Back to GdC Homepage
Ligue des Bibliothèques Européennes de Recherche,
Groupe des Cartothécaires de LIBER
TRA NSLATE ENGLISH to Français, Deutsch, Italiano, Português, Español! Explanation
© LIBER and author
Introduction.
Six years ago at the 4th LIBER-GdC Conference in The Hague, when I put to the meeting the
question 'Shall we catalogue the map or classify its information', the conclusion was reached that it
was not a matter of 'either-or', but that both need to be done. (1) Consequently this statement has
also been the starting point of the policy followed by the Leiden University Library with regard to
the automation of the catalogue of its collection of old maps, the Bodel Nijenhuis Collection: the
building up of a basic bibliographic database with the purpose of providing extensive retrieval
facilities.
In the present paper it will be explained why the Leiden University Library gave preference to the
Pica automated cataloguing system over the CCK (Dutch Union Map Catalogue) and - secondly -
how the experience with this system was.
Why the Pica-format ?
From the very beginning in the early seventies, the start of the pilot-project CCK went smooth and
full of promise. As member of the Board of the project and participant in the project I witnessed its
birth, youth and growing up. But reaching its manhood in the early eighties there was a slowing
down in its development; the efforts to find an organisational shelter for the operational CCK,
among others within Pica, did not yield the desired efforts. By this development the Leiden
University Library was placed in a tricky dilemma of making a choice between two systems, which -
as has been proved at that moment - could not be combined. Was it the CCK-project, which had at
that moment a very uncertain future in spite of its established qualities, or should we decide in
favour of the format for a bibliographic map-record, which was designed in a great hurry by Pica,
but not yet tested in a sufficient way.
What does Pica stand for? This 'Project for Integrated Catalogue Automation' started as a joint
research project of automated shared cataloguing of some large libraries and has become since
1976 a comprehensive all library system, in which most Dutch libraries are participating now. The
Pica Library Automation Network provides not only an on-line cataloguing support service, used for
the input, update and output of descriptive records and authority fields, but also sub- or local
systems for acquisition, lending and storage administrations and on-line search service for the
public. Pica uses a map-format, which is an adaptation of the Marc based Pica book-format with all
the implications of that. In his turn the book-format is an adapted, simplified or - if you want - a
curtailed Marc format. So the user of the cataloguing facility works without subfields and indicators
For the purpose of conversion to the US-Marc format, Pica uses a second internal system which
converts the punctuation in the required subfields.
In the actual situation of autumn 1984 - as mentioned above -there was no other choice than this
Pica-format for maps, because Pica was omnipresent in our library, standing by for some systems,
for others still in development. The consideration that the input of the title records and output of
hardcopy cards or lists were ready to hand and should not involve extra expenses for the library,
because they were covered by the Pica-subscription of the library, counted heavily in the decision.
The main reason however was that our experience with the brand-new Pica format for maps gave
us sufficient confidence in its possibilities with regard to bibliographic control and retrieval facilities,
as required for a collection of old cartographical material.
The unique and exact bibliographic description as basis.
The Leiden map collection (50.000 items) consists for 90 percent of material made before 1872, the
year in which Bodel Nijenhuis bequeathed his maps and atlases to our university. It is used
primarily in relation to historical research of very diverse character. To make these original
documents accessible for historical research an exact bibliographic description is wanted, by which
they can be identified. Identification of the source material is a first requirement of any historical
research. It applies no less to printed documents than to written ones, no less to maps than to
manuscripts or books. The base of our computer produced map-catalogue must therefore be the
unique and exact bibliographic record.
This premise squares very well with the principle of shared cataloguing of the Pica-system, that
there should be only one record per object or bibliographic entity. In addition Pica has got ample
experience in the last years with two other automated cataloguing projects of older historical source
material, which had been developed by the Dutch research libraries, namely for the books issued
before 1800 and for the learned correspondence-project or central catalogue of letters, once started
as a local effort by the Leiden University Library.
Although emphasizing the bibliographic descriptive record as the base, I am well aware of the fact
that the computer is a machine that makes it easier not only to produce a catalogue, but
above all to manipulate the information stored in the database. Besides the possibility of
a more efficient administration and management, the retrieval service of information is indeed the
big gain of the automation revolution.
Evaluation.
Now, six years after the decision in favour of Pica, the state of the art of automation of
cartographical information in the Netherlands, had been changed considerably. The attempt to
incorporate the CCK into the Pica-system by a complete conversion of its database and its thesauri
was foredoomed to failure by reason that the differences of the two formats were at that moment
unbridgeable. The failure of this experiment opened the eyes of the responsible authorities, who
decided at long last, that the CCK should be an independent system under the shelter of our Royal
Library - since 1982 appointed as national library. It was a wise decision, but it brought also in its
train that the interest of Pica in the cataloguing of maps had vanished all but completely. After 1984
the Legal Deposit Department of the Royal Library still continued for some years its map
cataloguing within the Pica-system, but moved definitively to the CCK in 1989, leaving only three
collections, which catalogue maps within the Pica-system: the University Library at Groningen, the
Royal Institute of Linguistics and Anthropology and the University Library, both at Leiden. This
situation caused a lot of trouble: because of the lack of interest on the part of Pica, we had had to
wait four years, until at long last a prooflist of one hundred title-records without major mistakes
was delivered. The greatest problem to overcome appeared to be the multi-level description.
In the meantime 1200 written input concepts, made originally in the years 1978-1980 on behalf of
the CCK, had been put in the Pica-database after a manual conversion. In May 1990 we got at long
last the first printed hardcopy of the regional classified catalogue of these 1200 maps, which could
be released to the public.
In spite of these considerable problems with the output, we still stand behind the decision of 1984
in favour of the Pica-system. The same considerations count equally heavily now as in 1984: the
library leans even more for all her automated technical service functions on the Pica Automated
Library Network, the cataloguing of maps within Pica does not entail extra expenses for the library,
but the positive evaluation of the 1200 map-records settled the matter ultimately.
Planning the future.
The present state of the art has not yet much to offer. The only computer produced map-catalogue,
now available to the public, is the area-classified one, for which is chosen the area classification
according to administrative division, as designed by the Bureau Centrale Catalogus Kartografie (the
CCK Office). Hopefully within a few months a second will follow: the alphabetical catalogue of
names of persons and corporations. These two traditional forms of hardcopy printed output will be
delivered by Pica every year in January with the production of the preceding year added to it. These
two catalogues, however, are an interim solution, for we plan as definitive forms of our catalogue 1)
the on-line retrieval facility, which is already available for books and 2) the printed catalogue in
bookform. 'Coming on-line' with all the benefits of retrieval services for the public will be
accomplished within the foreseeable future. The definitions for the primary and secondary
searchkeys have already been discussed with Pica. When once the OPAC is operational, maps can
be searched by area-name or code, co-ordinates, personal or corporate name and words from the
title, and in the second instance by scale, subject, year or form and material qualities, e.g.
manuscript or printed maps on vellum.
The third -and for research outside the library perhaps most important form of our catalogue- will
be the printed catalogue in bookform, also produced from the records in the Pica database. The
library tries to interest some commercial publishers in this project, to all appearance already with
some success. It will be a geographical classified catalogue, using the area classification of CCK
with indexes of names and geographical keywords.
From the day that the Pica-system became operational for the input, maintenance and output of
map-records, the cataloguing came to belong again to the daily routine. The retrospective
cataloguing of the map collection, which implies that as good as every record should be
supplemented, amended or corrected has gone ahead definitely.
Unsolved implications.
So far mainly the positive aspects and consequences of the choice of the Pica-system have been
discussed. The user of the Leiden mapcollection will indeed get amply his share with the
information delivered by these automated catalogues. But what will happen when he wants
information he could not get from the maps in our collection.
Libraries, especially research libraries, are not longer self-sufficient institutions on isolated islands.
They have rather become points of access to the information in all of them. In this view co-
ordination, co-operation and exchange are the catchwords to overcome the abundance of published
information and - not less urgent - to take full advantage of the computer assisted delivery of
information.
Well then, looking for possibilities of cooperation at national and international level, the conclusion
is inevitable from the viewpoint as mapcataloguer in Pica, that the outlook is not promising, at least
in the short term. Of course there are a lot of new facilities, owing to working within the automated
network of Pica. But on account of the realization recently of the CCK without any connection with
this network, the Leiden University Library found herself confronted with the handicap that a direct
transfer of the Pica title-records to the CCK is not yet possible, because of the different formats of
the two systems. From the technical point of view this drawback can not be unsurmountable, for
Pica already has a daily experience with the conversion of title-records of books into other Marc
formats through the medium of an internal conversion programme. The main problem however will
be the price-label attached to this conversion. Who has to pay? Likewise an interface, which will
give the mapcataloguer in Pica access to the database of the CCK, can be implemented technically,
but it belongs as yet financially to the impossibilities, too.
Even more urgent is the availability of the CCK-authority files for the use by the Pica-cataloguer,
for, when one day the Pica title-records have been converted to the CCK database, they can only be
accessible by way of these files. The CCK very properly points out to this condition, but a logical
consequence should be, that the Royal Library as responsible administrator of the CCK agrees in
principle with that availability.
Therefore a warning message to the responsible authorities to remover these obstacles is in order.
For, when we will not succeed in making the connection with the national catalogue of maps
(CCK), the real fruits of the automation of the map-catalogue in the Leiden University Library will
only be reaped by the users of the Pica Network.
Notes
1) D. de Vries: Shall we catalogue the maps or classify its information ? In: LIBER Bulletin 28
(1986), p. 6-10.