|
|
Patrick Le
Boeuf, ELAG Conference, Prague, June 6th, 2001 |
FRBR: toward
some practical experimentation in ELAG?
I. Context
Since the very beginning, ELAG has
been interested in the IFLA four-level model FRBR. There has been an FRBR
workshop in ELAG since 1996 and Susanna Peruginelli reported, on the occasion
of a two-day conference entirely devoted to FRBR, in January 2000 in Florence,
that ELAG regarded FRBR as not only of a high theoretical value, but also a
practical one, [
] making it possible to integrate [digital resources] with
traditional material; [
] searching and retrieval functionality will be
improved.[1]
Hence our wish to develop an
experimental database, within ELAG, that would allow us to value more precisely
the benefits the whole library community might expect from this new,
revolutionary model. We also want to know if this model would not raise
implementation problems; we must think of cataloguers comfort and, of course,
or our patrons comfort when navigating, in the future, a new catalogue
entirely developed according to the model.
Paula Goossens has therefore
elaborated Guidelines to help
workshop attendants to create new records. The aim is not just to transcode
pre-existing records, but to create new ones. Weve tried to get totally rid of
the MARC structure: our future experimental database is intended to be entirely
designed in XML from the beginning.
Only four bibliographic families
have been elaborated so far: it is obviously not enough for a database to be
implemented, but it is a beginning, and it already presents us with some
interesting cases. It would be too long to report in detail on all of these
four families, Ill therefore introduce only three of them to you.
II. Three examples of bibliographic families
a) A
unicellular family: The story of my life
Figure 1
This kind of family is far from
uncommon, especially for National Bibliographic Agencies, whose task is to
describe the whole production of a country. Many Works exist in only one version (= Expression) and are published (= Manifestation) only once, and they have neither parent works nor
sibling works. In such a case, the complexity of FRBR results in much
redundancy: the same title appears at the three upper levels: Work, Expression and Manifestation.
We must spare our cataloguers time
and nerves, and carefully designed by default proposals should prevent them
from having to type identical data several times for one document.
b) Boris
Godunov: the Pushkin play, the libretto, the opera
Figure 2
This bibliographic family involves
rather complex bibliographic relationships between different kinds of materials
(text, musical notation, sound recording, graphic material
) and mingles
different alphabets. It highlights a difficulty which the FRBR Final Report had left unsolved: what
treatment should be applied to creations which are regarded as secondary
(forewords, illustrations
) in comparison to what we call main entry, but
which have nevertheless their own existence and should not be held less
important than anything else in a document, especially in the context of
integrated heterogeneous resources?
c) Coronellis
Terrestrial Globe
Figure 3
This bibliographic family involves
an even greater number of different kinds of materials: cartographic materials
(both 2D and 3D), manuscripts, electronic resources (CD-ROM and website). It
features an interesting example of a direct reproduction relationship (taken
into account in the FRBR model but rarely encountered in current catalogues)
from Item to Work. It contains links to websites.
III. Future developments
Four bibliographic families only is
not enough, even for an experimental database. Creating these records does take a lot of time. We therefore
urgently need volunteers
We also need a physical place for
our experiments, a computer, a software
There is an opportunity to use the
VisualCat structures from the Danish Bibliographic Centre (thanks to Poul
Henrik Jørgensen!).
Then well have to start with the
practical implementation of an ELAG experimental database, which will focus on
two main topics: the ergonomics of the catalogue production system (cf. the aforementioned redundancy in the
Story of my life example!) and the retrieving facilities for final users (how
to navigate the catalogue, how to make the FRBR complexity simple, how to make
its peculiar terminology understandable, etc.: in one word: OPAC issues).
IV. Some reflections on OPACs
Last year in Paris, Dan Matei
launched a great idea: We must get rid of traditional Author/Title/Subject
OPACs! This motto raises two questions:
Would
it benefit our patrons?
If
yes, how could this be achieved?
I dont have the answer to the first
question, but here is a proposal draft, just in case the answer would be yes.
I dont claim in any way that switching from our current catalogues to such a
structure would prove useful and beneficial: we have to make experiments if we
want to make progress, and experiments may of course sometimes be unsuccessful.
And I insist this is only a proposal, and a draft: there is still
much work to put into it, and it needs to be discussed, valuated, put into
question by several people with different ideas, which stimulating
contradiction would help us to make progress as a group.
Figure 4
Under Person or Corporate Body, we
could choose between: Person or Corporate Body, Person only, and Corporate
Body only. Under Work, we could choose between: Work, Monograph only, Periodical only, Series
only, Expression, Performance, and Manifestation. Under Object, we could
choose between: Object, Music instrument and/or voice, Scientific instrument,
and Item. Under Concept, we could choose between: Concept, Place, Event, and
Form/Genre. Under Date of manifestation, we could choose between: Date of
manifestation, Date of work, Date of expression, Date of performance, and Date
of event. Under Place of manifestation, we could choose between: Place of
manifestation, and Place of performance. Under Ť Language of
expression ť, we could choose betwen : Language of expression, and
Language of performance. Under Mathematical data, we could choose among the
whole range of mathematical data which are required in cataloguing cartographic
materials.
After selecting, for instance, one
name of person, we would have the following structure:
Figure 5
This proposal raises the problem of
the FRBR terminology, with which our patrons are far from familiar. How
could we translate the names of the FRBR Group 1 entities (Work,
Expression, Manifestation, Item) into common language?
V. Conclusion
We need:
Imagination
Volunteers
Funds.