The Changing Nature of Networking in the Research Library

Community

by Henriette D. Avram '

Associate Librarianfor Collections Services

Library of Congress

Networking among libraries is certainly nothing new; to

the contrary, libraries have long been pioneers in net-

working activities. Today, I want briefly to summarize

that history, but in keeping with the theme of this

session, my focus will be on future plans and prospects.

We are in an exciting period now— one where the

technology and our collective imagination are at such a

confluence as to yield exhilarating results for libraries,

especially the research library community.

From the beginning, networking among libraries has

been propelled by our strongly held tradition of resource

sharing. The 1960s and 70s witnessed libraries turning

increasingly to the computer and the possibilities it held

for automating library operations. The development of

the MARC format at LC, the acceptance of the format by

library practitioners, the adoption of the format structure

as a national and international standard, and LC's

distribution of its cataloging data in this format, marked

the beginning of the true era of library networking as we
define it today.

There then appeared organizations that were new on the

scene, namely, bibliographic utilities. These organiza-

tions, created to serve the needs of libraries desirous of a

central source for cataloging records at a reasonable cost,

built growing files of cataloging data to which access

was limited to institutions who were members of the

particular utility. By the mid-1970s several large

databases— OCLC, the Research Libraries Group's

RLIN (Research Libraries Information Network), and

WLN (the Western Library Network), along with the one

at LC— coexisted in the United States, but could not be

accessed and shared directly. To remedy this situation,

efforts were initiated to enable libraries more easily to

share data housed on dissimilar systems and thus was
created the Linked Systems Project or LSP.

The International Organization for Standardization's

Open Systems Interconnection (OSI) Reference Model
was chosen by the LSP architects as the appropriate

protocol package to run LSP because OSI would substan-

tially reduce future development necessary to accommo-
date new systems, new applications, and new standards

being developed in accordance with the OSI model.

Two LSP applications modules were developed—

Record Transfer and Information Retrieval. Record

Transfer enables records of any type and any number to

be transported between systems. Information Retrieval

permits users of one system to access a remote system

and to view data found in the remote systems. Informa-

tion Retrieval permits users of one system to access a

remote system and to view data found in the remote

system on their own system, invoking the familiar query

command of their local system, effectively overcoming

the problem of multiple syntaxes.

The OSI-based protocols for LSP were fashioned to be of

general service and not application specific. Thus LSP
applications can be expanded to other purposes, e.g., the

Information Retrieval protocol, which is now an Ameri-

can National Standards Institute standard, is the basis of

several projects being planned in the U.S. for accessing

remote databases of all kinds, e.g., full text, abstract and

indexing, and others.

Currently, LSP is being used to support the exchange of

authority records. LC has distributed via LSP connec-

tions more than 2.5 million authority records to RLIN
and over 1.5 to OCLC. LC has received via LSP, over

88,000 authority records created by RLIN and OCLC
libraries which have been added to the file at LC and

distributed via LC's Cataloging Distribution Service to

libraries all over the world.

The next step is to support the exchange of bibliographic

records which will enable records to be searched between

systems, retrieved and then added to a particular database

for use by its patrons. By this augmentation of LSP, a

cooperative operation located at LC, NCCP (National

Coordinated Cataloging Program), will gain increased

efficiency. NCCP brings together eight research libraries

that have agreed to contribute national-level cataloging

records to the national database at LC for distribution to

the Nation's libraries.

While the library community was availing itself of

advances in technology to forge a national bibliographic

network via LSP, using OSI protocols, other networks

were evolving in the U.S. using different standards. The

academic and scientific community was busily laying the

foundation for a supemetwork supported by TCP/IP

(Transmission Control Protocol/Internet Protocol) that

Fall/Winter 1990



will support research and scientific investigations. This

network, known as the Internet, connects universities

across the U.S. and links them to supercomputer centers.

The Internet is a long-haul network that provides national

connectivity through the Unking of regional networks

which cover large geographic areas. NSFNet, the

National Science Foundation Network, acts as the

backbone of the Internet.

Because of the difference in standards used, the two

networks being built were incompatible. The first step

toward reconciling this incompatibility was to seek

cooperation with EDUCOM (a consortium of U.S.

institutions of higher learning). Accordingly, in 1987,

Henriette Avram invited Ken King, President of

EDUCOM, to come to LC for exploratory talks.

These initial meetings opened our eyes as librarians to

the enormity of what was happening, what was being

planned on the academic side, and also what was miss-

ing, i.e., much of what libraries were already doing or

had already accomplished. The aim was to connect

scholars' workstations on the Nation's campuses to each

other as well as to supercomputer centers via the Internet

to support research needs. Besides NSF, the players in

this grand scheme were influential and represented big

money interests— IBM, AT&T, and New York Tele-

phone.

As more was learned of what was envisioned, an addi-

tional strong concern emerged— the research libraries

on university campuses being wired to each other and to

supercomputer centers were also part of the LSP environ-

ment. These libraries are the keepers and organizers of

much of the information and data that feed the research

process. It is the ability of libraries to organize informa-

tion for retrieval and end user access that makes sharing

data over networks viable. We should not lose sight of

the importance of the organization of information and the

critical role standards play in this process of organiza-

tion. Indeed, it is this technical processing aspect that

underpins and makes possible the research and reference

functions that will become increasingly important as the

supernetwork takes shape and the variety of data on it

mushrooms.

EDUCOM has been instrumental in pushing and tracking

legislation currently before Congress (S. 1067, sponsored

by Senator Albert Gore). If approved, the evolution of
the Internet will take on immense proportions. Of keen
interest to research libraries is Title II of the bill, which
calls for the creation of a high capacity National Re-
search and Education Network (NREN), which will

interconnect over 1,000 colleges, universities, research

organizations, and, we hope, their libraries. Title II

further specifies that— working with other agencies—

by 1996, NSF would establish a multi-gigabit NREN
capable of transmitting 100,000 typed pages or 1 ,000

satellite photographs in one second. The network is to be

phased out when national, commercial high-speed

networks can satisfy research needs.

Now that the networking infrastructure that will serve the

Nation in its various components has been described, it is

well worth spending a few minutes discussing the

"content" of the Network, i.e., what kinds of information

and data will be accessible over the Network. As I've

tried to make clear, in terms of describing and formatting

bibliographic data for efficient searching and retrieval,

we're pretty much there. The standards are well defined

and broadly applied within the library networking

environment. For non-bibliographic data, however, we
have some way to go yet But strides are being made.

Having spent its first fourteen years focused primarily on

networking of bibliographic data, the Library of Con-

gress Network Advisory Committee (NAC, an umbrella

group comprised of library and networking professionals)

recently shifted its attention to non-bibliographic data-

bases (which it defines to include full-text, numeric, and

graphic data). NAC devoted an enure program meeting

to this topic last year. Entitled "Beyond Bibliographic

Data," its goals were to gain a better understanding of the

term "non-bibliographic" in the library network context

and to begin to appreciate the range and potential of such

electronic information. Among other things, by the end

of the meeting, it was agreed that: librarians must be able

to cope with the multiplicity of forms of information; a

user interface that will enable scholars and the public to

access and display bibliographic and non-bibliographic

data files must be devised; standardization and informa-

tion selection issues remain outstanding; and libraries

with local systems and how they affect the relationship of

libraries to bibliographic utilities is emerging as a

problem: At risk is resource sharing as librarians have

known it as attempts are made for economic reasons to

seek lower cost alternatives to cataloging on the utilities

and thereby not adding expensive cataloging records to a

large national database for sharing. The complete

proceedings of the meeting are available as part of the

Library Congress Network Planning Papers series.

EDUCOM has recently accepted a project proposal, "The

Library and the Electronic Document Environment

Infrastructure," submitted by Mrs. Avram on behalf of

the Library of Congress that calls for a full-scale effort

coordinated by EDUCOM, in conjunction with the

various stakeholder communities (libraries, researchers,

information processors, publishers, professional organi-

zation, etc.). The proposal concentrates on three areas of

activity:

IASSIST Quarterly



1) to carry out major studies in three core areas of the

electronic document environment:

a) technology and formats

What information resources are needed and in what
electronic format?

How will the information be organized for retrieval,

transmission, and exchange?
Will the technologies provide solutions to problems

of storage, preservation, and presentation of

information?

b) economic issues and choices

Who pays for and owns information resources in the

electronic document environment?

c) roles and responsibilities for libraries

How will the library exercise its role as organizer,

classifier, and preserver of information and

knowledge in die national network?

2) to devise several structural models conducive to the

growth of electronic document environment; and

3) to offer programs, seminars, and publications which
present findings about the library in the age of

electronic research, production, and publishing.

What has emerged from this discussion is the notion that

through the application of technology to networking,

libraries will become boundless and that users, by

accessing networks, will become patrons of "libraries

without walls." Right here in this state, a plan has been

issued by the New York State Library which details how
all libraries in the state— academic, school, public, or

other— can become electronic doorways for citizens of

New York. An electronic doorway library would make
needed information available electronically to users from

any part of the state via links to databases and resource

sharing programs with computers.

Research libraries are moving ahead on several fronts

through various organizations, both singly and collabora-

tively, in dealing with non-bibliographic data in a

network environment. One of the most promising

involves ARL, CAUSE (the association for the manage-

ment of information technology in higher education), and

EDUCOM forming the Coalition for Networked Infor-

mation. The Coalition will consist of a large and influen-

tial group of institutions of higher education, not-for-

profit organizations, corporate sponsors, and government

agencies. It has set for itself an agenda that includes

crafting a set of initiatives to deal with the provision of

information resources on the National Research and

Education Network. The Coalition will focus on issues

related to intellectual property rights, standards, licens-

ing, service arrangements, cost recovery fees, and

economic models. So far, as of May 4, over sixty

research libraries in the U.S. and Canada, including the

Library of Congress, have committed to joining the

Coalition.

As we move into the 1990s, it is fair to say that the

implications for research libraries of networking and the

changing network infrastructure are immense. But, as

can be seen, this final decade of the century holds great

promise to be an exciting and innovative one as well.

And while the task before research libraries in servicing

non-bibliographic data in the network setting is stagger-

ing, some excellent first steps are being taken.

1 Presented at the IASSIST 90 Conference held in

Poughkeepsie, N.Y. May 30 - June 2, 1990.

NARA Job Announcement - Center for

Electronic Records

Job Announcement - March 1, 1991

Within the next two weeks or so, the National

Archives will announce one or more vacancies for

senior archivists to deal with computer records.

Starting salary is $37,294 with all fringe benefits of

U.S. Government employment. U.S.citizenship

required. If any one is interested, or knows anyone

who may be interested, please contact the following:

Thomas E. Brown

Chief, Archival Services Branch

Center for Electronic Records

National Archives and Records Administration

Washington, D.C. 20408

(202)501-5565

TBROWN@DCUNSN.DAS.NET

Fall/Winter 1990