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
Abstract
The paper will introduce the library cataloguing and management package LIBERTAS, which
is
being used by more than twenty academic libraries in the UK. These are linked to a central
database, to which records can be added from many sources, e.g. OCLC, British Library, etc.
Although each library could operate the system in isolation, in practice, the availability
of the Joint Academic NETwork (JANET) means that any LIBERTAS user can interrogate not
only
the database, but also the catalogue of any other LIBERTAS user and find him or herself in
a
known environment.
This is particularly helpful in an area like London, where there are ca. 250,000 m of
occupied library shelves within the University, but these are spread over some forty
'schools' and 'institutions'. Ultimately, these will be grouped into just five or six
LIBERTAS catalogues.
One of the major problems of such an exercise is that it brings together a wide range of
cataloguing practice, and even where similar systems are operated, of interpretation. These
problems are particularly acute for cartographic materials, as the necessity for the
publication of the Manual of Interpretation for AACR2 made evident. Clearly, a similar
exercise will be needed by the participants in LIBERTAS.
LIBERTAS, however, also has two specific problems for cartographic materials in its
recognition of numerics, such as scales, and alpha-numerics, such as series or sheet
designations. These will be illustrated in the paper.
If solutions can be found to these problems, it seems likely that LIBERTAS will provide a
reasonably friendly tool for the cataloguing and management of map collections. Since it
uses mainstream computer technology and cataloguing principles, it may well be possible
to
develop it further, linking it with similar systems in the UK or in other countries.
Introduction
In a paper presented to the Map Curator' Group of the British Cartographic Society, Perkins
commented, somewhat scathingly, of map collections: "... implementing, because of external
pressure and as a second best, retrieval systems designed for libraries containing, predominantly,
books. Book-based automation will not help in the digital age."(l)
Leaving aside, for the present the accuracy of such a statement, the present paper will
examine one such book-based scheme, LIBERTAS. LIBERTAS is a product of the large computer
business DEC, operating through a subsidiary, SLS, at Bristol. It is, to quote their
publicity, "... an integrated Library management and information system providing a degree
of flexibity, power and ease of use unique in the field of library computerisation. LIBERTAS ...
has been developed for and
with the help of librarians. This
combined with SLS's record of reliability and professionalism, makes LIBERTAS the perfect
answer for todays' library."(2)
The system operates on VAX computers as a complete library package, with acquisitions,
accounting, periodicals, interlibrary loans, cataloguing and enquiry modules. Each
installation can serve a Local Area Network (LAN), or can be linked via JANET (Joint
Academic Network), PSS (Packet Switching System), IPSS (International Packet Switching
System), or direct links to other LIBERTAS installations.
Thus when cataloguing a new item, for example, the first task is to check whether it has
already been catalogued by one of the other LIBERTAS users and the record is thus already
on
the SLS central database. If not, the record can be claimed from one of the major sources
of
bibliographic data such as Blaise or OCLC. Finally, if all else fails, and this usually only
happens with older material, a new record is created - in UKMarc format. This, together with
records claimed from Blaise or OCLC take their place in the central database, so that other
libraries need not repeat the full cataloguing exercise, although they may need to modify
records which are taken from other sources.
There are, in fact, some thirty libraries, or groups of libraries, using LIBERTAS at
present, including one in Iceland and three in Sweden, and discussions are in hand to extend
this number. In some cases, libraries are grouped together. One example of this is in the
University of London which has some forty
constituent parts, ranging from the multi-faculty and multi-site major colleges such as
King's College or University College to the specialist post-graduate institutions such as
the Institute of Germanic Studies or the Institute of Psychiatry. At present there are five
LIBERTAS installations within the University serving more than a dozen institutions, and
others will be joining in due course. Like the other libraries using LIBERTAS, each of these
has its own system and catalogue, but has access to other LIBERTAS users and to the central
database.
In general, it seems to the author of this paper, both as a user and as a member of staff
involved in routine circulation transactions, that LIBERTAS is satisfactory. Of course,
there are difficulties as with any computer system, some of which are really quite
significant, but most have been, or are being, rectified by modifications to the software.
For cartographic materials, two difficulties seem, at present, to be of particular
significance, and these will be mentioned below. However, it is necessary first to comment
on the need for standardising input.
Common problems
At least in Britain, there is a shortage of map cataloguers who are familiar with the MARC
format. There are, it is true, those who catalogue maps in a variety of systems, many
derived from GSGS 3906 or its predecessor 'Parsons', but who have little experience of MARC
and there are also those who catalogue monographs to MARC standards, but who have little
experience of maps. Consequently, there are obvious potential dangers in a co-operative
catalogue of this nature. These are increased by the range of search headings which a map
user may require compared with a book reader. The nature and extent of the specific
difficulties posed by cartographic materials is well illustrated by the fact that a 250 page
'Manual of Interpretation for AACR2' had to be prepared to enable standard use of the AACR2
rules. While the interpretations are widely used by national libraries and major collections
in association with AACR2, thus ensuring complementarity, if not conformity, cataloguers in
smaller institutions are not necessarily so consistent when working on cartographic
materials.
LIBERTAS is MARC based, and hence is a relatively unforgiving environment for cataloguers as
Table 1 demonstrates. Only that part of the brief entry which actually provides the order of
the entries is shown. It will be seen that, in the absence of a series title in the
cataloguing record, sheets from different series have been grouped together in alphabetical
order of their sheet title. Whether this resulted from a conscious policy decision or an
unfamiliarity with MARC is not known, but it is quite likely that another collection might
want to use a different order, either by series and sheet name, (Table 2) or by sheet
number, or date, within a series. (Table 3)
LIBERTAS is enquiry driven, and Boolean searches are permitted. A ranking process ensures
that items closest to the enquiry are given first. Thus in a subject search for "maps" or
"Kenya", all records including "maps" and "Kenya" in field 651 (or 650) will be displayed,
followed by those records which have "maps" or "Kenya". The order in which the entries are
displayed on the screen is controlled by the order of the elements in the 245$a subfield.
(See Table 4) This order can be changed, it should be noted, by the addition or omission of
additional elements in this subfield. While some editing of entries by different libraries
is inevitable, this can be kept to a minimum if the initial record which goes into the
database is as full as possible and the elements are entered in an agreed order to a common
minimum standard.
Such an agreed order and common standard will rely on cooperation between map-cataloguers
and curators. However, they will have to rely on SLS, however to make the necessary software
changes to cater for some of the specific features of cartographic materials.
Cartographic problems
The first of these is scale. One of the ways that map records, whether sheet or series, are
often required to be ordered is by scale. LIBERTAS sorts all numbers in a digit-by-digit
procedure. This is ideal for dates,
e.g.: 1700 1720 1723 1724 1730 1770 1777 1800 1850 1940,
and also for decimally ordered class numbers,
e.g.: D16 D16.12 D16.16 D16.2 D16.25 D16.255 D16.4 D17 D21.
However, when it comes to scales, a whole number procedure must be used. Thus the map curator and map user require the order:
e.g.: 1:4,000 1:50,000 1:62,500 1:100,000 1:250,000 1:7,500,000
whereas LIBERTAS would give the order:
1:100,000 1:250,000 1:4,000 1:50,000 1:62,500 1:7,500,000
The only way to get the preferred order at present is to 'pack' the scales with redundant zeros (e.g. 1:0004000 or 1:0062500), which is clearly unacceptable.
The second requirement is the ability to handle a variety of numbering systems, at least alphabetical, numerical and Roman numerical. The last of these is, at present the most problematical, for the appearance of the sequence:
e.g.: III IV LII VII XLI XLIII XVIII XXII XXIX XXV
produced by treating the characters as alphabetical not numerical, instead of the sequence:
III IV VII XVIII XXII XXV XXIX XLI XLIII LII
is near enough random as makes no difference. To complicate the matter, alphabetical, numerical and Roman numerical characters are often used in combination, as in the large- scale 'county series' maps of Great Britain, e.g. Surrey XXVII-11, which must be distinguished from Surrey XXVII-2, as well as from Hampshire XXVII-11 and also from Surrey XXVII-NE, the change of suffix indicating a change of scale as in the Canadian national topographic system notations:
31 31SW 31G 31B/NE 31A/8 31C/4a
(1:1,000,000 1:500,000 1:250,000 1:125,000 1:50,000 1:25,000)It may be argued that these are sheet designations, and that refinements of the system to this level are not necessary. However, it must be noted that firstly, series designations can be as complicated (e.g., the Spanish series:
Serie 5V Serie L Serie C Serie 2C Serie 4C
or the German series:
TuK200 TK50Str TK500H M745 SK50K TK25Nw TK25A DGK5
to give but a sample). Secondly, some small collections, and some national collections,
already catalogue at sheet level, and may well wish to continue to do so; it is important
that LIBERTAS (or any other system) is able to permit this. Thirdly, it is essential that
any system which is entered does not straightjacket the curator and cataloguer so that
cataloguing can only ever occur at series level. It must be at least possible that in the
future, even if not now, the technology will exist which renders the old and illogical
division into sheets and series entirely obsolete. Map curators will then be able to cope
with the item that our users continually ask for, that is, the map.
Since such technological advances cannot be far away, indeed some would argue that they
are
already available, the immediate need is for SLS to solve the admittedly minor difficulties
outlined above. These short term goals will allow a step-by-step exploration of LIBERTAS as
a reasonably friendly and effective map library, rather than book library, cataloguing and
management system.
In the longer term, an environment which is even more appropriate to the needs of users,
including, for example, graphic-based searches, rather than text-based searches, and links
to other systems in the United Kingdom or elsewhere should be possible. One of the more
exciting possibilities in this area would be a link to Cartonet-Vax, the map cataloguing
package developed at the University of Edinburgh.
Whatever the way forward, it will be important not to be seduced by the enthusiasts into
developing a system which is too complex for the user to gain the full benefit. It must
never be forgotten that the whole purpose of LIBERTAS, or any other automated system, is to
provide a better service to more users, and the purpose of the developments outlined in this
paper are intended to enable the extension
of that service to include cartographic materials.
References
(1) Perkins, C. R., 1990 : Series level or sheet level?, unpublished paper presented to the Annual Meeting of the Map Curators' Group of the British Cartographic Society