Chapter 13 – Archiving challenges
333
regard into access policies, i.e. to define the type of protocol that has to be
followed by users of the resource and the guidelines on how to use it.
In addition, an archive has to take measures for the time when the main
depositor who knows the local situation and expectations is not available
anymore. Here an advisory board consisting of experienced field workers
and archivists may be of help. Alternatively, archives may require deposi-
tors to identify substitutes in case they themselves are no longer able to
determine (changes in) access policies.
6.2. Access management systems
“Access management” refers to a system that implements the above men-
tioned access policies. It should be obvious that an elaborated access man-
agement system is needed which, however, has to be tractable even when
confronted with an increasing number of access requests. Also from the
users’ point of view, the bureaucratic effort for granting access to selected
material has to be minimal, otherwise many potential users will be turned
away (for example, journalists seeking fast and easy access to materials of
use in writing a piece on linguistic or cultural heritage). What we need, then,
are efficient electronic ways of dealing with access requests. Currently, the
first steps are being undertaken in exploring how this can be achieved.
In the DoBeS program a web-based access management system was
developed which allows the delegation of rights to grant access permissions
to other people, such as the responsible researchers or depositors. Author-
ized persons such as the archivist and the depositor can define rights to-
gether by selecting a node in the linked metadata hierarchy and a resource
type. In this way, only one single command is necessary to indicate that all
textual materials of a given documentation project are open to the public.
The system also allows one to demand the acceptance of declarations and
to indicate the intended use of the material being accessed.
Clearly, though, no access system can fully enforce the proper usage of
archival materials. Only social control within the community of users can
prevent that material is being used for purposes other than requested.
334
Paul Trilsbeek and Peter Wittenburg
7. Outlook
It is highly desirable that the metadata descriptions contained in different
archives be included in worldwide accessible browse and search domains
so that all users can inform themselves about the types of resources that are
available, irrespective of the archive they are used to dealing with. It is a
widely accepted principle that the metadata descriptions have to be open to
the general public. Projects in Europe, the U.S., and Australia are presently
working on such an integration. The integrated IMDI and OLAC metadata
domain now covers about 80 institutions worldwide, which is a promising
start.
One of the goals of the DELAMAN initiative, in which a number of
major archives are collaborating, is to virtually integrate the archives such
that users can work with a unique identity and sign-on on all resources to
which access has been granted. The individual archives that house the
original copy of the resource will remain the access granting authority. It is
expected that such collaborations will not only radically simplify the ar-
chives’ access management, but also make it possible for interested users to
navigate in these new virtual archives without bureaucratic limitations.
Within the DELAMAN initiative, copies of data will be distributed among
the different participating archives to increase the probability of long-term
survival. Projects implementing this form of archive integration which goes
far beyond the metadata integration have started recently, i.e. results can be
expected within a few years.
For the utilization of complete language archives (or larger segments of
an archive, involving data from different documentation projects) the lack
of interoperability on various levels may create major problems. As was
already indicated above, at the technical encoding and format level a high
degree of coherence within an archive can and should be achieved. How-
ever, at the level of linguistic annotation we will continue to be faced with
different terminologies. Depending on the language, differing descriptive
traditions, and preferences with regard to linguistic theory, linguists will
continue to define their layers of annotations, their lexical attributes, and
the values these may take. The resulting differences in terminology and
annotation schemes will, e.g., limit the scope of searches. The new web-
based utilization frameworks mentioned in Section 5.2 above have to ad-
dress this problem. Various initiatives exist for developing flexible frame-
works that will make use of ontologies that are either generated in a bottom-
up fashion, i.e. driven by the actual resource selection, or created by top-
Chapter 13 – Archiving challenges
335
down methods determined by linguistic theory. With respect to the latter,
we can mention initiatives such as ISO TC37/SC4 and GOLD. While ISO
proposes a central data category registry that is an almost flat and compre-
hensive list of linguistic concepts, the GOLD project wants to develop an
ontology covering both definitions and relations.