Information and Libraries

Date: Tue, 17 Nov 1998 22:17:55 -0800
From: (Paul Duguid)

Paul Duguid

The academic senate at San Jose State University, California,
will vote next week on a plan to merge the city and the university
libraries.  The plan primarily concerns books, so on the surface it
is not necessarily of interest to those concerned with the digital
world.  I suspect, however, that the plausibility of that plan rests
on notions of information that have flattened distinctions between
different kinds of institution, use, and users.  But before offering
my opinion, let me give some facts.

San Jose State University [SJS], part of the California State system,
has long needed a new library.  The current collection is illogically
divided between two buildings, and each is inadequate.  Attempts to
fund a new library through a bond issue, voted on when the state was
in recession, failed.  Meanwhile, the city had acknowledged that, for
a major metropolitan center (San Jose is the nearest city to Silicon
Valley and growing fast), its public library was small and the
collection weak.  In February 1997, the mayor, Susan Hammer, proposed
fixing all these problems by combining the two facilities.  The city
would gain access to SJS's collection, and so not need to improve its
own.  The university would get a new building funded in good part by
the city.  Predictably, this was announced as a "win-win solution."

The city's plans for the libraries, however, are not as pure as they are
naive.  The current city library sits opposite the convention center.  The
city wants the site for a new hotel.  By merging the two libraries, it will
get it.  Under the current plan, the city will provide $71 million, the state
$90 million, and SJS $10 million and a new facility will be built on the
campus housing both collections.  The collections will be kept physically
separate within the building, but all users will have access and borrowing
privileges to all books.  There will be one circulation and one reference
service for all.  And though the staff of each organization will remain
separate, plans promise an undefined "seamless service".  The SJS senate will
vote on the matter on December 7th and those opposing the plan would welcome
any support you can offer.  (Fuller details can be found at

The planned merger takes it for granted that libraries are somehow
indifferent to the collection they hold or to the particular users
they serve.  Only from this perspective can a single library to serve
profoundly different communities seem unproblematic.  That perspective
may in turn reflect ways in which access to libraries has been
portrayed simply as access to information they store, access which
a digital future will provide remotely, rendering local features yet
more irrelevant.  Indeed, the California State system doesn't even
believe that this is a matter for the future.  In 1993, it decided
to build a new campus at Fort Ord without a library at all, arguing
that the information the students needed was or soon would be on line.

Where do such ideas come from -- when not from the willful ignorance
of administrators hoping to save money or mayors with Mitterand-like
pretensions hoping to cut a deal for a convention center?  One source
is the popular portrayal of the Web as home for "all the answers
you need" -- as a line on Oracle's web site reads.  Another might
be advertisements of the sort IBM aired in the States a few years
ago -- early in its series "solutions for a small planet," when the
"envisionary" (or simply deceptive) nature of the genre was not well

The ad ran in Italian with subtitles and showed an old Italian farmer
on his farm telling his astonished granddaughter that he had just
received his Ph.D.  The exchange went on:

     Grandfather:	Did my Research at Indiana University. 
     Woman:	        Indiana?
     Grandfather: 	Yup.  IBM took the school's library ...
                            and digitized it.  So I could access it
                            over the Internet.
           She cocks her ear to take this all in.
       Grandfather:	You know...  It's a great time to be alive.

[The storyboards for the ad are at]

The picture, particularly to those who know little about the
digitization of libraries, is compelling.  Unfortunately, it is also
grossly misleading.  IBM has not digitized the Indiana library.  (In
response to protest, the library confessed that "some of its music
collection" has been digitized.  IBM did not respond at all.)  But,
above all, the ad endorses the notion that libraries are little more
than information repositories -- a primitive form of file server.
(It also gives a weird view of education, but let that pass.)

Actual work on digital libraries, while more difficult than IBM's
ad acknowledges, is often very impressive.  But, in general, it
has critically failed to address libraries as social systems rather
than information systems.  Yet by using the name "library," the
work suggests that it is concerned with all that term encompasses.

I once suggested that the U.S. digital library initiative, funded
with government money, had appropriated the term "library" to help
draw popular and political support for what was in essence computer
science work.  Far from it, I was told by one proponent, the term
library was probably more of a disadvantage than an advantage to
their work.  Nevertheless, though it was considered, the National
Science Foundation has not abandoned the term.  Nor, perhaps more
importantly, has anyone involved considered that, in using the phrase
"digital library," whatever the disadvantage to them, the problems
for conventional libraries was likely to be far greater.

Not only does the phrase suggest that the future of all libraries
is simply digital.  It also leads to the flattening of distinctions
between the way libraries are used.  But does anyone really know how
libraries are used?  At the moment, I suspect, most people simply
assume that they know what goes on in libraries.  At a workshop of the
National Science Foundation's digital library project that I attended,
studies of use and users were referred to by one researcher as "touchy
feely" issues, while a previously rejected proposal by an eminent
ethnographer to study use was mentioned with a rolling of eyeballs.

(The failure of the book to die when its obituaries were confidently
read has led studies of the book and print culture to flourish.  I
hope that the same will soon be true of libraries.  Digital library
work is not sufficiently advanced here, however, to meet the sort of
rebuff that prompts reflection.  In Australia, the collapse of a major
national project that was due to go on-line at the end of 1996 has
provoked a major reassessment.)

At the moment, the idea of different patterns of use has received
little attention.  Take, for example, the matter of electronic
journals.  Justifiably, Andrew Odlyzko's paper, "Tragic Loss or
Good Riddance" has become one of the most influential papers on the
future of such journals, an important issue for all libraries.  But
that paper is directed entirely at scientific research and articles.
It does not tell (because it does not attempt to tell) much at all
about the humanities and social sciences or the ways their researchers
use libraries.  Indeed, because those most deeply involved in
digital research tend to be scientists, the way other fields work has
generally been taken for granted.  So Odzlyko's paper is read as if it
spoke for all journals and journal users.  It's not hard to think of
reasons why physicists use preprints from a file server at Los Alamos
extensively, but those studying the humanities do not.  But I don't
know of the difference being widely acknowledged.  It might be time
to revive, at least for purposes of analysis, C.P. Snow's old notion
of the "two cultures" -- but now from quite a different perspective.
Snow felt that the sciences were being slighted; now they are used to
represent all research.  And if one voice can speak for all, then one
library, presumably, can serve for all.

It's worth noting, then, that the new British Library, which has
been well received, has separated the reading and periodical rooms
for the humanities and the sciences, as if in recognition of their
distinct practices.  More relevant perhaps to San Jose, the New York
Public Library on 42d Street, which has just reopened after extensive
renovation, has maintained its historic distinction between the
research facilities (which are nonetheless open to all), and its
public lending facilities.  Though both in the same building, the two
are entirely separate.

Without conducting research (though I do use both city and university
libraries here in Berkeley), it seems quite unexceptional to me to
say that the patterns of use in a university research library and a
civic lending library are quite distinct.  The university's library's
primary obligation, for example, is to its collection; thus its
barriers are high.  The city's obligation, by contrast, is primarily
to its readers; its barriers are correspondingly low.  The university
lends only to people who pay fees to the university.  And it refuses
to graduate students who have library fines outstanding.  Guards
search bags at the exits.  The city lends to anyone with proof of a
local address; it cannot search bags.  (The library of the city of San
Jose has about $1 million dollars outstanding in fines and no current
record of how many books are lost, stolen, or unreturned.)

In numerous other ways -- patterns of readership, use of reference
facilities and reading rooms, development of collections, availability
of materials, resistance to censorship laws, and use of information
technology -- public and private libraries are different kinds
of institution.  Any suggestion that both are little more than
information warehouses, however, obscures the difference.  To those
who theorize information, who create ads for IBM (and who went along
at Indiana), and particularly to those involved in digital library
research, it will no doubt seem reckless to suggest that the merger
in San Jose has anything to do with them.  But is it too far fetched
to think that the damage that will result to both city and university
libraries as a result of this merger is in part "collateral damage"
from the indifference shown in these instances as elsewhere to the
institutional and organizational structure in which libraries and all
information sources are, inescapably, situated?