus that the topic come back as a proposal.

DISCUSSION PAPER NO. 106:  Type of date code 
for incorrect dates


When the date of publication on an item is incorrect, cataloging
rules usually instruct the cataloger to give the date, in the
imprint, as found on the item followed by the correct date in
brackets.  In the USMARC Formats for Bibliographic Data the
publication date is also encoded in the field 008/06-14 for access
purposes and only the correct date is given:

    260    . . . $c1703 [i.e. 1730]
    008/06 = s
    008/07-10 = 1730
    008/11-14 = [blanks]

This coding convention in the 008 field fails either (a) to
indicate that the date in 008/07-10 is a corrected date, or (b) to
allow retrieval by the on-the-item (incorrect) date.

These issues were discussed at the 1996 meeting of the ACRL/RMBS
MARC for Special Collections Discussion Group.  A consensus of the
group was that a new date type code for Corrected and given date
and inclusion of the on-the-item (incorrect) date in 008/11-14
should be proposed. 


Although incorrect imprint dates can appear on printed materials of
any era, they are a regular feature in the era of hand composition
of movable type.  It is an easy mistake to invert numbers or
letters (e.g., 1639 for 1693, or MDCCXLII for MDCCLXII).
Bibliographical and historical scholarship has also frequently
revealed that the actual publication dates of items do not
correspond to their printed dates.  Records for early printed books
therefore routinely contain corrections to the dates given in the

The coding of date information in field 008 described above is
problematic because many systems use the fixed fields for indexing
purposes, and a researcher may miss relevant items entirely if they
search by the on-the-item (incorrect) date.

It is proposed that the on-the-item (incorrect) date be recorded in
field 008/11-14 so that it can be used for retrieval, and that a
new date type code x be added to 008/06 for Corrected date and
Incorrect date.  In the table of precedence for single items, value
x would following value b (B.C. date).

The following partial examples illustrate the proposed coding:

    008/06 = x
    008/07-10 = 1693
    008/11-14 = 1639
    260    . . . $c 1639 [i.e. 1693]
        [digits transposed]

    008/06 = x
    008/07-10 = 1762
    008/11-14 = 1742
    260    . . . $c MDCCXLII [i.e. 1762]
        [letters transposed: MDCCXLII for MDCCLXII]

    008/06 = x
    008/07-10 = 1788
    008/11-14 = 1786
    260    . . . $c 1786 [i.e. 1788]
    500    On title page: 1786; in colophon: reprinted in 1788

    008/06 = x
    008/07-10 = 1503
    008/11-14 = 1502
    260    . . . $c Id. Mart. 1502 [15 Mar. 1503]
        [Non-Gregorian, "old style" date]
    [Submitted for this discussion by RBMS/BSC but not really an
    incorrect date.]

    008/06 = x
    008/07-10 = 1996
    008/11-14 = 1997
    260    . . . $c 1997 [i.e. 1996]
        [Item distributed before date on title page]


For multipart items complete in more than one year, either the
beginning or the ending date of publication may contain an
incorrect date.  In such a case, the date elements can either be
used to record a beginning and an ending date OR they can be used
to record a corrected and on-the-item date.  If a beginning and an
ending date are recorded, that date must be either the corrected
date or the on-the-item date.  The Bibliographic Standards
Committee has devoted considerable discussion to this issue.  The
following represents a convention that is supported by most members
of the Committee.

If the beginning date of publication is given incorrectly on the
item, record in 008/07-10 the EARLIER of the two dates (the
corrected or the on-the-item date); if the ending date of
publication is given incorrectly on the item, record in 008/11-14
the LATER of the two dates.  This convention would give the widest
range of retrieval and would produce the highest probability that
the item would be retrieved using either corrected or on-the-item

The following examples illustrate this convention:

    008/06 = m
    008/07-10 = 1639
    008/11-14 = 1699
    260    . . . $c 1639 [i.e. 1693]-1699
        ["1639" is recorded in 008/07-10 because it is earlier that

    008/06 = m
    008/07-10 = 1776
    008/11-14 = 1876
    260    . . . $c 1776-1876 [i.e. 1786]
        ["1876" is recorded in 008/11-14 because it is later that

Other conventions that were preferred by some Committee members
were (a) always to record the on-the-item dates, and (b) to record
the corrected and on-the-item beginning date of publication (using
value x in 008/06).  As noted above, most committee members prefer
the earliest/latest convention.  However, there is still need for
additional discussion on this point.


The RBMS/BSC submitted the following comments on impact, based on
their discussions.

The change would help users find materials.  The new coding would
affect catalogers and system vendors and should be made mandatory
if applicable.  Vendors will need to assess their costs of
implementation.  Training costs would be minimal.  The proposal
could be applied retrospectively, but this is not necessary for its
implementation.  The new coding would have benefits for any record
to which it were applied, but retrieval of existing records would
not be changed.  In the whole universe of cataloging records, the
proportion of materials affected would be infinitesimal.  For early
printed books, there would be a significant number -- but still
less than 1%.  An alternative way of achieving the same result
would be by keyword indexing of 260 $c, but would probably be less
cost effective.


1.  What do local systems and vendors do with the different types
of dates now?  Will the addition of another type be of value or
would expansion of an existing one be more useful?

2.  Many non-early and non-rare materials have the above date
characteristics.  Would practices change for all of these?  How
about atlases which regularly have "incorrect" dates on the item
since the atlas is for the next year?  Would the added complexity
of coding be valuable for current material?

3.  Can a value be limited to early printed books?

4.  Could any of the other existing date type codes be expanded to
include these dates, such as value m?

Go to:

Library of Congress
Library of Congress Help Desk (09/03/98)