Re: Functioning IRs - today's real realities

From: Arthur Sale <>
Date: Wed, 14 Dec 2005 10:11:48 +1100

To AMSci Forum readers


Before any more people tell me I am uncultured, let me explain the
situation more carefully than I did at first.


The phrase 'institutional repository' coexists with the phrase 'digital
library'. They normally mean different things to different people, and
what prompted my remark was yet another example of the phrases being used
interchangeably. We have often seen how misuse of terms leads to faulty
conclusions by analogy, especially on this forum!



The first is usually considered to mean a place where one deposits (hence
repository) born-digital objects such as peer-reviewed research
manuscripts for the purpose of providing open access. The born-digital
object needs to be captured at creation time otherwise it is soon lost.
The technology and the software for doing this is well-known and mature.
Copyright issues are well-understood. Preservation is also not an issue.
It costs tiny amounts of money (well under $US10,000) and only a small
amount of effort to create and run a repository of this sort. It runs
virtually out of the box. However, of most importance of all, an OA IR is
also an essential function of a university or a research institution. Any
institution that does not have a fully functional IR of this kind with
appropriate requirement to deposit and author support must at the present
time be considered to be failing to meet its basic standards of public
accountability. It also has universal impact in the research activity:
all researchers from the arts through the humanities to the life sciences
and the hard sciences benefit from an OA IR, and from a global system of


Hence my statement about priority, from which I do not resile. Any
university without an OA IR is somewhat like a [book] library without a
[book] circulation system. It is hard to imagine where their priorities
might be when essential services are missing. Working on other issues
while failing to address basic standards is professional neglect of a
librarian&#8217;s duty. I am sorry if I offend anybody, but that&#8217;s
my view.



Digital library applications on the other hand are far more expensive,
largely because the technology is not well-understood, and all of the
possible applications have a tinge or a lot of research in them. There is
no software that will do all of what everyone wants. What there is needs
work done to it to adapt it. Layering digital library functions onto an
OA IR is often not sensible, and generally shows confused thinking
perhaps caused by the confusion of terminology. Such applications are not
core business in the same sense as supporting undergraduate teaching is,
nor supporting research generally. They benefit a subset of researchers
(or teachers and learners) usually, though the minority may be vocal,
silent, large or tiny.


If I just pick out one of these digital library functions to discuss in a
bit more detail, take a collection of culturally significant artefacts,
documents and photographs pertaining to an Aboriginal tribe in Australia,
going back to early settlement times. Clearly digitizing this collection
could be considered to be a desirable target, but the digital library
involved has to deal with:

o        objects which are text, images, audio, video, 3-D
representations of objects, possibly X-rays, ultrasound and computer
tomography, as well as multi-media syntheses of any of these. Some
objects may be interactive, eg 3D views, panoramas, false color, etc.

o        Key language records.

o        Inter-relationships between the objects, including multiple
instantiations of some (say digital &#8216;prints&#8217; taken from a
negative with different gamma to reveal details in the shadows).

o        Inter-relations with other collections.

o        Provenance.

o        Complex &#8216;copyright&#8217; (or ownership) issues with each
object, some of which may be Crown copyright, some personal, some tribal,
some women&#8217;s business, etc

o        Complex metadata, often per object or sub-object.

o        Cultural sensitivities about who is allowed to see or hear what,
and maybe when. The restrictions may be very complex and need to be
tailored to the situation.

o        Preservation of this diverse range of digital objects cannot be
guaranteed; the best that can be done is to arrange for the long-term
preservation of some and regard the others as ephemeral but valuable
while they exist online.


Is anyone in any doubt that this might be core business for a museum, and
important business for a university department or a researcher, but it is
not the same as core business for a university? Is there any doubt that
the costs will be high, in the $100k to $1M range, and that much of the
work will be &#8216;R&D&#8217; in itself? This is not a mature field and
standards are missing. Data is not mutually harvestable. People are
feeling their way and trying to come up with reasonable (approximate)


So I return full circle. Working on digital library functions to the
exclusion of open access repositories is simply indefensible. Working on
both, with the OA IR being the priority, is defensible. Better still is
to do the OA IR and get it out of the way because it is such a small job.


Received on Wed Dec 14 2005 - 01:02:10 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:08 GMT