From:   MYRIAD::WATTERS      25-JAN-1994 13:08:26.40
To:     IN%"ILL-L@UVMVM.BITNET", "AUTOCAT@UBVM.BITNET"
CC:     WATTERS
Subj:   LC Series Group report: to trace or not to trace

As a guest participant in the LC Series Group, I prepared an appendix to the
report that attempted to outline the implications of adopting such a proposal.
This appendix was included in the report submitted to the LC Cataloging
administration.  However, it has been omitted from the report disseminated
through various list-servs because it resided on my computer (which
left Washington with me) and was not available to LC in machine-readable form.
 Therefore, I will attempt to distribute it myself.  I am sending it to
Autocat and the ILL list; if you know of other appropriate lists, please
feel free to forward it to them.
 --Cynthia Watters, Middlebury College, watters@myriad.lib.middlebury.edu

                Series, traced or untraced : a view from an observer
                                                                --Cynthia
Watters, Middlebury College

What is meant by not tracing series?  What is lost or gained?  What are the
ramifications for the Library of Congress, other libraries, and library users
across the country and beyond?

Bibliographic implications:
1.  Loss of series searchability
        Series statements would remain in the bibliographic record transcribed
exactly as they appear on the item but in a field defined by the MARC format
as "untraced" (490 0).

        Untraced, in the card catalog environment means no card is produced
and filed under that series title.  Some online systems and some libraries
have chosen to index this field; others, equating untraced with unindexed,
have not.
         Changing a system to index the 490 requires requires reindexing the
database and is likely expensive, perhaps impossible.  A library may need to
change the field tag (from 490 to 440) in each record during cataloging to
make the series title searchable.

        The 490 field lacks non-filing indicators so, even if the field is
indexed, a search may need to be done both with and without the initial
article; alternatively libraries may wish to remove initial articles from the
series statement during cataloging.

        Keyword searching may bypass some of these difficulties.

2. Loss of collocation
        Lack of an authorized form means that even if the 490 0 field is
indexed, the volumes of a series will not appear together in response to a
search unless the publisher and the catalogers have recorded the series title
on the item and in the record in exactly the same form for each volume.

        Various series with the same title will be intermixed in response to a
search.  For titles such as "Papers" or "Bulletin" the number of hits will
likely exceed the patron's capacity for browsing.

3. Loss of see references
        Without authority records, the catalog will not refer the user from
the form searched to the form used in the catalog.  These forms vary for
series titles that a) appear inconsistently from volume to volume, b)  change
title, c)  are searched by the corporate body issuing them (e.g. Geological
Survey. Bulletin  see Bulletin (Geological Survey)), or d)  are subseries

Other implications:
        An advantage of the proposal is that tracing or not tracing a series
is considered a local decision.  Thus, LC can implement the proposed policy
without violating cataloging rules or utility guidelines, and the policy does
not compromise the fullness of the record as indicated in the encoding level.

        Gains are seen to be chiefly in the area of saving the LC cataloger's
time. Series are a clear target area for simplification because 1) series
authority work takes a significant amount of time, and 2) the majority of
series are unlikely to be significant access points.

        Some series, however, are significant access points.  Sometimes the
series title is the only title a user knows for the work; some are cited in
bibliographies and indexes by series title and volume number only.  Separate
cataloging under each volume title has allowed the twin advantages of
retrieval by series title yet with subject access to contents of individual
volumes, thus serving the needs of the neophyte searching by subject as well
as the professional in the field and the user with specific citation.

        If series are to be controlled, a national authority file (whether
created by LC or by NACO-like cooperation) is the most cost-efficient.  The
costs for each individual library to create and maintain separate authority
files and to add series tracings to each LC record would be very high.