RE: SPAM-LOW: RE: Question regarding ILL

["David Goodman" <David.Goodman@liu.edu>]

I shall explain from the user's point of view why ILL or Document Delivery
in general is not suitable for important journals in research libraries.

When a researcher reads an article in his own field of specialization, one
of the first things looked at is generally the bibliography. (And the same
goes for a computer search within his field.) The researcher naturally
expects to be familiar with most of the articles, and probably has copies
of them. Some will be new to him--they may represent important work he has
missed, they may represent obscure but relevant work from sources he does
not follow, and some may be irrelevant altogether, though often not
obviously so from the title.

The next step is to look at the articles--typically every one that is new
to him, but might be of value. This is easily done if the library owns the
titles in electronic format. If so, he scans them, discards most of the
items, and marks or keeps copies of the few worth keeping. It will not
surprise him if the library does not own something particularly
obscure--if it looks really important he'll ask for an ILL copy, and he
will expect a research library to go to any lengths to obtain one. If it
does not seem critical, and especially if it is in a language he does not
read, he ignores it unless he is preparing a systematic review or a Ph.D
thesis or the like.

If any of the articles from what he knows to be significant journals are
not held by the library, he usually asks for them too, because he cannot
judge their quality and relevance without seeing them. When the library
gets them, he takes one look, appropriately throws most of them out, and
keeps the few relevant.  This is a decidedly slower way of working, and
why good libraries exist. A scholar who must depend on ILL for his basic
material can not do the bibliographic work quickly, and most of the work
and cost of ILL will furthermore be wasted.

I shouldn't limit it to science. Any humanities scholar wants to follow up
the references, and this is only practical if most of them are available.  
The nature of the humanities makes this possible only in the best
libraries, and the scholar will travel to one if necessary.

I also shouldn't limit this to the trained scientist.  He at least has
some idea of what articles are likely to be of importance. The beginning
student doesn't, and if he gets a reference list from a search and is
conscientious, he looks for them all.  If they are all at hand, this is
quick and easy and he will soon learn from looking at them which ones are
important. If they are not there, he will probably be able to get only a
few in time for his project, and he probably will not choose wisely

This also applies to non-academics.  For some years, I supervised an ILL
operation in biology. We received requests from public library patrons or
people at small colleges for papers in biomedicine where it was obvious at
first inspection that they would most likely not be suitable. (Yes, we
filled them. I had not yet come to consider it my responsibility to reform
the dissemination of STM information.) If the public library instead had
access to the journals, the layman would have been able to see what did
and did not meet his needs, and would ask for help in finding material he
would understand and would be to the point.

For everyone at every level in every subject , (I am aware of possible
over-generalization) some material can be ruled out by title and abstract
alone, and a good librarian should be able to help the beginner know what
to look for. But almost no article can be known to be useful without being
seen.

This all applies even to the best ILL operations (such as the one that I
had charge of.)  Practical ILL operations do produce the waste and
inefficiencies that Dean Anderson deplores. But books can not be obtained
overnight -- though this seems to be changing. Obtaining articles from
obscure sources is neither quick nor easy.  Handling large numbers of
items as rush requests is not feasible.  I could if really needed get
material in an hour from friends in other libraries, and the patron would
understand that I could get one item that way, not one hundred.

I note that most of the above does not apply to electronic purchase by the
article, where the patron may not even be aware that the material is not
actually owned. Different factors apply here, and it is just a matter of
balancing which method is the lower cost for the particular title and
library. If the patron is aware that each request costs the library $12 or
so, he would ask for fewer items. But, as argued above, he might not ask
for the right ones.

------

I know the long articles analyzing this in formal economic terms, which
involves assigning quantitative values to the time saved or wasted by the
different users and the staff.  My point is, that except possibly for
instantaneous electronic access, the time is not appropriately measured in
financial terms. An undergraduate has 4 years to get a degree; a faculty
member needs to have a certain number of articles per year. The role of
the library is to help them both do the best work they can by removing
barriers to access, whether involving delay, cost, convenience, or need
for an intermediary.

This argument would have applied equally well to the pre-electronic
days--but then logistics did make it impossible for most libraries to have
immediate access to all potentially relevant material. It is totally
different now.  It is perfectly possible for every library user in most
countries to have access immediately, without additional cost or delay, to
all material available in electronic format.  The only cost would be extra
servers, and that is not the major cost factor in publishing. Even if it
were, Moore's law would inevitably make it increasingly trivial.  
Successive electronic systems may cost as much, but they give much higher
output. Just like journal publishing, and libraries.

Dr. David Goodman
Associate Professor
Palmer School of Library and Information Science
Long Island University
dgoodman@liu.edu

This essay is intended for more formal publication in extended form, and I
would be grateful for off-list suggestions about what aspects need to be
explained in greater depth.  I apologize for the use of "he"; I encourage
the reader to make a global change to any preferred wording.

###