ISSUES IN CATALOGING FOR AN ONLINE CATALOG
ILLUSTATED WITH PROBLEMS FOUND
IN CONSULS RECORDS FOR SAINT EXUPERY ITEMS
As
an ever-increasing percentage of patron usage occurs in the online environment,
I think one primary focus of classification and shelflisting must now be to
create an online catalog that can be searched with maximum efficiency. In the case of the Connecticut System, and
any other where multiple physical collections are involved, I believe this
means cataloging must be centralized so that the shelflist is consistent across
all member institutions. Also needed is
an aggressive program to correct discrepancies already in place. Though I realize the cost is prohibitive, I
am not sure that letting things correct themselves over time is adequate to
serve the needs of our patrons.
In pulling records for this assignment, I
found it very frustrating that they were not all in one place. Catalog searching seemed extremely
inefficient. I believe that optimally, a
keyword search on Saint-Exupery should bring up all records in one step,
whether he appears in the 100 field (main author entry), the 245 field (title
or author), the 600 field (subject), or the 700 field (additional author). With that change, and additional
opportunities to limit and organize results, which I will further discuss in
the conclusion, this problem could be eliminated. The new IPACs
exhibit many of these features, so knows that it is possible. The just have to be implemented.
Current inefficiencies and errors meant
that running a search by call number first generated 33 records, while an
author search brought up 53 records.
That indicated a problem right there.
Unless Saint-Exupery had written many books outside the areas of
literature and autobiography, or appeared in a large number of collections of
multiple authors, some errors in classification were certain. Going through the tedious task of organizing
by 001 field, it turned out that four records were themselves duplicates,
caused by differences in classification across libraries. There were only 6 records in the call number
search that were not duplicated in the author search. Given the higher number of books about Saint-Exupery
with other authors in the 100 field (main author entry), this meant that more
errors in classification existed. In
addition, a subject search pulled up six more records. All in all, this made the search process very
time-consuming, and left me feeling very uncertain that I indeed had all the
records even then. I have to say that as
a patron, I experience this uncertainty on a regular basis, no matter what
catalog I am searching.
Other difficulties in the shelflist arise
from the tendency for academic libraries to have two or more classification
systems in place. This causes a lot of
problems. I am also not in favor of
dividing the collection into multiple mini-collections to allow organization by
course or discipline, if it is to be done, item location should be enhanced with
catalog messages in addition to the call number, not as part of it. With new
online catalogs creating one combined collection of several physical ones,
cataloguing and shelflist maintenance must be centralized or at least
coordinated. In the online environment,
this will eliminate duplicate records and the need to have three, four or five
steps in the search of one record.
This assignment uncovered a lot of
errors. I do not mean to indicate that
they are all a result of incompetence.
Now that everyone has done the assignment, I think we all understand
just how time consuming the real life work of creating and maintaining an
accurate shelflist would be. Also, many
of these records are old, probably having been placed into MARC format and
loaded from card catalogs to online systems without content change. The cataloging rules are an evolving system,
so it is hard to say what rules were in effect when those decisions were made.
In making my changes, I took very
seriously the statement: “If the
classification practices of CONSULS are up to appropriate standards, all the
works by and about Saint-Exupery should be in PQ2637.A274.” I made the many changes necessary to make it
so. Many items were in other parts of
the LC classification, or completely outside it. Some professionals believe that certain works
of Saint-Exupery are topical enough to belong outside the system. I, on the other hand, believe that his
intentions are to report his experiences and his existential musings about
them, rather than to accurately describe events, so his works are much more
appropriate within his one cutter designation.
There is support for this viewpoint in the practices of some libraries.
Many of my shelflist changes were created by this content judgment call.
All of Saint-Exupery’s work is to some
extent autobiographical, even Le Petit Prince, and as such, it is often
difficult to know what is fact and what is fiction and
where each work belongs within the one cutter designation. Since some libraries, even in the Connecticut
State System, are placing some of his works in the autobiography, journals,
memoirs and letters areas, I indulged my inclination to do this when I had to
change some or all of the records on a work anyway. However, I do admit that there is some
organization, and therefore browsing advantage to leaving the separate works in
that area, which has different internal organization. Therefore, I did not move any of those titles
that were consistent in that arrangement.
For one thing, to be accurate with this particular would, I think,
require actual scanning of the works, which I felt went beyond the perameters
of this assignment. Many additional
changes just involved adding a date to an otherwise working designation, or
taking out a number that was not needed, or changing a second cutter to better
accommodate new and future additions.
Some things, however, seemed to be
definite errors, or appeared so because of changes in cataloguing
practice. A few records were in close,
but not quite correct designations, leading me to believe that errors in typing
or in reading tables had occurred.
Hopefully, I have not made too many similar errors of my own. Sometimes it was hard to understand how an error
occurred. Pages Choises,
for example, could hardly be both a separate work and have multiple titles
listed in a note. There were also errors
caused by not organizing by original title, or not designating a translated
work, but again, these may be new practices. There was at least one error in
alphabetization.
If we were actually to enact these many
changes, which of course probably are occurring in other designations as well,
the cost would be fairly prohibitive.
However, I think that our intention needs to be to clean up these
problems because they clearly affect the effectiveness of catalog searching and
patron service, and employee job performance on many levels.
COMMENTARY ABOUT SUBJECT HEADINGS
I
believe that many records could be much more helpful to searchers both when
they are searching and when they have obtained the records. Others used are superfluous. The idea of subject headings is to anticipate
the major categories that the item covers, hopefully the same ones that patrons
are likely to search. I have not noted
many of these comments on the chart because they apply to so many records, and
would make the chart unwieldy. I do not
think that one needs to include a French 20th Century Authors subject anywhere
other than to the call number itself.
The organization system lets one know what to expect in this area. On the other hand, patrons searching for
books in French or the other languages represented here will not necessarily
expect to find them in this section.
Another language issue brought up by this
assignment is the mystery created by foreign language titles. I guess as an actual patron, I would be
rejecting these books out of hand anyway, but there must be more instances than
library school assignments, when a non-speaker would like to know generally
what is contained in a foreign language book, so that kind of record should
include a note. In the same vein, the
English translations could use a subject as well.
When I moved the items that had been in
the World War II sections back into the single cutter area, I created a problem
for those people browsing that subject area.
Again, consistent use of the World War II, pilot, and memoir subjects
can help bridge that gap. Good examples
of these subjects are in MARC record of 13003196. Other items that could use the memoir note
with this particular author are the letters and memoirs that are in the section
of the category rather than in the criticism section. The current scheme makes for good browsing of
a particular work, but not of the whole category of Saint-Exupery
criticism.
Also, in the Biography and Criticism area,
since everything is in one alphabet, each record needs a biography or criticism
note, or both, to let viewers of an item know content, and to allow the
creation of sorted results lists. Record 30053699 on the chart has a good
example of both. The notes tell me this
item includes both biographical and critical content. In addition, any criticism item that is filed
with its separate work needs a criticism subject to make a bridge between the
normal area for criticism, and the actual diverse location of those items.
As a side note, if I had been designing
the cataloguing scheme I would not have combined Biography and Criticism in
this way. I would basically have
inverted the lower part of the schedule so that autobiographical material
including letters would have directly followed the “by” section to truly group
all items by the author in one place.
The dictionaries and indexes could easily have come at the end. I know that’s contrary to librarian subject
thinking, but I believe some conventions need to be changed to comply with
logic.
The main improvement that could be made in
the Connecticut State System, and in any library system, would be to choose
good and needed subjects like I have mentioned above, and use them consistently
across any particular subject area that would apply. What I mean is that any area that
unexpectedly has foreign language books in it should have a subject so stating,
and so on. On the other hand, unneeded
subjects, especially to the extent they muddle searches, should be
dropped. I don’t know enough about the
workings of the software to know that if, for example, a name is in the main
entry and in the subject area, if it might cause multiple records in a search,
or mar the understandability of the results list. Any practice that does should definitely be
ceased.
A major goal of the profession should be to
have any search bring up pristine lists of results with no duplicates, and ones
that contain all relative results.
That is certainly not the case in the
Individual errors in the records are
disturbing. I do not understand why some
of the translations do not have 240 fields for original title. It seems pretty basic to me. The one case of a duplicate subject is
laughable. I am sure that the records
contain many too many 500 notes, in my instructor’s your way of thinking, but I
personally do not mind them. Without
going overboard, any content notes that will be helpful to the catalog user
when viewing the record are all right with me.
They in no way impede a good search.
Series notes on the other hand are ridiculous. Cataloguers should use the 490 field or
nothing.
CONCLUSION
This class, and this assignment in
particular, certainly gives the student an appreciation for the importance of
the cataloguing function in creating a browsable collection, both in the
physical realm of the library, and in the online catalog. The profession needs to use strictly
consistent practices to keep the shelf list in correct order, and subject
headings inserted to connect the areas that are connected by subject, but
separated by classification design. As a
patron in the library I am often at effect of widely scattered dual
classification systems, or multiple mini collections. Using online catalogs I
have experienced having to do multiple searches to come up with what should be
a single subject list.
I appreciate the difficulty of the
detailed work of shelf list creation and maintenance, especially after this
assignment, but it is essential. The
function needs to be centralized, or at least coordinated in the various
entities that make up a united catalog, here in the Connecticut State Library
System, and elsewhere. We can not count
on deletion of old records to correct the problem. The shelf list we did is a good example. Many of those records have been in effect for
over half a century.
Another thing that would help this problem
is improved search software for online catalogs. CONSULS’ is behind many in not having the
call number on the initial search page, making an additional step
necessary. One keyword search should
bring up all relevant records, and the gathering should be hierarchical by
field to eliminate duplicate results.
Results need to be limitable, and sortable in
any number of ways: by author, by call number, by publication date, by format,
by institution, and so on. Cataloguers
and other staff members could take advantage of these options as well. If they had been in place, the students in
this class could have brought up all records for this assignment in one search
and asked for shelflist order.
Cataloguers could do the same. Such sorting options would go a long way
toward eliminating many of the problems discussed here. The new IPAC are beginning to offer many of
these features already. Now we just need
to clean up our records so that problems that have plagued us since online
catalogs came into being can become a thing of the past!