.zE$..».;!:. 1.3:... .

...-1. h

"3- A J 11 urn.
... e .5”. m4? ....

....»
;

3.1!. .
z 1:
. xii... .
. l.‘ I .I
31 ...... 1‘11
...}. ::.5.:.1
1 pl...

Elli.
.un‘hflhu.ilu
f!.c(J"3\

5.1.“.
.... .....i. ...
,1 . Ankara ..uaci. .
......M”. . 15.9.1.9...“11. 2m”... ..
.. .. . 2
. .. .. 1
5311...? "flaw
ravine”?! :
$1.3: unhuxuukr.
(.9116... .
«5F.- ..s,....r,..... z...

5......
....21121...
...-1.3... .1:
.511?! .23.
1 ..
{.11}... 11.. : ..
.13. grizinrl.
61.3.3-1...1531”. 3...“: .3
.... .1 .. 2...: ....
1......4»... 2 . ...!
1 .1 1. q .‘6 11. I.
1

"kn. xx.

.- v...» I ..l. .
12.55.. «.2. ...»

13m}... in...“

s :er ......
. la}...
I.

’

. t

.\
I~i~1

 

 

m IllllllllllllllIllillllllllllllllll

) 1 3 1293 01410 7407

 

 

 

 

 

 

K‘

(If/‘75:)

This is to certify that the

thesis entitled

Networks, People, and Place:
A Model for Networking Communities

presented by
Gregory Michael Laudeman ‘

has been accepted towards fulfillment ,
of the requirements for ‘

Master of Arts _ Telecommunication
degree in

Wfiflcagf
Dr. David McCarty

Major professor

 

 

 

Date November 13, 1995

 

0.7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE ll RETURN BOX to roman this checkout from your rocord.
‘TO AVOID FINES return on or baton data duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I I- l- I

 

MSU IoAn Afflrmatlvo MONEquIl Oppommlty lmtltdton

 

NETWORKS, PEOPLE, AND PLACE:
A MODEL FOR NETWORKING COMMUNITIES

By

Gregory Michael Laudeman

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Telecommunication

1995

ABSTRACT

NETWORKS, PEOPLE, AND PLACE
A MODEL FOR NETWORKING COMMUNITIES

By

Gregory Michael Laudeman

Communities and enterprises are both complex social systems. Enterprises use communication
technologies in the form of computer local- and wide-area networks to support their operations. Networks
optimize accessibility, distribute processing, and facilitate sharing and coordination. They help
decentralized, flat organizations to identify opportunities, make decisions, resolve conflicts, and solve
problems. Enterprise networks provide a catalyst for innovating organizational objectives and processes,

as well as a means to improve operations.

Communities have information flows and needs that are similar to those of enterprises, though they
tend to be much less hierarchical, centralized and structured than traditional organizations. They are
social networks that function to integrate their members, to help them grow and prosper through processes
of recruitment, socialization and social control. Community information systems should parallel the

structure of communities and support their processes.

Contemporary technologies such as groupware, hypertext, remote connectivity, and client/server
architecture could be effectively adapted to facilitate community processes. Networks require a critical
mass of services and usage in order to be economical and effective. The practicality of adapting enterprise
networking to a community is dependent on willingness of it constituents to adopt technological means of

improving their situations.

ACKNOWLEDGEMENTS

I would like to express my appreciation to the Department of Telecommunication at Michigan State
University, and the University itself, for the providing an excellent academic environment. In particular,
I’d like to thank Dr. Tom Muth, Dr. Johannes Bauer, and especially Dr. David McCarty for their
indulgence and assistance. Also, I’d like to thank my friends and family for their support during my time
in graduate school, and while I was researching and writing this thesis. May everyone have a communities

as wonderful as mine.

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS .................................................................................................................. iii
TABLE OF CONTENTS ..................................................................................................................... iv
LIST OF TABLES ............................................................................................................................... v
LIST OF FIGURES .............................................................................................................................. vi
INTRODUCTION ................................................................................................................................ 1
METHODOLOGY ............................................................................................................................... 2
COMMUNITY .................................................................................................................................... 4
Community and its environment ............................................................................................. 5
Community processes ............................................................................................................. 7
The Components of Community ............................................................................................. 9
Community inputs and outputs ............................................................................................... 10
Community function, form and structure ................................................................................ 11
ENTERPRISES ................................................................................................................................... 15
Communication technologies in organizations ........................................................................ 24
ENTERPRISE NETWORKS ................................................................................................................ 34
The Function of Networks ...................................................................................................... 34
Network forms ....................................................................................................................... 40
Type of application ................................................................................................... 42
Specific applications ................................................................................................. 42
Network structure ................................................................................................................... 45
THE NETWORK MODEL .................................................................................................................. 51
Application of the model ........................................................................................................ 63
NETWORKING COMMUNITIES ....................................................................................................... 69
Communities of interest and networks for social action .......................................................... 69
Community Networks ............................................................................................................. 72
CONCLUSIONS, OPPORTUNITIES, AND RECOMMENDATIONS ................................................. 77
BIBLIOGRAPHY ................................................................................................................................ 83

iv

LIST OF TABLES

Table 1. Examples of general types and specific computer applications. ............................................... 42

LIST OF FIGURES

Figure 1. Browning’s “Virtuous Spiral” ............................................................................................... 30

Figure 2. A general comparison of network protocol suite reference models (adapted from Robert
B. Walford 114) ................................................................................................................................... 47

INTRODUCTION

Humans are social creatures. We work well in groups and rely on interpersonal bonds of intimacy
and commitment to meet shared needs: food, shelter, belonging, self-esteem, education, etc. It is this
shared dependency that gives rise to community, the phenomenon of people living and working together
within a geographic area. Communities are complex, multidimensional “meta-organizations” that
encompass the range of human endeavors. They are the context in which we judge our welfare, our quality
of life and cost of living. It has always been fashionable to decry the demise of community and to cite the
resulting social ills (Claude Fischer 190-192), but regardless of the state of a particular community, few of

its members would pass up an opportunity to improve their quality of life or lower their cost of living.

In the context of commercial and institutional enterprise, quality of life and cost of living translate
into income and expense, and are called “the bottom line”: what you gain and what you lose. Over the last
two decades, enterprises have developed a number of technological models for improving their bottom
lines. The primary components of these models are task-specific sofiware, computers, and links between
the computers. One such model, the enterprise network, creates a particular form of information system
that is well-suited to highly integrated yet diverse and decentralized organizations. Networked
organizations are able to respond rapidly to changes in the environment, to efi'ectively identify needs and

to fulfill them cost-effectively.

Can the enterprise network model be applied outside the focused context of commercial and
institutional enterprise, into the larger context of the community? It has been recognized by progressive
thinkers that there is a technological—communication technologies in particular—component in the
improvement of communities (Park Dixon Goist 42; Making Cities Work 12). Could an information
system enhance the functionality of a community similar to the way in which it would contribute to an
organization that is more goal-oriented? Nearly forty communities across the United States and Canada
have active community network projects. How do community networks compare and contrast with
enterprise networks? What can we learn through a critical analysis of community networks using the
enterprise networks as a model? This thesis examines these questions, seeks to identify opportunities for

l

2
community networking, and offers practical recommendations on the creation of community networks that

will have positive impacts on communities’ “bottom lines.”

METHODOLOGY

This thesis is a synthesis of research in a number of fields, including urban design, organizational
communication, sociology and management science. In order to understand how a computer network
might benefit a community, we must understand what a community is, how it works, and how it can be
strengthened. There has been a vast amount of work done in sociology, anthropology, urban design,
political science, and other disciplines on the subject of community. The most prominent and persuasive

of these provide an excellent understanding of the structure, form, and function of communities.

If anything has been examined as thoroughly as community, it is information technology (IT). The
bows and whys of implementing IT are exhaustively debated in the popular press, professional
publications, and the world-wide Internet. To build a model of networking, we must identify the features
of certain information technologies, the characteristics of organizations that apply IT suwessfully, and
how the two are related. A review of the literature of organizational communication and of current
information technology applications will provide the parts of the network model, as well as an

understanding of the dynamics of the model, and its impact on organizations.

In comparison to the more general phenomena of community and information technology,
community networks have evolved almost completely unnoticed. They have come about in a number of
different ways, using a variety of technological models, and with varying levels of success. By visiting
community networks through the Internet, examining surveys that have been conducted on the Internet,
and reviewing the literature on community networking we can perform an analysis of the state of

community networking today.

The conclusion of this thesis is a synthesis of the whys and hows of communities, the functions of
information technology in organizations, and existing efforts towards networking communities. The
results of this synthesis are a broader and deeper understanding of what community networking is and

how communities might benefit from it, a model for economical and equitable community networking,

3
identification of opportunities in community networking, and recommendations on the design,

implementation and operation of a community network.

COMMUNITY

As a complex, multidimensional phenomenon, community is difficult to precisely define and is
generally informal, i.e., changes form across space and time, and may have indefinite characteristics. A
definition is necessary in order to discuss community, but it is necessarily a vastly simplified, generalized,
and abstract definition. For our purpose, community is a social system in which persons with overlapping
and conflicting agendas interact with the environment and each other within a geographically definable
area in order to achieve shared and personal goals. Our challenge is to fully develop this definition in
order to understand the costs and benefits of community networking. To answer this challenge we will
move along a course from generalizations and abstractions to bottom-line, real world implications and
recommendations. To help us stay on that course, we will apply concepts to a specific community:
Chattanooga, Tennessee. Our discussion of this particular community will provide a subjective foil for the

more objective, academic discussion of community as a phenomenon.

In its broadest sense community can be understood as a social system (Terry Clark; Irwin Sanders;
Fischer; Making Cities Wo¢, et. al.). A system is a group of interrelated, interacting, or interdependent
constituents forming a complex whole, a fimctionally related group of elements (Webster’s II New
Riverside University Dictionary). Stephen Littlejohn provides an excellent overview of the development of
general systems theory, and related studies of cybemetics and information theory. He notes that open
systems have several characteristics:

- Wholeness and interdependence

- Hierarchy

- Self-regulation and control

- Interchange with the environment
- Balance

- Change and adaptability

- Equifinality (35-38)

 

The general elements of a system include:
- an environment
- processes
- components
- inputs from the environment

0 outputs to the environment.

The general function of a system is to process inputs into outputs. A system achieves and maintains
the characteristics described above through the persistence of interactions between components, acting
upon inputs. These actions and interactions make up the system’s processes and result in the
transformation of the inputs into outputs. The system persists because its outputs have a greater value than
its inputs. The value of inputs and outputs are determined by the transactions involving those items that
take place in the environment. Through a number of transactions over time the environment sets value
and thereby affects the outputs and inputs of a system. A system will tend to produce what it can with the
greatest positive difference between the value of the inputs and the value of the outputs. In this way the

function of a system is determined by its environment.

Community and its environment

The environment in which a system operates is the primary determinant of its function, form and
structure. Richard Farace, Peter Monge and Hamish Russell delineate general types of environments along
a continuum from “placid-randomized,” in which a person [system] can wander without direction,
learning to achieve goals and avoid danger by trial and error, through “placid-clustered” and “distributive-
reactive” to “turbulent,” in which a system must be ready to change its basic nature in order to adapt (36-
37). Arensberg and Kimball see community as the link between society, culture, and the environment (2).
Sanders turns this linkage around, saying that community “is part of and acted upon by complex
environmental factors” (44). This “setting” includes a community’s:

- Ecology

- Demography

° Culture
. Personality
- Time

° Society at large (Sanders 44-47)

Rather than delving into inconsequential details, let’s restrict the description of our referent city to a brief
overview of Sander’s components:

Contemporary Chattanooga exists, of course, in the United States of the nineteen nineties. It is part
of, and exists within, that particular larger social setting. More specifically, Chattanooga exists in a
society that sees itself as fragmented and out of touch with tradition, dealing with constant change
and uncertainty, yet dedieated to the ideals of “rugged individualism” and pursuit of the “American
dream.” Overall, it exists in what would be characterized as a placid-clustered environment.

Chattanooga is a city and surrounding metropolitan area in southeastern Tennessee, where the state
borders Georgia and Alabama, just east of the central time zone, just west of the Smoky Mountains,
and within a two hour drive of four major metropolitan areas. It sits up against the eastern flank of
the Cumberland Plateau and straddles the Tennessee River. The center of the city is its Downtown
area, which has recently undergone significant revitalization. Downtown is surrounded by industry,
especially to the south, and lower income neighborhoods.

There are several smaller centers in non-annexed towns and shopping areas to the north and east of
the city. The smaller centers are surrounded by suburban middle-class neighborhoods. Upper-class
neighborhoods are located on the ridges and plateaus that rise up around the center of the city.

The weather is temperate, the soil is fertile, and there are abundant natural resources. The
community is surrounded by rural areas, which support farms, extraction operations, and semi-
wildemess, and numerous smaller cities and towns. Chattanooga is a center for ground and water
transportation in the southeast, and supports diverse tourism and light manufacturing. Much of
Chattanooga’s heavy industry faded in the nineteen seventies, although there has been a steady—if
slow—immigration of heavy manufacturing over the last decade. It has made a recovery from being
one of the country’s most polluted cities, to being trumpeted as the “Scenic City.”

The people of Chattanooga are generally conservative, predominantly white, middle- and lower-
class, with strong upper-class and minority components. They tend to be religious, with strong
beliefs about personal freedom and moral commitment that are popularly characterized as “family
values.” Regardless of the social conservatism professed by much of the population, Chattanooga has
a vibrant cultural life, even a hint of an “underground” culture. The city has numerous museums—
including an aquarium, train museum, and several art museums—and is rarely without a cultural
event occurring near the city center.

Chattanooga has had—and still has—its share of social problems. And, like many communities, it
has a tendency to try to ignore them. The community does benefit from a healthy grass-roots
tradition, and the capital to fund such efforts. In particular, the community created an opportunity
for self-appraisal and goal-setting—referred to as “visioning”-—that has spawned significant
amounts of civic activity and investment.

7
Beyond this general description, the community fits Sander’s characterization in that it is a place

with a population that exhibits general cultural and particular personal traits, and fills a range of social
roles. It is important to note that the community and its environment are intermingled. While the
community of Chattanooga, as a system, has boundaries that might be delineated in a number of ways-—
geographically, socially, economically, pyschographieally and demographically—those boundaries tend to
be vague and dynamic. As a small example, if an individual lives in Chattanooga but works in Dalton
thirty minutes away, does that extend the boundaries of the community or compromise them? The
elements of Sanders’ setting are not singular or static measures; there may be only a vague and shifting
delineation between where Chattanooga’s cultural milieu ends and Dalton’s begins. Not only is a
community, or any system, affected by external factors in the environment, it has an efl‘ect upon itself. For
instance, Chattanooga’s culture and ecology are both components of community, but they are also parts of

its environment. Culture transforms ecology. Ecology impacts culture.

Community processes

Systems encompass a range of general processes: procurement, allocation, production, consumption,
transformation, innovation and change, maintenance, communication, etc. As a particular type of system,
a community has specific processes, which Sanders identifies as recruitment, socialization, and social
control. He states that, in order for a community to prosper it must:

1. Recruit new members either through birth, in-migration, or annexation, and maintain existing
members;

2. Train the new members to play the appropriate roles as they take their places and achieve status in
the community;

3. Exert some form of control over individuals who deviate too far from the norm (192).

These specific processes are carried forward, Sanders suggests, by two particular general processes:

allocation and communication (219-240). The allocation of resources, roles, power and prestige to

members of the community provides incentives and disincentives for becoming, staying, or bringing in a

8
member of the community. Similarly, allocation provides rewards or punishments for filling certain roles

or behaving in certain ways.

Communieation is the process by which members or potential members of a community come to
know how the community performs allocation—how much of what resources are assigned to which
components of the community—as well as how efficiently, effectively and equitably the community
performs its more specific processes of recruitment, socialization and control. Communication is an
especially important process because of the function it provides for individuals. Frank Dance identifies the
functions of communication as:

- Integration of an entity and its milieu;

0 Development of mental functions;

- Control of behavior. (285)

The general process of communication allows the community to carry out its specific processes—
recruitment, socialization, and control—by bringing individuals within its boundaries, enhancing their

ability to deal with diversity and to prosper.

Beyond allocation and communication, as identified by Sanders, another general process that is
significant to this thesis is innovation, as examined Making Cities Work. The editors of this book identify
innovation as “not just a response to pressure but structural or system-wide effects on cities” (1 1).
Innovation is not invention, but the process by which new ideas or technologies diffuse into use within a
community (1 1-12). It is a process of restructuring the social perspective in order to create long-term,
sustainable improvements in the community. More particularly, innovation is the process by which a

community not only responds to changes in its environment, but exploits them. Schon 56-65).

In Chattanooga, the specific processes of community are realized in numerous ways:

0 An economic development board recruits businesses, encouraging them to move to the Chattanooga
area, possibly even arranging tax breaks for them. [allocation]

- The Big Brothers/Big Sisters program provides positive role models and friends for children from

disadvantaged or broken homes.

9
- The police and sheriff’s department regularly set up sobriety check points along major roads on the

weekends, reminding people to stay sober when driving and inearcerating those who don’t pass the
test.
- Through its Visioning process, the community is attempting to proactively restructure the ways in

which it carries out recruitment, socialization, and social control.

The Components of Community

Arensberg and Kimball assert that the community is actually a “series of independent systems”,
suggesting that community is a master-model of other social systems linking, as already noted, society,
culture and environment. They go on to build a taxonomy of these systems based on their categorization as
relational, customary, or value systems. Relational systems are the primary components of a community.
They are enumerated as:

- Family

- Economic

- Religious

- Political

- Edueational

- Associational
- Informal

Each of these relational systems has customary and value systems associated with it, and is made up of
numerous components itself. When one of the relational systems changes, the change affects its

components, its associated system of customs and values, and the other relational systems (269).

The major components of relational systems can be categorized as actors, such as persons,
families/households, institutions, enterprises, associations, and resources, such as eapital, real estate,
public goods, private goods, and time. These actors may be part of and resources available to several

relational systems, fulfilling a similar number of roles. Logically, any particular actor has limited

10
resources with which to act. This fact is reflected in problems of allocation, and particularly in the

problem of role strain (Sanders 108-11). Actors allocate their own resources within a relational system
basedontheassociatedsystemofvalues, andusetheassociatedsystemofcustomsasaguidefortheir
actions. As they move between relational systems—and roles—they must adjust how they allocate their
resources and conduct themselves. The process of this adjustment, and the process of communication
required for knowing when and how to change, exert a strain on the actors resources, over and above what
has been allocated to each role. Role strain accrues from the actor having limited resources for allocation

to a practically unlimited number of roles, and from having to change roles.

Relational systems and their associated values and customs are quite apparent in Chattanooga. An
individual may have a role in an economic system as an employee, in a family system as a child, in an
educational system as a student, in an associational system as a darts enthusiast, etc. There are complex
patterns of interaction in the form of general processes such as allocation, communication, transportation,
etc., taking place among the people of Chattanooga as they fill their various roles. They give up their time,
energy, and knowledge in return for money, and give up their money for education, entertainment, and
domestic tranquility. Participation in these processes fulfills basic social, institutional, and personal needs
for the actors. This concept applies to institutions such as mental hospitals and schools, and enterprises

such as barbecue stands and banks, as well as for individuals and families.

Community inputs and outputs

Each of Chattanooga’s relational systems—and the systems of which they’re composed—have
identifiable inputs and outputs. There are a set of items that each one takes in, transforms, and then
returns to its environs. The foundry takes in ore and power, and produces pipes. The City Commission
takes in facts, figures, public opinion and taxes, and produces budgets and policies. Churches bring in
sinners and send out saints. The family takes in food and energy, and outputs children who are well-
adjusted, intelligent, and healthy. But what about the community itself, the meta-system of social systems,
what are its inputs and outputs? Well, if we consider its processes—recruitment, socialization, and social

control—we can see that those processes exist to benefit the constituents of the community, which is to say

1 1
its components: persons, families, enterprises, and institutions. What is it that a community provides to

these constituents? It provides roles in relational systems that give the actor value over and above the
value of the resources the actor gives up in order to participate in the community, its relational systems
and processes. In other words, a community synthesizes the outputs of all of its components in order to

produce a standard of living.

Iftheoutputofacommunityisanabstractstandard, thenitsinputcanbecharacterizedjustas
abstractly as a “cost,” more specifically the cost of living. This cost is paid by each constituent of the
community, each actor that benefits from acting within the context of the community. The community
maintains itself, grows or declines, based on the ratio of quality of life to cost of living. The higher the
quality of life and the lower the cost of living, the more benefit (or excess profits) for constituents of the

community, and the more wealthy the constituency becomes.

To bring this back to our referent community, we ean see that the interaction of relational systems
results in the creation of such things as industrial parks, police forces, and the renovation of historic sites.
These items and activities provide the benefits of jobs, safety, and culture to individuals and families,
which improves their quality of life. A couple may take advantage of these benefits to start a family, bring
their parents to live with them, and/or encourage friends to move to Chattanooga. In essence, by these
actions the couple “outputs” new members of their community. If Chattanooga as a community is able to
output a sufiiciently high quality of life while requiring sufficiently low inputs to the cost of living it will

attract industrial executives, farm kids, unfirlfilled professionals, and new babies.

Community function. form and structure

We now begin to see the complex form of the community emerging from the indefinite haze of
society. As we simultaneously open up a broad abstraction and zoom in on a focused application of

community, the function, form and structure of our social meta-system becomes more fully articulated.

What is the function of community? Clark suggests that communities provide these functions for
their component actors:

- Adaption, which is accomplished through the economy;

12
- Goal-attainment, which is supported by the polity;

- Integration, which is performed by law, political parties and social groups;
- Latent pattern-maintenance and tension management, which he associates with the family, education
and cultural organizations. (47)
He maintains that each firnction may be fulfilled difi'erently by any one community and that the fimctions
may be fulfilled by various communities for a given actor. The characteristics of a community are
determined by patterns of functionality that emerge across its constituents. One way of analyzing the
functionality of a community, he suggests, is by the ratio of occupations in the various professions
associated with the services through which each function is fulfilled (53). This analysis points out that the
community functions to empower its constituents. It empowers them to fulfill their physical, social, and
personal needs, sheltering its members from dangers in the environment and drawing in resources for
them. As noted above, persons and organizations may have to deal with environments ranging from placid
to turbulent. Community makes its constituents prosperous. It provides a context in which they can find

food and shelter, gain acceptance, and build self-esteem.

How does community fulfill these functions? The formal characteristics of a system are what
determines how it interacts with its environments. So what is the form of community? Well, let's step back
and look at the components and subsystems of community as they take up their roles in the process of

community and see what form emerges.

Within each relational system, general and specific processes tie together components and
subsystem. A high school student eats a hamburger (consumption), the fry cook gets paid (allocation), the
manager does payroll and inventory (maintenance), the assistant manager calls in an order for more
lettuce and tomato (communication), the supply truck pulls away from the warehouse and out onto the
road (transportation), all within a microcosm of the community. Each interaction branches out from the
last. Each individual moves from relation to relation, role to role, process to process. Each resource is

transformed, the new resource is redistributed, and is in turn transformed again. What emerges is a web of

13

activities and actors within space and time. Ifwe put an individual or organization at the center of this

web, it becomes apparent that we all exist within a “social network.” Fischer (1977), points out that,
“[s]ocial networks are important in all our lives, often finding jobs, more ofien for finding a helping
hand, companionship, or a shoulder to cry on. Our networks do more than just support us; they also
place demands upon us—for assistance, advice, to help a friend of a fiiend—and they influence our
values, attitudes, and decisions.” (19)

By moving within and through our social network we find sustenance and meaning, and through the

network we make our contribution towards the cost of living.

So, a community takes the general form of a network, but a social network is a rather vague
abstraction. If we define form as the sum of the characteristics that determine the nature of interactions,
we see that these networks take the form of relational systems and, more specifically, the form of specific
relational subsystems such as businesses, churches, schools, governments, and households. The specific
form of a relational subsystem determines the processes it may take part in, the roles it can play. Those
processes, roles, and the subsystems themselves define the structure of the larger relational system, c. g.
economics, politics, religion, etc. These relational systems, the concomitant custom and value system, and
their interactions define the form of the community, much as Clark suggests, above, and the form of the
community determines its functionality. The numbers of individuals fulfilling roles within the various
relational systems is one means of mapping the form of a specific community, of estimating to what extent

it provides its constituents with a high quality of life for a reasonable cost.

In contrast to form, which describes the external, subjective characteristics, structure describes the
internal, objective characteristics that support form. Not to detract from the work of the distinguished
scientists who study social networks, but they confirse the terms form and structure. Form defines how a
system interacts with its environment, while structure defines how the components of a system relate to
each other. Beyond this semantic point, social network analysis provides excellent definitions of the
structural variables of social systems:

Range—number of actors connected in a network

Density—the extent of interlinkage among the actors, usually expressed as the ratio of the number of
existing links to the number of possible links.

l4
Reachability—the average number of links needed to connect any two actors by the shortest route.

Clustering—the extent to which the total network is divided into distinguishable cliques. (Fischer
36)

The general attributes of relationships between actors/components are listed as:

Multiplexity—the number of relations in a given link [relationship]
Symmetry—the balance of power or profit
Intensity—degree of commitment to the link [relationship] (3 6)

Analyzed within the general framework of structure, form and function, we see that a community
will tend to alter its components through processes of recruitment, socialization, and control in order to

achieve a form that most optimally helps its constituents fulfill their nwds, avoid danger, achieve their

goals, grow and prosper.

ENTERPRISES

The analysis of social networks, of actors, roles and relations was developed as much in management
science as in social science. Everett M. Rogers and Rekha Agarwala-Rogers stress the interrelatedness of
organizational structure and communication They begin—like most who write on this topic—with the
accepted model of communication as a linear, dyadic phenomenon, but they develop this model into
complex communication networks, with individuals or groups as nodes and communications as links. As
with others in the systems school of organization, they place a great deal of emphasis on the significance
of the environment as an element of organizations and as a determinant of organizational function, form
and structure. Communication networks not only operate within the organization but also link the
organization with other entities and its environment, which Rogers and Rogers see as crucial to the ability
of organizations to adapt and innovate (70). Farace, Monge and Russell add to and elaborate on the link
attribute identified above:

- Symmetry is the balance of control and effort that the communicating entities contribute to the link.
Symmetrical communication has equitable flows between all participants; each node does equal duty
as source and sink. Asymmetrical communication flows in only one direction; the nodes have fixed
functionality as source or sink.

0 Strength is determined by the periodic qualities of the link. Stronger links are used more frequently
and/or for greater lengths of time.

- Reciprocity is the amount of awareness or agreement between participants as to the existence or

characteristics of the link.
- Content is, of course, the messages that flow across the link.
- Mode is the means of communication. (181-2)

There are four variables that may be used to formally define communication networks:
- Flexibility of the network determines “who may communicate with whom” and “what specific

pathways messages are to take.”

15

l6
- Directionality of information flows describes how communication spans the organization

horizontally and vertically, as well as whether the flows are symmetrical or asymmetrical.
- Initiation or sequencing of messages may be imposed from the top of the organization regardless of
the actions on lower levels, or sought by members of the organization, with communication in

general encouraged throughout the organization.

- Load of information processing is characterized by the amount of resources being applied to

communication. (76)

Communication is universally seen as a vital general process for organizations (Rogers and Rogers,
F arace, Monge and Russell; Peter Manning; Lee Thayer). It has even been suggested that organizations
are nothing more than complex information processing systems (Pamela Shockley-Zalabak 7; Rogers and
Rogers 57; Jay R Galbraith 179-89). For organizations as systems, communication functions to make
them more efi‘ective in garnering resources and creating products; to increase their efliciency through
innovation; and to maintain or change their form and structure in response to internal and external
strains. These functions are facilitated by communication systems, which are subsystems of the
organization. Overall, communication allows organizations to achieve their goals by supporting processes
of production, planning, coordination, control, etc. Communication provides the information, commands
and instruction, influence and persuasion, and integration (Thayer 187-249) that makes organizations

cohesive wholes.

Farace, Monge and Russell parallel the functions of organizational communication with the general
processes in organizations—production, innovation and maintenance:

1. Production. Communication involving the work being done, the work waiting to be done,
problems in the work (such as errors), and problem detection and correction.

2. Innovation. Communication involving the elicitation of new ideas for doing work, new ways to
do work, new types of work that might be done, and new procedures and policies that would
improve the operation of the organization or the overall environment.

3. Maintenance. Communication that improves or enhances the individual’s concept of self, or
the nature and quality of interpersonal relations, or the identification with and loyalty to the
company; in addition, maintenance includes communication that supports the production and
innovation functions as well. (76-7)

—and relate them to three forms of communication:

17

Environmental communication provides descriptions of actions, events or processes that observers
detect in the world in which they live. . ..

Motivational communication refers to statements that assert specific goal or values. . ..

Instructional communication consists of assertions that provide courses of action which link the
motivational and environmental statements together. (1977, 30-1)

Thayer states that the likelihood of an entity engaging in communicative behavior—either being
communicated with or communicating to—are dependent on the anticipated result of that behavior. In the
goal- and task-oriented environs of the organization, motivation to be communicative is founded on the
need to solve problems or make decisions, to achieve some productive result (78). On an operational level,
Thayer points out four preconditions to communicative behavior:

- A Relationship - Communication between entities requires that they be able to relate to one another

in some way, on some level.

- A Nwd - At least one of the entities to be involved in the communication must require information

or have some purpose for communicating.

- Roles and Rules - Rules involve the time and location of communication, as well as the medium,
language, etc. to be utilized. Entities “play” roles when they make use of certain rules in order to
communicate.

- A Language - the entities must share a means of representing meaning. (77-81)

The dynamics of communicative behavior are defined by the translation of informational needs into some

form of symbiosis through which power, control and regulation are exerted and exchanged between the

communicating parties, which Thayer terms “gaming” (82-9). For communication to be effective in
organizations, its dynamics must be managed. The questions which must be answered in the process of
managing communication are:

a) who acquires or transmits what “information,”

b) fiom or to whom,

c) when, and

d) with what consequences? (102-3; emphasis in the original)

18
Farace, Monge and Russell see this as a problem of control and coordination. Any organization, they

argue, can have only a limited amount of information (bounded rationality). This information exists in

two forms:

0 Absolute information is a single “piece of knowledge” which can be possessed by a member of the
organization. This is a measure of the amount of information that exists in the organization.

- Distributed information is the dispersion of a piece of knowledge through the organization. This is a
measure of the penetration of the information that exists in the organization into its social structure.
(26)

Absolute information either exists in the environment or in the organization’s memory, and must be

located, either actively or passively, through a search. Organizational memory consists of short-term and

long-term storage, in the form of documents or knowledgeable members of the organization. The cost of
storage and the cost of access are inversely related: while short-term memory can be easily searched, it is
more expensive. The environment “stores” information without cost, yet it may be difficult to locate that
information when it is needed. There is an economic trade ofl‘ between using information, retrieving it
from storage and reprocessing it from the environment. If the revenue potential of information is less than
the cost of storing and retrieving it—particularly if the information is of use only in the near term—it
should be discarded. Farace, Monge and Russell outline the general strategy for garnering information
from the environment and incorporating it into organizational memory:

1. “[T]he total information available to an organization first goes through a processing step as it
reaches the organization’s members by any of several different channels” Channels may be
systematically designed and used by the organization.

2. Information processed into the organization must be stored for future use. There must be rules for
deciding what is usable and should be kept, and what should be discarded immediately, as well as
methods for establishing and periodically evaluating those rules.

3. Information that is highly perishable, loses value quickly or is needed to resolve issues in ongoing

activities should be put into short-term storage where it can be retrieved with little effort.

4. Less immediate information can be put into long-term storage.

l9
5. Information must be disseminated through the organization (absolute vs. distributed information).

Rules and procedures must be established to manage the distribution of information. (3 9-4 1)
Once information has been incorporated into organizational memory it can contribute to production,
innovation or maintenance by facilitating tasks such as:
- Diagnosing quality problems or equipment malfirnctions;
- Learning the identities of extra-organizational experts, influence peddlers, resource providers, or
other useful non-members;
- Locating information or resources that cannot be located using oflicial, standard sources (George P.

Huber 256-58).

The information required to accomplish these tasks must be distributed in such a way that it is available
when and where it is needed. Distributed information results from communicative behaviors. The forms of
communication—environmental, motivational and instructional—serve to make sure their concomitant
forms of information are distributed economically, in short-term and long-term memory, throughout the
organization. In order to accomplish this, organizations utilize functional—but not necessarily formally or
structurally—different communication systems:

1. The operational communication system, which “conveys” data about task-related activities and
operations within the organization...

2. The regulatory communication system, which “conveys” to the organization’s members those
orders, rules, and instructions which are intended to regulate their material- and/or
information-processing functions. . ..

3. ...the maintenance and development communication system serves. . .the purposes of providing
feedback as to the condition of the people or the communication channels upon which the

organization is dependent... (Thayer 103-4)
Thayer also notes that these systems facilitate the management of communication as well as the process of

communication itself.

There is an apparently recursive relationship between the functions of communication and the
structure of organization in which it takes place. Communication is crucial for defining an organization: it
is the means by which division of labor and delegation of authority take place. Yet, those are the very
constructs which determine how who communicates what to whom, when, and with what results.

Bounded rationality constrains the information processing capacity of organizations. In part, the structure

20
of an organization serves to distribute processing capacity, just as it guides and limits flows of

information. Farace, Monge and Russell suggest that there are number of factors which determine how

effectively a system processes information:

- Load is the amount of information that the system must process. Information overload may
compromise the operation of the system, but so—particularly if the node is a human who is wont to
get bored—may underload. Overload of one component may have a domino affect, leading to the

overload of other components when the first fails.

0 Rules or norms about how information should and may be processed must be known by the node,
agreed upon by other components in the system and must be accurate. For example, Rogers and
Rogers note that the active discouragement of the practice of “going over the boss’s head” not only
reinforces chain of command and delegation of authority, but also prevents information overload of
the boss’s boss (90-92).

- Attention—or processing capacity—must be applied to the task of processing the information. There
is also somewhat of a domino effect here, in that the more capacity a component has, the more it gets
used, increasing the likelihood of its failure. (10-12)

When the amount of information available outweighs processing capacity, organizations have several

mechanisms for dealing with the overload: successive—rather than simultaneous—judgements, memory,

varying the precision of judgements, and chunking (F arace, Monge and Russell 115). Unfortunately,
utilization of these techniques may result in important information not getting through—omission—or
changes in the meaning of the message—distortion. These failures can be compensated for by a number of
methods:

1. Redundancy is the repeating of a message in different forms, over different channels, or over
time. . ..
2. Verification is insuring the accuracy of a previous message.
Bypassing is the elimination of intermediaries in a communication flow. (92-94, emphasis in
original)
Unfortunately again, all of these remedies may contribute to information overload. The soundest tactics

for dealing with overload involve adding capacity or spreading the load out over time: increase the

2 1
channels, filter the information more, allow queues to form during peak periods. Beyond these tactics, the

organization must accept a trade ofi between the quantity of information processed and the quality of the

processing (Farace, Monge and Russell 117-24).

Rogers and Rogers offer these general observations about horizontal and vertical flows through
organizations:

- “Horizontal communication flows in an organization are more frequent than vertical flows.” Rogers
and Rogers hold that peers in an organization have similar views of the organization and their place
in it and do not hold sway over each other, consequently they are more open to interaction.
Horizontal flows usually serve to wordinate activities, “whereas downward flows are mainly
authoritative and upward flows chiefly provide feedback on operational performance; so vertical
flows carry messages that are potentially more threatening.” Horizontal flows are more informal and

move over channels that are more flexible and rapid.

- “Downward communication flows in an organization are more frequent than upward flows.”
Superiors naturally control communicative situations, so subordinates are less likely to initiate
upward communications. Communication between superiors and subordinates are more likely at the
upper level of organizations. Upward flows are also more susceptible to distortions which encourage
a positive image of the subordinate. (96)

Rogers and Rogers also note that there is a relationship between the physical structure of an organization

and its information flows:

Usually, the physical structure (walls, corridors, and floors) of an organization is established
according to the organizational structure, and the physical structure in turn largely determines the

communication flows. ...Physical proximity plays an especially important role in who interacts with
whom at the early stages of social acquaintance. (102)

Communication networks operate on three social levels:

1. Total system network—comprising the communication patterns among all of the individuals in
a system, such as an organization. . ..

2. Clique—defined as a subsystem whose elements interact with each other relatively more
frequently than with other members of the communication system. . ..

3. Personal network—defined as those interconnected individuals who are linked by patterned
communication flows to any given individual. (Rogers and Rogers 113-14)

22
These networks consist of primary links between the entity and the other entities with whom it directly

and frequently communicates and “weak ties” to entities with whom it communicates infiequently or
indirectly, yet from whom it receives a significant amount of information Weak ties can extend out
several iterations and may have significant effects on the subject at the “center” of the network (115).
Rogers and Rogers discuss the structure of interpersonal communication networks and note that open,
radial networks are more effective at gathering information than closed, interlocking networks (116).
They note a number of structures—circle, wheel, chain and all-channel—based on experiments by Leavitt
with small groups in a laboratory setting. These conclusions are offered as results from those studies:

1. Network centralization. . .contributes to rapid performance. . ., but the error rate is high,
presumably because two-way communication and feedback is discouraged.

Low centralization or high independence. . . is associated with member satisfaction.
The network structure served to elevate certain individuals into leadership positions.

Being in a key position in a network, however, also led to information overload for the leader,
through whom all the messages had to pass. (121)

While these studies offer some useful generalizations about how information flows through and
people interact in simple networks, Rogers and Rogers point out that they must be seen as only that—
generalizations—and used only abstractly. Most notably, they cite criticisms of the studies for not
considering independent variables such as opportunity, personality, noise, distortion, and distance (122).
Taking these criticisms into account, Rogers and Rogers offer a number insights into analyzing networks.
They suggest that analysis requires:

1. identifying cliques within the total system, and determining how these structural sub-groupings
afl'ect communication behavior in the organization;

identifying certain specialized communication roles such as liaisons, bridges and isolates...

measuring various structural indexes (like communication integration or connectedness, and
system openness) for individuals, cliques, or entire systems. (125)

F arace, Monge and Russell identify the general structural characteristics of nodes in a
communication networks:
- Isolates do not participate in any group.
° Group members participate primarily in a single group.

- Intergroup linkers participate equally in multiple groups. (185)

23

These roles are identified more formally as gatekeeper, liaison, opinion leader and cosmopolite by Rogers

and Rogers (133). Ronald E. Rice characterizes various communication roles as routing and distributing,

summarizing without changing meaning, delaying and storing based on priority, and modifying, editing

or distorting content (67). Robert W. Zmud considers the functions that nodes may perform and offers a

much greater characterization:

Sensor nodes detect environmental signals and transform these into messages. Sensor nodes can
exhibit distinct types of behavior: as scanners (gathering information that has no immediate use), as
trackers (routine monitoring of issues, events and entities), and as probers (one-time exploration of a
specific issue, event, or entity).

Filter nodes operate on input messages in order to screen out those which are irrelevant. Relevant
messages are maintained within the [system]; irrelevant messages are lost.

Router nodes distribute an input message to particular nodes and to particular information bufiers.
Carrier nodes move, or “carry,” messages to and from processing nodes and information bufl‘ers.

Interpreter nodes apply existent guides-for-action and contexts-for action to input messages to
enhance the information value of a message. As a result, uncertain or equivocal messages are
interpreted, i.e. made more meaningful.

Learner nodes apply existent guides-for-action and contexts-for-action to uncertain or equivocal
messages in order to create new score-cards, triggers-for-action, guides-for-action, or contexts-for-
action. Without these new schema, the input message would remain uncertain or equivocal.

Modifier nodes transform input messages but do not add value to these messages. The information
potential of an output message is at best equivalent to that of its associated input message. (99-100)

Based on previous discussions, Zmud’s characterizations might be made more consistent by eliminating

the carrier node—that function is fulfilled by links—and adding a storage node, which Zmud would

identify as an “information buffer.”

Zmud characterizes the messages that might flow through communication systems as facts,

observations, projections, beliefs, opinions, as well as:

“Guides-for-action represent norms, policies, plans, instructions, rules, or procedures indicating
‘what to do’ and ‘how to do it.”’

“Score-cards represent measures of task performance—what was accomplished and how well it was
accomplish

“Triggers-for-action represent signals to commence a certain action.”

24
“Contexts-for-action represent the provision of data on the nature, history, and objectives that invoke

a particular schema for a work situation.” (99-100)

Communication technologies in organizations

Everett M. Rogers defines communication technology as:

...the hardware equipment, organizational structures, and social values by which individuals collect,
process, and exchange information with other individuals. (2)

Further, he quotes McLuhan:

All communication technology extends the human senses of touching, smelling, tasting, and
(especially) hearing and seeing. Such extensions allow an individual to reach out in space and time,
and thus obtain information that would not otherwise be available. (Rogers 2)

Rogers recognizes the “digital revolution” that is taking place in communication technology and

goes on to enumerate the impacts of inexpensive digital technologies on human communication:

Communication media allow for greater interaction than has ever before been possible, with the
debatable exception of face to face communication. A prime example of this interactivity is the
increased flexibility of many-to-many communication. Interaction relies to a great extent on the
apparent “human-ness,” or behavioral variations resulting from self-knowledge, of participants in
the exchange. Increasingly, machines facilitate interaction or have a certain amount of human-ness
themselves.

Control of information flows through communication systems is moving from the producers of 1
messages to the consumers, a process Rogers calls “de-massification.” This is a result both of the
implementation of different architectures and the declining price of computer technologies.

New communication technologies allow communication to be asynchronous. The necessity of
participants to be proximate to each other—or conversational referents—in time is being eliminated
by devices which can “store and forwar ” large amounts of information at the whim of the user. (4-

5)

All media may be described using a set a four characteristics, according to Rice:

' Constraints are inherent characteristics that limit, hinder, or obstruct the freedom of
interaction between users.

25

- [Richness] . . .is the diversity of cues that a particular medium can transmit.

- Interactivity is the degree to which participants in a communication process have control over,
can exchange roles in, their mutual discourse.

- Network factors concern the effect that the pattern of communication has upon the input,
conversion, and output of interactions. (66-7)

William R Johnson, Jr. envisions systems that might be used to move any information, anywhere, at any
time. This type of system must provide:

- connectivity

- interoperability

- manageability

- distributed applications and connective services (152).

In light of the capabilities of new communication technologies, the question suggested by Thayer,
above, “Who communicates what to whom, when and with what afl'ects,” must be amended to include
“how.” This addition makes the question circular because the answer to “how” will change the answers to
the other parts of the question. In order to effectively and efficiently use new communication technologies,
organizations must consider:

- The goals they wish to achieve;

- The strategies and tactics that may be used to achieve their goals;

- The tools and means applicable to their strategies and tactics.

- The impact of the technologies on goals and strategies.

The selection and implementation of a communication system requires the organization to understand
what needs to be done and what the technology can do. Ronald E. Rice states that,

[elstablishing a good fit between media and communication activities requires an understanding of

the characteristics of different media, possible interactions among characteristics and activities, and
the media usage patterns of information workers. (66)

This requires consideration of technical, economic, satisfaction, political, cultural and cognitive criteria
(Rice 70).
Peter Keen maintains that communication technologies were originally implemented to fulfill a basic

firnction—conversation over a distance—and not to achieve any specific goals. At this stage, telephone

technology was utilized to support business operations with little or no consideration of the various types

26
of criteria for their implementation—they simply evolved. The technology was not viewed as a strategic

factor to the success of an organization. In the next stage, communication technologies evolved to be an
internal utility. This was fostered by increasing phone costs, growing competition in communications
markets, and the emerging strategic significance of data processing. Today, communication technologies
represent a coordinated business resource that is characterized not only by converging systems and
distributed processing power, but also by the emergence of communication systems as a key strategic
factor. They are increasingly seen as sources of revenue for the organization, rather than operational
expenses (Keen 15-20). Browning notes that information systems can be used to boost productivity, gain
competitive advantage and improve responsiveness. But, he says these are the ideals, of which the realities
often fall short (5). Keen cites four general reasons for contemporary organizations to integrate new
communication technologies into their business strategies:

- Operational necessity—to keep up with the basic level of service in one’s industry

- Defensive necessity—to protect one’s market

- Competitive opportunity—to steal an edge

- Breakaways and preemptive strikes—to change the rules of the games for competitors. (3)

A similar list is offered by Browning, with a slightly more functional perspective. He quotes Venkatraman
in describing five stages of adoption:

1. Automating existing jobs. . .typically to boost productivity. . . [l]ittle changes but the number of
people and the capital costs of doing business.

2. Electronic infrastructure. Islands of frmctional automation are linked together. ...[but] without
change there is usually little economic incentive to overcome the inevitable technical
incompatibilities and to battle over who does what. . ..

3. Business-process redesign. . . .computers enable things to be done in new and more efficient
ways.

4. Business-network redesign. Creating links with suppliers and customers not only creates new
opportimities for changing business processes, it also changes the balance of competition.

5. Business-scope redesign. As part of the process of self-irnprovement, information technology
enables some companies to move into new businesses. (1990, 7-8)

The strategic value of information technologies is determined in part by the operations they support
Traditionally, IT has been used tactically, to automate the “back office,” tasks such as accounting, payroll
and inventory, corresponding to Venkatraman’s stages one and two or Keen’s “operatational necessity”

and “defensive necessity.” Much of the emphasis in the re-engineering of corporations has focused on

27
Venkatraman’s stage three and Keen’s “competitive opportunity.” In these cases, systems that support the

“front office,” executive decision making, product design, market research. The next step, which is
already being taken by a number of innovative companies, is to incorporate information systems into the
product itself. Margaret E. Guerin-Calvert and Steven S. Wildman note that this entails using information
systems for processing of transactions, facilitating consumer search and evaluation, and product delivery
(27). Only in rare cases are these fimctions, which correspond to Keen’s “breakaways and pre-emptive

strikes” and Venkatraman’s stages four and five, provided by information systems, all three are not.

Another perspective on the functionality of information systems is ofl’ered by John Browning, who
suggests that information technologies are most effective when they support humans rather then replace
them. He suggests that IT best fulfills the functions of:

- Assistants that coordinate activities and control devices or processes;

- Advisers based on artificial intelligence such as expert systems that can provide guidance in making
decisions, diagnosing problems, and identifying opportunities;

- Communicators that link together individuals in flexible information flows. (5-6)

More generally, Rice contends that computer-mediated communication systems can serve both as a
medium and as content. Communication systems are an innovation that organizations must process, as
well as systems which process information about innovation (73). Based on this concept, Rice offers a bi-
dimensional model, which correlates processes of input, conversion and output with the two roles of
medium and content, as a means of identifying possible uses of computer systems in communication.
Under this model, as an example, computer-mediated communication provides information for making

decisions more effectively and a means of transforming the organization to operate more effectively (74).

Applying information systems to any segment of an organization’s operation may have significant
benefits, but possibly the greatest benefits will be had from total integration of all units via information
systems. The result of maximizing the strategic potential of communication technologies is a basic
transformation in how the organization operates, a process Stan Davis and Bill Davidson call

“informationalization.” The tactical characteristics of this process are:

28
Customized products + rapid response;
Manufacture at the point of delivery,
Shrinking overhead, inventory and working capital;
High service standards;
Inter-organization bonding;
Globalization. ( 13)

Lois R Brass and Helene T. Roos ofl'er a strategic triangle as the context for re-engineering business to

capture the full benefit of information systems. The triangle “consists of clearly articulated and tightly

linked technology strategy, business strategy and organizational change strategy.”

Huber maintains that the emergence of such tactical implementations will be evolutionary, rather

than resulting from strategic decisions. He feels information systems will most often be deployed to meet

operational necessity, i.e. to reduce costs by increasing efliciency or reducing staff. But, in general, “for

advancement of their own interests, organizational participants will use advanced information

technologies in ways that increase their own effectiveness in fulfilling organizational goals” (244). Huber

notes a number of characteristics of information systems that support strategic behavior by allowing users

to:

communicate more easily and less expensively across time and geographic location...
communicate more rapidly and with greater precision to targeted groups...

record and index more reliably and inexpensively the content and nature of communication
events...

more selectively control access and participation in a communication event or network...
store and retrieve large amounts of information more quickly and inexpensively...

more rapidly and selectively access information created outside the organization...

more rapidly and accurately combine and reconfigure information so as to create new
information...

more compactly store and quickly use the judgement and decision models developed in the
minds of experts, or in the mind of the decision maker, and stored as expert systems or
decision models...

more reliably and inexpensively record and retrieve information about the content and nature
of organizational transactions. (241)

These capabilities can be applied to answering the questions Franz F. Selig and Jack T. Nipper lay out for

defining business strategies which capitalize on information:

What information do you have that might benefit your customer or supplier?
What information does your customer or your supplier have that might benefit you?

29

- What information does one department of your company have that might benefit another.
- Can you make information more valuable by transforming it in time, format, and ease of use?
- Can you collect new data or aggregate old data in ways that make them more useful? (4)

Browning, quoting Jefl‘eries, suggests the questions may need to be more basic. It is necessary to ask two
questions:

Why should the change be made?
What is it that must be done to meet the new goals? (1 l)

Browning also suggests that an important function of information systems is not just to answer questions,
but to help users ask better questions. Organizations should use information systems to constantly recreate
the fabric of organizational memory. This will be done, he says, by implementing systems that unify all of
an organizations functional units, serving as the organization’s collective mind, learning as individual

members learn, acting in unison with each member’s actions (18).

There is some risk involved in deploying systems that are based on the answers to these questions.
Sharing and manipulating information will have benefits, but it will also have effects, beneficial and
otherwise, which cannot be easily foreseen by designers. First of all, it must be understood that successful
implementation of strategic information systems—such as might make use of the Selig and Nippers
questions—will have a profound efl'ect on the organization. Therefore, any significant system deployment
should be done either in conjunction with a general re-engineering of the organizations systems—as
suggested by Bruss and Roos—or with some caution and an understanding of how the organization may

be affected.

Sproull and Kiesler point out that communications technologies have both first- and second-level
efiects. First-level effects are those that are planned and which provide the impetus for implementing the
technologies in the first place. These effects can be generalized as either reducing costs or adding value.
“Second-level efl‘ects. . .come about primarily because new communication technology leads people to pay
attention to diflemnt things, have contact with different people, and depend on one another differently”
(4). Communication technologies also have a recursive relationship with their users, according to

Contractor and Eisenberg: usage shapes functionality as functionality shapes usage. Perceptions of new

30
media, pauernsofusageandbyextension, themediathemselvesdevelopandarereformulatedin

sequences, cycles and simultaneously (147-49). This is a process that is profoundly social in nature, and
rarely results in the fulfillment of rational expectations (143). Because of the unpredictable nature of how
information technologies will function in a particular setting, it is necessary to move fiom specification
and design, to implementation, to testing, usage and improving, back to specification and design, to

implementation, etc., (Figure 4) in what Browning calls the “virtuous spiral” (l l).

   
  

Specification and

Testing, using and

. . Implementation
Improvmg

Figure l. Browning’s “Virtuous Spiral”

Huber offers a very powerful—although somewhat complex—model of the impacts of advanced
communication technologies decision support systems on organizations. On a subunit level, Huber
maintains that these technologies will lead to involving a larger number and variety of people in decision
making, even while they reduce the time spent and the number ofpeople involved in face-to-face
meetings. Over the height of the organization, he predicts that distribution of authority will become more
uniform. Decentralized organizations will become more centralized and the converse will occur in
centralized organizations. Also, the number of levels involved in authorizations and message processing

will be reduced, along with the number of intermediate human nodes in organizational networks (245-64).

3 1
Organizational intelligence and memory, according to Huber’s model, will become more eflicient

and efl'ective. Both short-term and long-term storage will become less expensive and more rapidly
accessible, allowing problems and opportunities to be identified and solved or exploited in a much more
timely, accurate and comprehensive fashion. In general, higher quality decisions will be made faster

(Huber 245-64).

The chain of causality behind these events goes as follows:

- The availability of information technologies will lead to increased usage of those technologies.

- The increased usage of information technologies lead to greater information accessibility.

- Greater information accessibility will cause changes in the function, form and structure of
organizations.

- Both the changes in organizations and the greater accessibility of information will drive

improvements in the development of organizational intelligence and decision making. (264-5)

Even if the effects of technologies are taken into account, designers must understand that the effects
will not be realized if the systems do not gain the general acceptance of the users. Bruss and Roos
maintain that not only must the technology be available and the business operations redesigned, but is also
necessary to have organizational culture and a level of readiness that will support the change. Senior level
support and employee involvement are stressed by both Keen and Bruss an Roos. Both also underscoring
the importance of user education, Keen in particular notes that there are three levels to education
necessary for user buy-in:

1. Training: How do I use the system?

2. Support: How do I use the system in my work?

3. Education: How do I use it my way, in my job? (235)

An important factor that in both the acceptance of a system and its firnctionality for the organization
is its performance. If the performance does not meet the expectation of the users, they simply will not
utilize it to the extent envisioned, if at all. Ifthe performance characteristics do not firlly meet the needs of
the organization, both initially and alter the system is accepted and in full use, the results may range fiorn

disappointment to disaster. Then the general guidelines for the system must be set forth based upon the

32

goals that it will be used to accomplished and how it will assist in accomplishing them. From those

guidelines performance criteria can be set. Any configuration or system that falls outside of these factors

would be immediately eliminated fiom consideration Some of the factors that must be considered are:

Reliability - If an information system is to be used for mission critical operations, it must be
available and operating when it is needed. Some applications have sporadic or constant usage
patterns that require the system to be up and running always. Getting this level of reliability may be
very expensive.

Capacity - Transmission, processing and storage capacity are vital considerations for any system.
Once the system has been implemented even marginal performance gains over previous systems may
result in unprecedented amounts of usage. The capacity of the system must be carefully considered,
and there should be contingency or upgrade plans in the event of over utilization.

Ease of operation - The system should be well-suited to the task. Factors in the work environment
and those dictated by the type and amount of use must be considered. Also, the operational literacy
of the user must be taken into account, especially if the customer will be the operator. Educating
users can be very expensive and is of paramount importance to system acceptance.

Flexibility - The system may need to support a variety of tasks or applications. Typically, a system
that does many things, does not do any of them as well as a system that performs only limited
functions.

Interoperability - Communication and information systems may add the most value through the
ability to share tasks and data. This is especially true when considering several dedicated systems

rather than a single, flexible system.

Any system will need to be managed and maintained. While this may seem like a foregone conclusion,

compromises in the manageability of the system can adversely impact its performance and contribution to

the bottom line.

Maintenance - Ifthe system fails it must be easily repairable, ideally without any noticeable effects
on the users. Maintenance features include automated fault management and correction, operator

notification, multi-level alarms and fault diagnosis.

33
Management - Changes, upgrades and moves should be able to be made with minimal impact on

performance.

Accounting - System usage must be tracked. This may only serve the purpose of verifying service
provider billing, but it may also be used to monitor productivity, charge usage back to subunits or
individuals, or to bill customers.

Modularity - A system which is composed of well engineered modules is more flexible, expandable,
reliable and economic. Upgrades and repairs can be accomplished by switching out components. The
components may be sold competitively by several vendors, giving the manager greater choice in both

component quality andcost.

ENTERPRISE NETWORKS

Enterprise-wide computer networks—or simply enterprise networks—are the result of on-going
efforts by organizations to integrate the various, oflen dissimilar information systems of their fitnctional
units. The impetuses for this trend are to:

- share resources such as programs, data and equipment regardless of geographic location;
- increase information system reliability with duplicate files and redundant processors;
- Save money with the better price/performance ratio of small computers. (Tanenbaum 3)

Enterprise networking is not a “technology” or type of information system, rather it is a general term for
the complex form of communication system that emerges when all of the information resources used by an
organization are combined as a greater whole. We have seen how information systems influence and
interact with goal-oriented social systems. By examining the components, processes, and operation of an
enterprise network, we will develop a network model that might be applied on a larger scale, to

community, the “social meta-system.”

The Function of Networks

In the average office, the computer network fimctions on numerous levels. It connects the computers
together, it carries e-mail messages, it allows the CEO to track production and sales. Networks are
complex and dynamic systems which are closely linked to the culture and economy of an organization. A
computer network can reveal a lot about an organization just as Rogers and Rogers, quoted above, say an
organizational chart can. How hardware and software resources are allocated says nearly as much about
organizational goals and values as the use of real estate, personnel, or fuel. When organizational functions
are automated, eliminated, or created using information technology, that organization is making a
statement about and selecting a determinant of its structure, form, and function. The classic challenge for

implementers of computer networks has been to build a system that is both true to and positive for the

organization.

34

35
So what basic function is both true about and positive for all businesses? Cost reduction.

Organizations must maintain a positive difierence between inputs and outputs, i. e., income must exmd
expense. Not maintaining this excess value will eventually (the U. S. national debt notwithstanding) lead
to system failure. Networked micro-computers have provided the greatest cost saving by allowing
organizations to move away from monolithic, “legacy” information systems—mini- and mainframe

computers.

The first generation of business computer applications eliminated armies of bookkeepers and clerks,
by performing accounting functions much faster and with greater accuracy than a human could.
Ironically, the centralized computing paradigm, and the mainframe/mini with it, was replaced in much
the same way by networks. Beginning in the late nineteen-eighties, firms began to realize that the
personal computers (PC) that they were using for clerical functions such as typing letters or creating
simple spreadsheets, could be interconnected and used for enterprise-wide accounting and coordination,
eliminating the nwd for a mainframe (Herb Brody; Bill Musgrave; G. Hunter Jones; Bill Machrone; Marc
Kustofl', et. al.). The process of replacing mainframe applications with network services, which is
generally known as “downsizing,” allows organizations to eliminate regular sofiware fees and hardware
payments, high-paid technology specialist positions, and exorbitant power bills (Machrone; Jim Seymour;
John Gantz). During the early nineteen-nineties large organizations in finance (Evan I. Schwartz; Alice
LaPlante; Jean S. Bozman, “Mainframe-to-PC LAN shift taking hold”), law (Kim S. Nash), government
(Christopher Lindquist), medical (Kirk Johnson), utility (W. E. Delcomyn), retail (Kathleen O'Malley)
and manufacturing (Michael Fitzgerald, Susan R Nykamp) made the transition from closed, proprietary
legacy systems to integrated, networked, enterprise-wide systems (Paul Pinella; Gantz; Laurie Flynn;

Bozman, “Bank of America tests downsizing: moves host apps to PC LANs across 11 states”).

Micro-computers can save costs in the same way as their larger counterparts do—by replacing
human workers—but in the smaller organizations that were unable to justify a mainframe, the micro-
computer did not supplant secretaries, clerks, and bookkeepers, it supplemented them: thus the term
“personal computer.” Networking saved costs for personal computers by allowing the sharing of expensive

resources such as printers and disk drives (Cheryl Spencer, Donald J. Little). Networking had one other

36
important cost function: just as it distributes processing in a technical sense, it also distributes costs. The

cost of personal computer networks can be easily assigned to specific departments, business processes, and

individual personnel (Musgrave, Kustofl).

This is not to say that personal computing and enterprise networking are inexpensive or easy.
Achieving ideals for an information system laid out in the previous chapter will be costly no matter what
form the system takes. It took some three or four years for the networking technology to mature in the late
nineteen-eighties and early nineties, building a LAN could be an expensive and problematic proposition
(Musgrave; Seymour; Frank Blau; Mark Schlack). In general, PCs and networks provided a much better
price/performance point than mainframes (Pinella; Gantz; Flynn; Bozman, “Mainframe-to-PC LAN shifi
taking hol ”), and the performance and prices of network systems are continually improving. Networks
have also become a means of synergistically combining personal and mainframe computers, preserving
the value of the legacy (Joe Vincent; David Hughes), and proving to be very cost-effective overall (Lynda

Radosevich).

In general terms, information systems function to reduce the costs of searches and transactions.
Locating resources is in itself a process that uses other resources. By making resources more available,
easier to locate and identify, information systems reduce search costs. Capturing the value of resources is
also a process that uses other resources. By facilitating the coordination, tracking, and control of activities,
information systems reduce transaction costs. Network systems also facilitate cost control because they can
be implemented in phases or incrementally, and costs can be distributed or imputed to more firlly. Because
network fimctionality comes in smaller units (i. e., micro-computers), and these smaller units can be
flexibly combined, the scale and scope of a network system are easier to alter than in legacy systems.
Usage of network resouces can be completely logged and charged to user, department, or project accounts.
This level of cost control keeps networks tied very closely to demand for an organization’s product, the
cost of production, and cost justification, i. e., if a car parts distributor has seen steadily, but slowly
increasing business, its management can flexibly add workstations, printers, etc, as needed, for a

resonably modest cost. If business suddenly booms, the entire system can be supplemented just as easily.

 

37

Production is, of course, a general process that is vital to any organization. One of the major reasons
for deploying LAN s is the boost in productivity that can be accomplished through an effective
implementation of information technology. As networking emerged in the late eighties, professional
publications and pundits evangelized the technonology (Latamore, Little). And as the business world
began climbing up the networking learning curve, many specific examples of impressive gains appeared.
Firms such as K—Mart (O'Malley), Allen-Bradley (Nykamp), Lloyd’s of London (Boris Sedacca), and
Celestial Seasonings, Inc (Michael Fitzgerald) found that networking allowed for better coordination and
control. Similar results have been identified by academic researchers such as Sproull and Kiesler, Peters,
Drucker, and Keen.1 What this range of sources shows is that the productive benefits of networked
computer systems are, not surprisingly, primarily captured by knowledge workers: managers, clerks,
secretaries, agents, designers, authors, etc. But we also see second level contributions to the less ethereal
forms of production. Manufacturing processes can be tuned or completely redesigned to improve economy
or quality based upon accounting and analysis performed by using information systems. Production can be

better tracked and coordinated in order to reduce duplication or unproductive time.

The general function of supporting productivity has most recently spawned two market trends:
groupware and remote connectivity. Groupware is a form of computing that is focused on supporting the
operation of teams or “flat” organizations. It is particularly well-suited to synthesizing inputs from and
coordinating the activities of disparate (either physically or logically) organizations. It functions to allow
users to manipulate the components, processes and relations of a software environment that is analogous
to their organization’s memory, sources, sinks, and processes. The goal is for the participants/users to
become highly integrated with the organization, to develop an in-depth familiarity with the organization’s
form, and to be able to react effectively and quickly to changes such as new opportunities or dangers (Paul

Bernstein; Michael Miley).

 

1See the third chapter, “Enterprises,” above

38
The same concept underlies remote connectivity, which may be seen as the structural component that

underlays groupware and similar forms of information systems. Remote connectivity is simply the ability
to extend the reach, or the scale, of a computer network, and points out the contribution networking makes
to distributive processes. It allows geographically separate LANs or individual users to be linked together
as a wide-area network, or WAN. Remote connectivity boosts productivity in two ways, both of which
increase the availability and the flexible allocation of resources. First of all, it makes information available
where it is needed—on a construction site, sales meeting, car wreck, or battlefield—when it is needed.
Secondly, it allows personnel to remain closely linked to the organimtion even when they are in difierent

locations for extended periods.

Networks generally function to make an organization more productive through coordination based
on highly accurate and timely information, by handling a greater amount of information that can be used
to tailor a product or service to more fully defined markets, and by pooling knowledge resources—e. g.,
people—regardless of geographic location. Information systems allow organizations to respond more
quickly and fully to their environment by:

l. facilitating the collection of data,

2. the compilation of data into information,

3. the synthesis of information into knowledge, and

4. the application of knowledge through action.

Implicit in the function of processing of data into action are “networks of causal relations, including
closed-loop ‘feedbacks,’” that are so vital to goal-orientation. In contrast with mainframes, networking
realizes this concept in distributed structure, which can be flexibly reformulated and applied to
opportunities or problems, at any one of these four stages, and then can be quickly re-allocated to another
aspect or different problem. As a result, networks are well-suited to informal or unstructured processes,

and for allowing those processes to become more or less defined as required by any particular situation.

In turbulent environments reactive strategies are typically ineffective because, by the time a system

can react, what it was reacting to has gone or been transformed. Environments which undergo frequent,

39
rapid or chaotic clunge call for a proactive strategy, a strategy that is exemplified by innovative behavior.

It is this type of behavior that is best facilitated by flexible, informal, and relatively unstructured
communication processes and relationships. Ifwe remember our discussion of the enterprise social
networks in chapter three—particularly the examination of communication technology’s significance for
organizations—we see how information technology provided the greatest value when used to integrate the
various departments, operations or constituencies in an organization. A highly integrated organization
will involve more people in decision making, distribute authority more uniformly, reduce the resources
needed to turn information into action, and be generally efficient and effective in identifying problems and
opportunties, and at resolving or exploiting the situation. To restate Huber’s assertion: a well-

implemented enterprise-wide network will allow higher quality decisions to be made faster (245-64).

The contemporary concept of innovation as verbalized by progressive businesspersons and academics
can be metaphorically—and sometimes literally—characterized as a process of “changing the rules”
(Changm’ the Rules; Keen; Drucker; Peters). At the heart of this process is empowering an organization to
drive the market, rather than being driven by it, not through monopoly or cartel, but through continual re-
invention and strategic redefinition of the organization itself. While such an organizational state may be
undesirable in all but the most turbulent of environments, organizational innovation is an important
process. Whether a hospital nwds to redefine its operations to provide medical services in the suburbs
(Kirk Johnson), a university is attempting to change where and how students can work on their degrees
(Jon Nordheimer; Denis Newman), or a publication wishes to find new ways to reach its readers (Hugh

Morgan; Deirdre Carmody), networking fimctions to support innovation.

Innovation is associated with computer networks in multiple, intermeshed ways. First of all, as
discussed above, computer networks lend themselves to the specific innovations in production and
distribution, such as on-line shopping, distance learning, or mass-customization. More generally,
networks support the types of organizations that are best at being innovative, well-suited to operating in
turbulent environments: diverse, team-based, flat and unhindered by geography. Lastly, networking is

itselfan innovation with which organizations must deal. Building an effective network necessarily means

40
formalizing organizational information flows; adopting new work habits, new customs and conventions,

and even new personnel and roles.

Network forms

Networking has four forms, or technology paradigms, including “sneaker net.” These forms dictate the
ways in which information may be processed and applied. They are:

Stand-alone, or sneaker-net - Even if a group of computer users are not physically interconnected,
they can still share information by writing it onto some interchange media such as diskettes, or
paper, or by spoken word. In this form, individual performance and functions are defined by
each user, but the ability to capture benefits from rapid or flexible interaction is rather limited
because there is no network through which they might arise.

Terminal/host - All processing and storage is performed by a central computer, or host. Information
is entered and displayed on terminals or printers. In this form, the capacity, performance, and
firnctions of the system are centrally defined. There is no need for processing or storage at the
periphery. Individual users have little or no control over what what the system does or how it
does it. This form captures network economies to a limited extent, because it does not allow for
process or fimctional change to be initiated quickly or flexibly: it must occur system-wide and
all at once, or not at all. Terminal/host are not strictly speaking “networked” and do not
provide all the functionality of a network, but they do represent the precursor to networking
and can operate over a network infrastructure.

Client/server - All components of the network provide processing, storage, and input/output, but do
so within two general roles. A server provides specific, well-defined services that may be
applied to a range of processes or types of information. Clients on the other hand, may flexibly
use these services either by using multiple services or using a single service for several
purposes. In this form, processes and units of information are located on particular servers and
client computers. Major changes in or new forms of services must be initiated at the center of
the network, but new ways of using services and the adoption of new services are defined at the

periphery.

Peer-to-peer - In this form there is no center or periphery of the network. As in a sneaker-net,
performance and functionality are defined by each user, but local definitions are intimately
interrelated with system-wide definitions. Processes and storage may be firlly distributed across
the network, e. g., when you access a body of information, the data from which it is composed

may exist on various computers and the process of creating it may be flexibly carried out
similarly.

What this taxonomy of network forms shows is a progression in the integration of the devices on a
network and the network itself. On the lower end of network functionality, under the standalone or
terminal/host forms, changes in the needs of the end-user have little effect on the overall system. Indeed,
the host/temiinal form is designed to meet very specific needs of a meta-user: the organization. This form

can be effective for organizations because all of their personnel are working—consciously or otherwise—

41
toward a set of well-defined goals, within a set a well-defined roles. Mainframe-based host systems can

process large amounts of data, but also run complex programs that cannot be easily or flexibly changed to

meet the needs of organizational subunits.

Networks, on the other hand, because they are made up of relatively small, standardized components
and tie together systems that are on average of much more modest means than a mainframe, can be built
incrementally, flexibly and minimal corporate oversight. Many LANs have evolved in this way: a
department hooks its computers together to share a printer and to facilitate moving files between
computers, a division ties all of its departments together so its managers can better track and coordinate
the activities of the various departments, corporate executives see that all the littles LANs are getting out
of hand both technically and financially, and that they can better control costs and capture additional

benefits by interconnecting all of the divisional and departmental LAN 5 as well as stand-alone computers.

As much as network form is conceptually defined by the processing power and interconnection of a
network’s components, so is it practically defined by the types of software that run across it or make use of
resources through it. There are numerous forms of software, including:

0 Operating system (OS) - The OS defines the basic processes that go on inside a computer. They
also include higher level functions for identifying and accessing units of information.

- Network operating system (NOS) - Much as an OS defines the information process within a
computer, a NOS defines how a computer takes part in information processes that occur across
the network.

- Applications or programs - Applications are units of information that interpret other units—
called files or documents—of information, and what operations or processes might be carried
on those files. There are both general and specific types of applications; examples of these are
listed in table x. It should be noted that such a typology is deceiving because specific
applications often have the characteristics of various general types.

- Documents or files - Files are units of information that are recognized by an operating system,
and documents are files that are associated with a specific application.

42
Table 1. Examples of general types and specific computer applications.

 

 

 

 

 

Type of application Specific applications

Word processor Microsoft Word, WordPerfect, Claris
MacWrite, Lotus AmiPro

Spreadsheet Lotus l-2-3, Microsoft Excel, Quattro Pro

Desktop publishing Quark Xpress, Aldus PageMaker, TypeStyler,
Print Shop

Computer-aided Design Autodesk AutoCAD, Microstation,
MicroCAD

Database management Sybase, Paradox, Microsoft Access

Software facilitates various processes: filing, data management, production, viewing, development,
messaging, and recreation. Filing functions are typically integrated with a computer’s operating system
(OS), and the specific operation and the syntax of the filing system is determined by the OS. Today, OSs
typically include the ability to use the network as an extension of the filing system. More and more 08s
are opening their filing systems to provide hooks into other 085 and their filing systems, so that
computers running Macintosh, Windows, or Unix operating systems may share storage space through the
network. A network directory is evolving into a list of the network’s users and resources in a structure that
is similar to that of a computer filing system. While the OS takes care of things like computer and file
locations and names, the specific firntionality of a system—the tasks for which it can be used or how it

might be applied—are determined by programs and applications.

Data processing applications support the data entry into, processing and storage of, and information
retrieval from databases. Data processing software, including database managers, spreadsheets,
accounting packages, etc., is used primarily for structuring and inter-relating data so that it may be
interpreted, and used to guide actions. Computer-based production processes pieces of information into
files which will be converted into real products: a report, newspaper, letter, or blueprint. Production as a
general form encompasses word processing, desktop publishing, report generation and other specific

forms of production. The logical counterpart to the production of files is the viewing of those files.

43
Typically, files and documents are viewed with the same application with which they were created. A new

class of software is evolving that allows users to view files regardless of the application that was used to
make them. Development is similar to production, except that the development process produces
interactive computer programs. Until the late eighties, “development software” was synonymous with
“programming language.” A trend towards more open, comprehensible, and useable development tools
has resulted in the development of multimedia authoring programs, many of which can be practically used
by a novice. Recreation is not a process that is typically important to the organization as a whole, but with
the tendency of some workers to goof ofl‘, either by playing a game of Solitare or with a mult-user Doom
session, it is a process that can have a significant impact on an organization. By definition, recreation
software doesn’t have any productive result other than the amusement of its users, which may be counter-

productive for the organization as a whole.

Messaging allows for writing, sending and receiving messages, including files and programs, and is
the only computing process that requires a network. As a matter of fact, low level messaging is a central
process in NOS5. The functioning e-mail, file transfers, and other types of higher level messaging is
dependent on the lower level messaging operations carried out by the network. Of course, other forms of
applications can make use of the network in much the same way as filing systems do. Client/server and

peer-to-peer networks allow applications to have hooks into data on, or the processing capacity of, other
computers.

A relatively new type of software called groupware is further defining the form of network systems.
Groupware emerged in the early nineties. Like messaging groupware is necessarily linked with networks,
but unlike messaging it provides a synthesis of higher level fimctionality that enables an enterprise to
flexibly access and allocate its knowledge resources. Rather than focus on processes such as production,
development, or data processing, groupware integrates some of the functional characteristics of these
processes into the process of coordination, and focuses on supporting, automating and innovating the

internal workings of the organization as well as its external functions (Miley, Zmud, Srpoull and Kiesler).

44
The final, and possibly most significant, formal aspect of computer systems—and networks by

extension—is the user interface, which determines how information is displayed on, and how a user may
interact with, the computer. It is through the interface that the user builds a concept or cognitive picture of
an information system. The major innovation in user interfaces has been the move away from text-only
displays and command lines interfaces toward graphical interfaces and WYSIWYG (what-you-see-is-
what-you-get) displays. This innovation also included the introduction of pointing devices, such as the
mouse, that allow users to physically select and manipulate information. The graphical user interface
(GUI) has revolutionized computer applications by revolutionizing how users understand and interact with
the information system. Typically, GUIs use real-world metaphors for the computing environment,
relating applications to tools, and files to things. Another strength of GUIs is that they can guide user
interaction with on-screen prompts that have both implicit and explicit meaning: a storage device may be

represented as a filing cabinet, and may even have a sign attached to it saying, “put your files here.”

The users of a network and the organization itselfare aspects of a network’s form. More than any
other aspect, how people use an information system determines its functionality. Even if a system is
designed to operate a certain way, users may well use it in other ways, for different functions. This fact
takes us back to a number of the points made above, about how enterprises use information systems.
Systems cannot simply be dropped into an organization, design and implementation must be followed by
assessment, redesign, modification, and reassessment as nwded to most effectively meet organizational
goals and capitalize on user habits. One way to facilitate the “virtuous spiral” of system implementation is
to make it flexible and scalable, creating a system architecture that responds to the demands, interests or
needs of its constituents and components. User education is also an important part of implementing a
network. The entire premise of a computer network is to empower an organization to interact more
efficiently and effectively with its environment. The members of an organization must be able to use the
system optimally, must be able to make it conform to their way of doing their jobs, in order to firlly
capture the benefits of networking. The form of a network, like that of any system, must change to fit its

environment, and users are a major element in the network’s environment

45

Network structure

Enterprise networking has been evolving incrementally ever since micro-computers made their way
into the corporate environment. But it can be an emergent phenomenon, can coalesce out of several
unrelated devices simply by adding a resource that they all need. One of the primary characteristics of
enterprise networking is that it can be, and often is, developed from the bottom up. Two workers in the
same office may need to share resources so they connect their machines together. Then others in the office
add connections, until the whole ofl'rce can share the resources via a LAN. A gateway to the mainframe or
rnini-computer is added to the LAN. Other offices or departments that have also been internally
networked, might then interconnect with the first. And so an enterprise network grows. This is a
generalization of the situation, but that is in part due to a second characteristic of enterprise networks:
they evolve along many different paths and take many different forms. Indeed, enterprise networks are
nearly as varied as the situations in which they are implemented Each organization has different

communication needs, technology strategies and diflerent mixes of information systems.

Another characteristic of all enterprise networks, associated with their incremental growth, is the
distribution of processing and storage throughout the network. Rather than consisting of dumb terminals
that rely on centralized processing facilities, enterprise networks are built in large part on powerful
personal computers, workstations and numerous, small-scale input, output, routing and storage devices.
All of these devices are segmented into either physical or logical zones that can—ideally—be managed,
maintained or reconfigured separately, without affecting other parts of the network. These devices are
software as much as hardware and the digital revolution is leading to new ways of defining and

interconnecting software devices, i.e., distributed object computing.

One practical result of distributing processing power throughout the network is that costs are also
distributed. The cost of the network is only marginally more than the computers and other hardware that it
interconnects. Micro-computers are not only relatively inexpensive stand-alone appliances, but they may
be purchased and configured independently and then integrated into the network when necessary. The

common equipment that supports the network’s operation is reasonably inexpensive, is becoming both less

46
expensive and more powerful, and its cost may be shared among all network users. For instance, an

application that is used by all members of a group may be centralized, saving the cost of licensing it for all

workstations.

The defining concepts for enterprise networks are “openness” and “connectivity.” The emergence of
micro-computers and the need to share resources between disparate systems contributed to an on-going
process of standardization in the computer industry. The ideal goal of this process is to define a standard
for “open systems” that allows computer systems—even if they are based on proprietary technologies—to
be easily interconnected and to share data input, output, storage and processing. To accomplish
transparent connectivity, hardware and software functions are defined in suites of protocols (Figure 2).
Each protocol layer in a suite is functionally separate from the other layers of the suite. A layer defines
certain functions that may be fulfilled by any number of technologies. It sends data to and receives data
from the layers immediately above and below it, but it “knows” nothing of those layers, it only knows how
to process the data in very specific ways and then to pass it on. Just as the layers of a protocol stack don’t
need to be aware of each other, so it is that users, operating at the top of the suite, don’t need to be aware
of the protocols that make data available to them. Indeed, the whole idea of a protocol suite is to take low-
level details such as bit sequencing and session naming off of the mind of the user, to totally eliminate any
work that is external to the task they nwd to accomplish, to make the computer and the network
disappear. The network provides either a shared set of protocols for all devices, or a means of translating

one protocol into another.

Operation

Transmission

Connection

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Application End User Application Application
Presentation NAU Service Service Protocol
Management
Session Data Flow Control Transport
Transport Transmission Communication
Control Protocol
Network Path Control Internet
Data Link Data Link Control Network Interface ”‘1‘ supp”
Driver
Physical Physical Physical Hardware
OSI SNA (IBM) TCP/I'P (Internet) NetWare (Novell)

Figure 2. A general comparison of network protocol suite reference models

The lowest level of a protocol suite defines the physical connection between the network and the

devices it interconnects, the basic network infiastructure. Things such as cabling, connectors, voltage

levels, data carrier encoding, etc. are defined by physical layer protocols. Often physical protocols are

closely associated with data link or interface layer protocols, which set out the traffic laws for accessing

the network. The data link layer is often subdivided into two layers. In particular, the Institute of

Electronic and Electrical Engineers (IEEE), which publishes the most widely implemented physical and

data link protocols, defines a lower level media access control (MAC) protocol and an upper portion, the

logical link control (LLC) protocol. The physical and data link protocols are the primary determinants of

the data carrying capacity of a network.

Local area networks (LANs) can be distinguished as either being shared access or switched.

Additionally switched networks can be circuit switched or packet switched. In shared access MAC

 

48
protocols such as ethernet and token ring, all devices on a segment of a network see all the data on that

segment. They simply watch the traffic, waiting for a packet that is addressed to them and ignoring any
that is not. LAN 5 contain traflic on segments and subnets. Segments are interconnected through bridges,
which can only process the MAC address information at the head of each data packet. A bridge knows
what MAC addresses are on either side of it, simply forwards or stops packets from passing to each

segment based on that information.

A recent development in data link protocols is the point-to-point protocol (PPP) for interconnecting
devices through the telephone network. Before the development of PPP in the early nineties, serial
connections between computers through modems and over the telephone network only allowed for
character-based links. PPP defines packet-based serial connections, and can be used through analog or
digital modems, across analog or digital lines. A character-based connection was fine for text-only
applications, when linking a terminal to a host. Packetizing the data makes not only for more effective
transimion of non-textual data, but it also makes transmission more efficient. Lastly, PPP is well

integrated with network layer protocols, consequently its operation is almost transparent to the end user.

Routers understand network layer protocols, as well as data link information. Network protocols, the
most common of which are internetwork packet exchange (IPX), intemetwork protocol (IP), and
LocalTalk,2 define how the network is subdivided and how traffic moves between these subnetworks. Just
as there are hardware addresses for the MAC layer, there are software addresses at the network layer.
Unlike MAC addresses though, network addresses can be set dynamically and designate a logical
grouping. Network protocols can operate over most any MAC protocol, e. g. 1P works over ethernet, token
ring, or PPP, as do IPX and AppleTalk. Routers are devices that move trafiic based upon their network
protocol information, and are used to create large, logical intemetworks out of smaller, physical local

networks. Often, the various physical networks are interconnected through a backbone. The actually

 

2IPX and Localka are proprietary protocols defined, respectively, by the Novell and Apple
corporations.

49
backbone may be a high-speed LAN, a router or a packet switch, and functions as a center of the network

in which traffic is concentrated on its way from one subnetwork to another.

Just as physical and MAC protocols are closely related, so are network and transport layer protocols.
Transport protocols define the interconnection of processes between communicating devices, basically
setting up and managing the flow of data between the devices. Typically, network services, which operate
at the session layer, are closely linked to transport and intemetwork protocols, too. The functions at the
top of the suite, application and presentation, are standardized by operating system, and are typically
totally independent of the lower layer protocols. Applications are the primary consumers of network
bandwidth, and applications seldom use network capacity conservatively because they are not designed to
be aware of or compensate for available capacity. The exception to this generalization is the TCP/IP
protocol suite. This suite, which is commonly referred to as the Internet protocols, or just IP, extends the
functionality of the network up to the application layer with protocols for messaging, file transfer,

terminal emulation, and hypertext links.

The [P application protocols are notable in that they are standards for applications which are freely
published and distributed, but they are not applications. The application protocols define how to
implement and manage the formal aspects of the network, but they are not the programs that assume those
aspects. They describe functions, but they don’t actually perform them. This means that any one can write
applications that adhere to the 1P protocols, and all such applications will interoperate. It is also important
to point out that many IP protocols implement a client/server architecture makes efficient use of network
bandwidth.

In a true enterprise-wide network all personnel who have any type of information dependency in
their job have access to a computer, even if it is just a shared workstation that they use to access e-mail.
Although computer usage may be concentrated in management, clerical or creative professions, computers
are distributed throughout the organization Mission-critical and large-scale services are provided
centrally, while lower-level services for printing or sharing files are usually provided departrnentally or by

geographic location.

50
The resources required to build, maintain, and operate the network infrastructure and central

services come out of the organization general budget, or from a budget for information systems. Divisions
of the organization that have their own information systems may purchase resources for them on the cpen
market and/or through a corporate information services (IS) division. Likewise, the costs of the network
infrastructure and central services may be imputed to divisions that use them by the IS department or chief
financial officer. The wsts of the network are recouped through the benefits provided by networking at

corporate, departmental and personal levels.

Information services departments implement and operate the network, hosts, terminals, clients, and
servers. The duties of corporate IS departments include, but are by no means limited to, installing
operating system and application software, controlling access and security, training users, evaluating
products and services, and troubleshootng network faults. Some organizaitons have departmental
technical specialists or system administrators, especially if they have their own server, but major changes
or improvements are handled by a centralized group of network engineers. While the functions,
components, and costs of a network are distributed throughout the organization, the responsibility to

designing, purchasing, building, running, and managing the network are generally centralized.

THE NETWORK MODEL

We have examined communities, enterprises, and enterprise networks as systems, as having firnctions for
their constituents or users, forms through which they interact with the environment in order to meet their
functions, and structures consisting of components and processes operating in support of form. Now our
task is to use these concepts as guidelines for developing a model of a public information systems that are
efl'ective, eflicient, and equitable. To demonstrate this model we will turn back to our referent city,

Chattanooga.

A community is a social meta-system, which is composed relational systems such as commerce,
education, and law. A business such as a bookstore or a school is also a social system. Commercial
enterprises are systems that are nested within a larger economic system, which in turn overlaps with other
relational systems. To extend this analysis, computers and networks are information systems that support
the operation and facilitate the functioning of organizations. A clerk at the bookstore uses the cash register
to record a purchase by a teacher for her class. The cash register is a computer that tracks the day’s sales,
sending the data to an inventory system, which consequently produces an order to replenish the stock
books. While this accounting system simply automates transactions, it could be adapted to helping
shoppers find the books they need, or to processing transactions automatically, both of which would make
the store more productive. The store could set up a service that not only helps clients locate books and has
the books delivered to them, but also allows local authors to distribute their work, provides reviews, and
announces literary events. An electronic bulletin board system (BBS)—a dial-up terminal/host system—
could provide these services inexpensively, but this change represents a significant innovation for the

store.

An innovation is not simply an internal change such as cutting costs or boosting production, it
changes the form of the organization and how it interacts with the its environment. In other words, the
“virtual” bookstore functions in a way that is difi'erent fi'om its real counterpart. If this form meets a need
or serves a purpose in the community that generates more income for the bookstore than what it costs
them to keep up the system, then it will most likely persist. But in order for the innovation to succeed the

51

52
community must react positively to it and adopt it; the innovation must diffuse out into the environment,

the community. The teacher would need to have a computer and a modem at home and/or at school, some

training in using the software, and some willingness to use a computer in order to use the system.

Even if the bookstore is a wonderful store, with a great selection, and helpful staff, and even if they
do an excellent job setting up, publicizing and running the BBS, few of its cheats are likely to go out of
their way to use the system. Customers that have and know how use the necessary hardware and software
may use it some, but probably not enough to generate significant cash flow. Even if the BBS reduced the
cost or improved the productivity of handling these clients, it probably wouldn’t be a justifiable

investment for the store.

The clientele can expand either by having more local users or by expanding its geographic s00pe.
While the latter might be viable it would require more capital and a place the bookstore in a competitive
market dominated by catalogs. Having more local users can only occur if the individuals, families and
businesses change their information seeking behavior and begin to use computers more. Rationally, this
will occur when the perceived benefits of using the computer exceeds the barriers to using it. And these
barriers are more than buying the computer, setting it up, and learning to use it. Just turning it on and
waiting for it to boot can be significant barriers to adoption. So, we see that basic economics hold true,
and that in order to prosper the network model must:

- optimize the benefits provided to the community and constituents, and

. lower the barriers to and cost of usage.

An information system benefits an organization by lower its costs, boosting productivity, and
facilitating innovation. I think its safe to say that most members of the Chattanooga community would
like to be able to live for less, get more done, and make more opportrmities for themselves. These are, of
course, the very functions of an information system: reduce costs, boost productivity, and facilitate
innovation. Organizations capture these benefits effectively because they are goal-oriented and are
predicated on economically producing a good or service. This also holds true for both commercial and

institutional organizations. Even relational systems such as education, religion, and politics must bring in

53
at least as many resources as they expend in firlfilling their functions. Indeed, this even holds true for the

family and individual. If families don’t keep income above expense, they go into debt and eventually
insolvency, just as an organization would. But, society cannot aflord to let people die as it can with most
organizations, so these people slip into poverty and the cost of supporting them is transferred to the

community.

Clearly, a community network based upon our model must have a real impact of the cost of living
and the quality of life in a community. Costs of living are attributable to food, clothing, health, housing,
transportation, education, recreation, and entertainment. How can a network reduce these costs? Well,
coststotheconsumeraresetinpartbywhatitcostsafirmororganizationtoprovideaproductsand
services. A network could lower what people pay for goods and services by reducing the cost of
production, marketing and distribution for the retailer. The consumer also has search and transaction
costs that are exemplified by such things as driving around town looking for a the right swim suit, or
havingtopayforabouncedcheckbecauseabankdepositwasn’tmade intime. Aswimsuit isn’t
something you’d want to purchase without trying on, but it would be nice to know which stores are most
likely to have a swim suit you’d like. Descriptive listings of items such as bathing suits it something that

is handle well by information systems, so in this way a computer network might reduce search costs.

An information system may allow a person to electronically make deposits and check an account
balance, thereby avoiding fees of bouncing a check, but there are number of other transaction costs that
networks might reduce. These are the costs that are caused by taking on roles in various systems or taking
on multiple roles within a systems. These costs are exemplified by driving to work, participating in a PTA
meeting, leaving the omce to vote, heating and cooling the house, being burglarized, or moving to be near
a better school. Networking minimizes these costs by supporting telecommuting, on-line conferencing,
electronic voting, home telemetrics, computer-assisted surveillance, and distance learning. Any one of

these services probably wouldn’t prompt the widespread adoption of community networking.

These applications can also increase a family’s “productivity.” Under a functional implementation of

the network model, a woman can meet with cheats, check on the status of a project, schedule a dental

54
appointment, coordinate preparations for her church’s summer bible camp, participate in a lesson at her

son’s school, and automatically alert the gas company that there’s a leak in the basement, all within the
course of a day, while spending time at home with her infant daughter. The network supports these
activities by providing access to distributed resources and multiple processes within a single system. The
network allows the mother to interact more effectively and efficiently with her environment Through the
network she can participate in multiple relational systems, go from one role to another more rapidly and

with less trouble, and even get guidance on the systems of values and customs associated with them.

Finally, a network should enable a community and its constituents to be able to innovate more easily.
It is interesting to note that what we have described as a healthy community is structurally similar to
enterprises that are good at innovating: diverse, organizationally flat, actively seeking opportunities for
improvement. Networks support these organizations by capturing synergies among groups with different
skills, experiences, and values, by automating the coordination and accounting functions performed by
middle managers and bureaucrats, and by allowing the organizations to examine themselves and their
environment from multiple vantage points and perspectives. Through a community network that fits our
model, a barber who hears a rumor about a plan to widen the street outside his shop, can check its veracity
through government records, assess its impact on his business, and express his support or opposition
easily and effectively. Or a group that wishes to promote natural conservation by attracting coo-tourism
can organize, publicize its views, recruit visitors, work with land owners, and document and share its
activities, successes, and problems all through a network, and in theory without ever having a face to face
meeting. This type of network will allow members of a community to locate persons with like interests,
work and play together, and economically share resources. By no means would all of these activities take
place on the network, rather it would add value to and formalize activities that already take place in the
community, and facilitate the initiation of new activities. As noted above, a computer network does not
function optimally by supplanting existing forms of interaction, but by supplementing and enhancing

them.

TThe key to the value of a network, even more so than many other technological innovations, is

dependent on the number and type of persons or systems that have adopted it. Network economies are

55
derived from having easy access to multiple and varied resources, with the cost of access shared among

numerous users. In other words, the cost per user and functionality of a network are directly correlated to
the number ofusers it has: each user adds value and reduces cost for each other user. There is a critical
mass of network users that must exist in order for the network to create more value than it consumes and
for it to persist as a functional information system. It is the synergy of a whole range of dynamic services
that can flexibly change to meet the needs of the market, the community, the users, that would take a

network to the critical mass of usage required to make it a sustainable innovation.

Interestingly, in order for a community network to reach and maintain a critical mass of users it
must use processes that parallel those of a community: recruitment, socialization and social control. In
order for a community network to succeed it must operate as a relation system. A community network
would be a particular form of relation system though. It would be one that formalizes the interaction of
traditional relational systems. In order to be successful, a community network must supplement and
enhance political, economic, and religious systems by providing linkages between them, rather than
supplant and replace them. A community network can be generally defined as a relational system that
reduces the cost and increases the benefit of interaction between other relational systems, their component

organizations, and the individual constituents of a community.

So we see how a network could function to reduce costs, boost productivity and foster innovation for
communities and their constituents. But what form of system, which applications and services must be
available on a network in order for it to fulfill these functions? There are three general aspects of the form
of a community network that can be identified immediately. First, it must have an intuitive interface and
be easy to use. Second, because communities are composed of numerous constituents with conflicting and
complementary agendas, a network must be applicable to various activities and needs. Lastly, the network
must have a form that brings people together, emphasizes commonalities, minimizes the effects of

conflict, and synthesizes divergent objectives into a shared agenda.

More specifically, a community network would need to be economical and provide some flexibility in

how the network is used, therefore it should take the form of either a terminal/host or client/server system.

56
Peer-to-peer networking could be an excellent architecture for community networking, but it is more

dynamic and less formal than the other forms. This points up an important point about the form of a
community network: for it to be adopted it must have one. A network is a rather abstract thing,
particularly when it is outside the familiar context of an enterprise. For people to get into the network,
both actually and conceptually, it will have to provide them with a means of thinking and talking about
the system as well as a physically accessing, using and benefiting from the it. Members of the
Chattanooga community are like any other people, for them to make a decision about whether something
could be of use to them, they have to know what that thing is, what it does and how to use it. Giving the

network a form is not only a technical or practical consideration, but a marketing consideration as well.

On the other hand, the network must be informal enough to redefine itself as it evolves and as the
community begins to learn about and really use the system. Ifwe refer back to Browning and his “virtuous
spiral” (page 39), we remember that the less information designers and implementers have about the
application of an information system, the more it will need to be refined and redesigned during
implementation. Such a mix of form and flexibility are best provided by client/server networking. Typical
terminal/host systems are probably too monolithic to adapt to the dynamism and informality of a
community, while peer-to-peer networking may not provide the form required in order to be adopted.
Additionally, terminal/host systems do not provide the type of intuitive, attractive interface called for by
contemporary aesthetics, and peer-to-peer systems don’t make efficient use of network bandwidth.
Client/server provides a set of general functions that may be applied to specific tasks, but also a level of
formality that allows users to place it within a meaningful conceptual context. And it doesn’t preclude
either other form, so that terminals may be used to provide low cost access, and a peer-to-peer form can

develop as the network and user sophistication evolves.

The form of a community dictates certain formal characteristics for network that will support its
processes. Communities are composed of numerous actors operating multiple relational systems. Each of
these systems are realized in specific instances as families, businesses, churches, schools, etc., each of
which exist in a discrete location, physically separate from each other. In order to tie all of these locations

into a network, there must be an infrastructure that can interconnect them. There are three alternatives:

57
- Make use of existing communications infrastructures such as cable and telephone physical plants,

cellular services, or satellite links.
- Use wireless network technologies for point to point and/or multipoint connections.

- Deploy new infrastructure that meets the specifications of physical and data link protocols.

A new infiastructure would be rather expensive to deploy, and while it would provide the best
overall performance there is no way to justify the investment without concrete evidence of the demand for
the services it would provide. Wireless technologies are also rather expensive, are not fully mature
technologies, and are designed for use within a space such as a building or for connecting such spaces to
one another. While LAN standards may be adapted to community networking, and are valid for deploying
portions of the network that may exist within a single location, it would be far more practical to use the
existing infrastructure. In other words, a community network will not take the form of a local area
network, which is limited to a floor or a building, but that of a WAN or dial-in network. Users would
access the network through the same wires used to provide telephone and cable TV service. Organizations
with their own LAN 3 could tie those networks into the community network in the same way they
interconnect their sites. As technologies and usage develop, access through wireless may become more
practical, particularly for interconnecting specific points, and it may even prove feasrble to deploy new

physical infrastructure such as a high-speed backbone that is specifically for the community network.

In Chattanooga, the network model would suggest that the City and County governments, the
Chamber of Commerce, the United Way and other public sector institutions would be interconnected.
Each agency’s member organizations would be connected through that agency. Non-profit or commercial
organizations could provide access to the general public via terminals or dial-in connections. All of the
oganizations using the network could work together to build a backbone for the network, to which they
would connect directly, instead of having to cross-connect with each other. In this case, it is likely that a
third party would have to be given responsibility for the backbone. This organization would be the

community’s version of the corporate IS department.

58

The central component of a dial-in remote access network is a router or communication server which
brings data traffic in from phone lines and routes it onto a LAN or another phone line. The LAN provides
a shared access medium that allows data to flow fi'om one device to other devices. Individual workstations,
terminals, or other LAN 5 may be attached to the LAN through the router. Between the router and phone
line is a device called a modem that dials the phone number of the destination LAN or workstation3, and
encodes digital information into the analog connection. On the other end, another modem answers the
call, decodes the data and sends it on to the attached DTE (data terminal equipment)“. The interconnected
agencies and network service providers in Chattanooga would each own and operate a router, as well as
the communication equipment and phone lines that would interconnect them. If there was a backbone,
each organization would need only one connection. If there was not, each agency would have to have and

manage multiple physical connections to the network.

The function of each device on the network is determined by the software that it runs and the
physical connections, or ports, it supports. A server, for instance, has a LAN port and runs service
software (“‘daemons” in Unix, “modules” in NetWare, “services” in AppleShare), a router on the other
hand has multiple LAN or WAN (wide-area network serial connections over phone lines) ports and runs
routing software that moves data packets from a source port to the correct destination port. These
functions can be combined so that a single device provides both routing and server firnctions. A
communication server is just such a combination. It has multiple serial ports and a LAN port. Router
software sends the data from one destination to others. Server software matches destinations to serial
ports, sets up and manages connections. Each organization that wished to provide dial-in access to the
community network would need a communication server, as well as software for managing usage and

accounting. The hardware and software needed to effectively connect a LAN to a community network

 

3Terminals do not have the intelligence to answer calls and provide a service, and so would not
typically be a destination for a network link.

4The term terminal typically refers to a video-display terminal, but it technically refers to an
input/output device at the end of a serial connection. A router is D'IE because it terminates the serial
transmission by moving the data to the LAN.

59
backbone can be purchased at retail for $1,500 to $3,000. A similar system that would interconnect a

LAN, eight dial-in users, and a community network backbone costs around 87,000.

People who wish to use the network could do so from a public terminal or computer, a home
computer, from a pay-per-use computer, or from a workstation on a LAN. Terminals would have serial
connections to a host, and that host in turn would be attached to the community network. Most computers
would need a modem and software—called “packet drivers”—in order to establish both a physical and
logical connection to the network. They would dial an access provider, be assigned a port on the router

and be connected to the network, all in a process that is transparent to the end user.

A communication server is just one type of server, and a relative simple one at that. Other services
include:
- File server - shares files and data storage capacity;
- Application server - shares expensive or specialized software;
- Database server - shares large amounts of data and large-scale data processing;
- Mail server - shares electronic mail with private mailboxes;
- News server - shares “delayed response” or asynchronous discussion groups;
- Hypertext server - share text and multimedia documents and provides links to other servers;
- Directory server - shares structured lists of network resources, organizations and individuals;
- Chat server - shares real-time conferences;
- Print server - shares the use of a printer or other “hard-copy” output devices like film printers, CD-
ROM recorders, digital audio or video recorders.
Servers can be set up independent of each other and by various organizations. The Chattanooga City
government may want to set up an application server that provides access to public records on their host
computer. The Chamber of Commerce could set up a hypertext server that features information about its
members. A camera store may let customers use a photographic printer across the network. Interactive
conferences may be nm by local arts organizations to discuss aesthetics and culture. A computer company

could set up some modems and phone lines, let people have he access to the community network, but

60
charge to use use their file, mail and directory services. A bird-watcher could put her collection of

photographers, recordings and observations onto her computer, run a file server program, and allow the
general public to access the files. Each user of the network would need a physical connection to the
network, network operating system extensions and client software in order to access services. The software
can usually fit on a single floppy disk. The software and hardware needed to connect a DOS-based PC to a

community network would cost about $200 and an hour to install.

The services do not dictate what kinds of information can be shared through them, they simply
provide persons with a medium, giving the author somewhere to put her work and giving readers a way to
find it. The service defines how, when and with whom information can be shared, but not what the

information is, how it might be used or interpreted.

The types of services are determined by the network operating system, specific operational
characteristics by the operating system, the server software, and cheat software. The NOS provides five
things:

- Seamless integration with workstation operating systems;

- Unique “names” for network devices that can be mapped onto media-level address;
- Hardware and software platforms for network services;

- A means of managing interconnected networks and routing data between them;

- Accounting, security and auditing tools for controlling access and tracking usage.

Multiple NOSs can operate over a shared network infiastructure, and services from several NOSs can be

simultaneously accessed from a networked computer.

There are several network operating systems (NOS) available today. For the sake of brevity we will
examine the four industry leaders: Apple Computer’s AppleShare, 1P, Microsoft WindowsNT-Advanwd
Server, and Novell’s NetWare. AppleShare provides a whole range of services on a single platform. The
basic functionahty of AppleShare is built into the Macintosh operating systems. The server software
provides a wider range of more powerful services on a larger scale than the cheat software. AppleShare

also supports third party apphcatioas for a range of network services, including access to the services

61
provided by other operating systems. While Maciatoshes can access the services of most any NOS, it can

be diflicult to make AppleShare services available to machines running a different NOS.

IP is unique. Because it is httle more that a set of protocols recommendations, it can be integrated
with relative ease into an operating system or apphcatioa. Also, these protocols were created by their
users, technologists and scientists, not manufacturers. The global Internet is based upon these protocols.
The Internet makes IP valuable for two reasons, one is that it allows users to get at information and users
all over the world. The other is that it is an excellent source of inexpensive “shareware” or “freeware”
server and cheat software, which can be used to set up servers inexpensively, as well as access these
services. There are implementations of IP service protocols—both cheat and server components for all
major N085 and 085. Also, protocols for other operating systems can be “funneled” through an IP

network.

The Novell corporation publishes the most popular NOS for corporate environments. NetWare is a
proprietary, multi-functioa software that runs on a standard PC-compatible computer. This software
allows users to “attach” to the computer across the network. Network services are provided by software
modules that load on top of the NOS. While only Novell publishes the core operating system, a wide
variety of server modules are available from third party developers and Novell. There are modules that
allow a NetWare server to provide 1? and AppleShare services. Some of the most advanced network

services are provided by Novell: video services, telephony services, and enterprise-wide directory services.

Microsoft WindowsNT is a “next generation” operating system which can be extended with the NT-
AS (Advanced Server) software to provide network services. NT-AS operates at the session layer and
above, and is totally independent of lower level protocols. Consequently, NT-AS services can be provided
across most any network structure: IP, IPX, or AppleTalk. NT-AS has many server functions built into it,

but as with the other NOSs, software that extends its functioaahty is available from Microsoft and third
party developers.

In order to provide a service on the network, an organization or individual would have to have a

network connection, a computer to use as the server “platform,” a network operating system and server

62
software for each service and the number of users they plan to support. The users of these services—the

people who actually put information into the system, use its apphcatioas, and get information out of it—
would connect to the network using the appropriate network software by running the client for the type of

service they wish to access, enter the name of the specific service, hit the ‘enter’ key, and they’re in.

Even with hardware and software that are inexpensive and powerful, a network cannot run without
the proper “wetware”: the peolple who make any changes and improvements, keep systems running, and
keep the users trained. There are two things that will stop any networking effort dead in its tracks:
technologies that don’t work, and users that don’t care. It is the job of the IS department to mitigate each
of these factors in enterprises, and there must be a similar group supporting the operation of any network.
This function will be especially important in setting where the computer is a major innovation, such with
underpriviledged or economically disenfranchised communities. It is the responsibility of the network
operations group to make sure users are getting the optimal value out of the system, and often times the
greatest value can be gained by the least empowered. In an enterprise setting these groups are relatively
easy to spot and can usually be motivated by rewards in pay or position. In a community, groups that
aren’t contributing can be difficult to identify, and even if they can located it may be very difficult to get
them to even attempt to find the benefit of networking: they don’t have an interest in being integrated, and

sure aren’t going to get paid for it.

Corporate support and training professionals long ago learned that the basic truism of computers—
“garbage in, garbage out”—also applies to users. You must understand how and why individuals do or not
use computers, and then evaluate and re-evaluate how they might gain the most value from the
information system Pat answers and canned instruction may make everybody feel like they’ve done
something, but have httle value when you’re back at the office trying to get something to work. Education
and trouble-shooting efl‘orts must suit the scale and scope of the user base, not the network. This means
finding out what matters to the user as well as identifying technical problems, and resolving both issues at
the same time. It means changing anitiudes as well as skills, recruiting users, building enthusiasm, and

providing proactive, hands-on, personahzed, real-world training.

63
Application of the model

The possible means of creating and supporting a community network can be considered in a continuum of
available capital and technical expertise as follows:

1. individual hobbyists,

2. community cooperatives,

3. small entrepreneurs or institutions,

4. pubhc computing consortiums,

5. local governments,

6. large entrepreneurs or institutions,

7. utihty companies.

Any and all of these actors may have a role in the implementation community networks. The enterprise
network model would suggest that community networks would start on the lower end of the scale, with

grass roots organizations, and involve higher- from those at the upper end.

Individual hobbyists would be motivated by the personal gratification of working with the technology
and the social interaction that would result from working on a community network, as well as any modest
network economies that would accrue. At this level the community network would necessarily be very
modest. It would be implemented in much the same way as a BBS, except that:

- users would have packet-based connections that would support terminal emulation as well as more
sophisticated apphcatioas that make use of the cheat systems processing and storage and,

- the system would be a server that could be interconnected to and share resources with other servers
as well as clients, users connected to one server would be able to access services across the network.

The the system owner would have to purchase a computer, network operating system software, two or

more modems, a multi-port serial device, and any software that might be used to provide services. The

total cost of such a system, providing IP-based services independently of the Internet, would be around

$5000, with monthly charges for phone lines and power.

64
A cooperative implementation of the network model in a be similar to that of the hobbyist, but with

greater coordination and poohng of resources. It would be formally similar to other cooperative
enterprises with a board of directors, a couple of paid employees, and a general membership. The pubhc
would be charged for use of the network or just given basic access, while members would have free and/or
full access. All members would participate in fitnd raising, but a core group would be responsible for
setting directions and pohcy, selecting technologies, building and operating the system, and training and
supporting the general members. One possibihty for a cooperatively deployed community network is a
system that focuses primarily on interconnecting non-profit arts, environmental, and/or social groups that
have some LAN infrastructure in place already and could benefit from interconnecting their information

systems.

A municipal community network would be a combination of a community information system and
an enterprise network, extending beyond the city departments to other community institutions and
households. Part of the motivation for such a network would be to support the operation of the city
govemmeat. The cost of the city enterprise network would be more justifiable if it could be shared with
other community subunits. It could make for a more informed, involved populace. One of the sticky issues
of community networking rears its head at this point: universal access. Ifthe community network is
provided by a government agency, all citizens must be given equal access to the network. Does that mean
a jack on their wall, does it mean giving every citizen a computer, does it mean opening all government
files for access on the network? On a computer network “access” can exist on multiple levels, across the
breadth of services on the network. Even if only basic information services are provided by the
government, excluding any provisioning of infrastructure, there will be a number of difficult questions

that will need to be answer in defining “equal access” on a community network.

Entrepreneurial provision of community networks is a wide-open consideration. Many BBSs are
operated for profit. There is a growing market and healthy competition for regional, national, and global
network services. Local markets are being opened to competition for telephony and cable television,
although there has been httle movement into more sophisticated network services. For any entrepreneur

there must be potential return commensurate with the risk of the investment. The only precedent for the

65
types of service that a community network would offer, and the way in which it would benefit its

constituency, is the Internet. The costs of a community network would be at least comparable to those of
any small business, while the return is questionable: how much will users pay for what services? sign-up
fees? usage-sensitive charges? service specific charges? flat rate? how do you set up a pricing structure?
how do you track usage, maintain accounts and collect? do you interconnect with other network

entrepreneurs? can you build enough of a user base to reach a critical mass?

Possibly the best way for a network to be provisioned by an entrepreneur is to focus on target
markets. One such market would be organizations with shared agendas, with much the same specific
functions as a cooperative community network, but providing services to commercial and semi-
commercial ventures for a fee. This may mean providing services along industry hnes, or vertically
between supphers and their customers. Another focused market would be apartment complexes. It would
be reasonably economical to build high-speed LAN s in apartment complexes, and even to tie the LAN s
together. A final entrepreneurial opportunity exists in provision of gateway services to the global Internet.
A community network could easily operate as a secondary function of such a firm, but most access
providers are national or regional in scope and not too terribly motivated to support community

networking.

Utility companies—broadly meaning all companies that currently provide water, energy or
communication services to the home under a functional local monopoly—could be the most viable
implementers of community networks. They have the capital necessary for such investments, the
infi'astructure and technical expertise for the job, and community networking would not be far from their
core businesses, which involve economically distributing resources to members of a community. Indeed, a
community network may provide the utihty itself with benefits such as more economical fulfillment of
their core business, greater customer satisfaction, and enhanced social standing. A community network, as
defined by our model, represents a significant innovation. Business considerations, regulations and
corporate culture might all work against the possibihty of utihty companies from building community

networks.

66
The network model doesn’t preclude the involvement of several of these sectors in creating a

community network. Rather, the model allows for various network structures and forms to be integrated
into a single system. Involving several diverse groups in the creation of the network would actually
contribute to the value of the system. Each sector could play a unique and significant part. If a single
group or organization deploys the community network, it scents reasonable that they would insist on
controlling it. On the other hand, bringing diverse groups into a project of this scope may result in

conflicts that could compromise the functionahty of the network.

The cash flow of a community network must be positive, but there are several ways to structure
finances in order to achieve positive cash flow. One way of generating income is to ask. If the community
network is operated by a non-profit organimtion, it could bring in firnding in the form or donations.
Similarly, the system could be at least initiated with a grant from a philanthropic or govemmeatal agency.
The system could have a membership, and membership fee, that would give a user access to special
services or discounted usage fees. Fees are of course the most tried and true method of generating cash
flow. Users could be charged a one-time membership fee, a reoccurring membership fee, per time or data
unit, per service, etc. Commercial enterprises may pay for space on the community network and for
services such as transaction processing, and the revenue they generate may be used to subsidize the system
for the rest of the community. Rather than charging for access to the system, the network operators may
charge to limit access to spaces users wish to keep private. Also fee structures may change over time, most
significantly they may be correlated to the resources on the network or may be set by defined

democratically by the users of the system.

Let’s review the network model what it tells us about a community network. A community network
would consist of home computers and pubhc terminals connected to private and pubhc local-area
networks via modems and phone lines. Servers, personal computers, and workstations would be directly
attached to the LAN 5, which would be interconnected by digital data cicuits. Constituents of the network
would fill the following roles:

- User - uses the network as a means of coordinating activities, locating facts and figures, setting goals

and solving problems,

67

- Access provider - operates and maintains a communication server or a router and provides physical
access to the network,

- Service provider - operates and maintains a server platform and manages its services,

- Information provider - provides content for services,

- Network operations - defines the architecture and technologies used in the network, is responsible
for operation and finance of shared portions of the network, and provides technical assistance and

education to users.

Access to the network would be free or inexpensive. Each service—including those that supply network
control and support—would have its own pricing structure and usage pohcies. Individuals or

organizations may fill one or more of these roles.

The network would formalize interactions between the constituents of various relational systems,
between the relational systems themselves, and between the community and its environment. Rather than
wondering where to find a book, a person would run their directory client software, type in the name of
the book, and the program would find any entries for the book in local bookstores inventories, hbrary
catalogs, and possibly even private collections. The network would facilitate the problem solving, conflict
resolution, innovation, and strategic behavior. By its very nature, a community network would provide a
broad context for the concerns and interests of its constituents, within which people might find

commonahties and synergism.

Just as enterprise networking has been an innovation that caused conflict and change in the business
world, networking will also be an innovation for communities. According to this model, a ftmctional
community network will have the same efl'ects on its community as networking has had on organizations:
more people will be involved in decision making, decision will be made more quickly, authority will be
distributed more uniformly across the social strata and among relational systems, centralized social
processes will become less so and vice versa, and the number of intermediaries between members of the
community will decrease. Community memory and intelligence will be improved as building, storing and

sharing information becomes easier and less expensive. Social problems and opporttmities will be solved

68
or exploited in a much more timely, accurate and comprehensive fashion. The network should allow the

community formalize decision making processes that higher quality decisions will be made faster and

more economically.

NETWORKING COMMUNITIES

It is beyond the scope of this thesis to apply the network model to a community. Such a project
would necessarily require the participation of the members of a community, would take several years and
significant amounts of capital to reach critical mass. Instead of building a network from the ground up,
let’s use the model to analyze some concepts behind community networking, some trends that may impact

the evolution of community networks and some community networks that are operating or being created

today.

Communities of interest and networks for social action

There are two conceptual approaches that use the ideas of community and networking together with
somewhat different intent than what we have identified as community networking. One view of
networking and communities sees computer networking as making community independent of place.
Intellectual community is an ancient concept, and is at the center of the relational systems that we have
identified and the primary components of a community. Computer networking formalizes intellectual
interaction while ehminating burden of having a body, the baggage brought to a relationship by physical
characteristics, such as age, weight, sex, physical ability, etc. The lack of a physical representation allows
one to be a member of a group that is without space: what Howard Rheingould calls a “virtual”
community, an “on-line community,” or more generally, cyberspace. The only thing that ties such a
community together is the shared interests of its members—and their concept of community. The Whole
Earth ‘Lectronic Link (WELL) and East Coast Hang Out (ECHO) are two such systems. While they are

located in San Francisco and New York, respectively, many of their users are outside those areas.

The Internet is, of course, the other example of virtual community—or communities. MUDs, or
multi-user dungeons (taken from the game “Dungeons and Dragons”), are particularly good examples of
electronic communities. MUDs are shared apphcatioas, in which the users build a virtual world where
interactions are described in text, and through which they type their ways. They have been noted for their

overpowering sense of community: there are stories of professionals and students totally abandoning their

69

70
work just to play in the MUD (Josh Quittner 138). These systems have distinct cultures, their own customs

and values, their own means of recruiting, socializing and controlhng members. Virtual communities
have the form of community even if they don’t have the substance. Their constituents think of themselves
as part of a community. But, while the intellectual environment of a virtual community may be
compelling, the physical environment is nonexistent. No matter how much you may get out of on-hne
culture, you still have to deal with what the denizens of cyberspace call “the meat world,” physical

existence.

The members of an intellectual community cannot benefit from network economies in the same way
members of a physical community can: car poohng, baby-sitting, crime watches, cooperative cooking, etc.
Most people can gain more value from interaction with their neighbors than with people with whom they
share only intellectual interests. Members of a virtual community simply don’t share material needs and
stresses the way members of a physical community do (John Perry Barlow 55). There are a number of
other inherent problems with on-hne communities. Carolyn Schafl’er and Kristin Anundsen point out
some of the inherent social problems that go along with virtual anonymity: intentional deception and
misperceptions, casual disparagement and “flaming,” and dominance by white males (121). This last
discrepancy has been pointed out by numerous observers. There have been concerns about who gets access
and at what price ever since the first telephone networks were built a century ago, and the problem of
access persists with the information superhighway. Barlow points out the general lack of women,
minorities, children, and handicappers in virtual communities (54). Reginald Stuart characterizes the
situation as the major civil rights issue of the next century, and foresees the rise of “electronic redliaing,”
the systematic exclusion of minorities from cyberspace. Lack of diversity is not only inequitable, but the
network model suggests that diverse communities will find the optimal synergies and greatest benefits of
networking. Whether it happens intentionally or not, lack of diversity hmits the resihence, flexibility, and

innovativeness of a system.

The other view of networking has nothing in particular to do with computers. It has to do with social
networks and the way people set and pursue their agendas in the context of their community. In this view,

networking is purposive social interaction, sharing concerns and ideas among a diverse, richly interrelated

7 1
group of people, a means of creating and reinforcing community. It is a definitely local phenomenon, one

that may not even be amenable to discussion without a specific local context, because every community is
unique in its character, richness and poverty (M’ g Cities Work ll-12). The sociologists who study
social networks advocate a proactive approach that is similar to that of the management scientists and
sociologists who study organizational networks. Peter Keen encourages business to “compete in time” by
using communications technology as a “coordinated business resource” (15). Donald Schon sees
networking as an innovative way to restructure social problems and put then in terms that express the
interests of all members of a community (53). Schaffer and Anundsen enthusiastically encourage the use

of social networks—personal, business, and activist—to build community.

There is a definite synergy here. While virtual communities powerfully hnk people but lack the
profundity of physical existence, real-world communitarians are working diligently to provide hnkage
between diverse interests and strugghng to cross the gulfs between relational systems. On-hne
communities are rather bland environments with technical and financial barriers that are too high for
many minorities. In contrast, enterprises have driven a market for low cost yet powerful means of sharing
information and are linking personnel throughout an organization. Management scientists have
documented the power of networked computer systems to innovate the form and operation of
organizations, allowing for greater participation (Sehg and Nipper; Zmud; Rice; Manning; et. al.).
Communities seem to be searching for the kind of equitable, efficient, and effective interaction that

network systems have been able to facilitate in enterprises.

The network model for communities is an innovative application of computer and communications
technology, an innovation that must be adopted by the community, and an innovation that makes a
community more economical and innovative in its own right. By bringing virtual community and social
networking together in the form of a community network, a community should be able to reduce the cost

of living and improve the quahty of hfe for its constituency.

72

Community Networks5

The first community networks were, by todays standards, rather modest systems. The first two
systems shared this modesty, but beyond that were very different. The Community Memory project in
Berkley, Cahfornia began in the early eighties, with its first test system going into operation in 1984. It
provided little more than indexed files that a person could view, respond to, or create their own. The
creator of a message chose keywords to be associated with the file. Files could be browsed or selected by
keyword. The system focused on recording the opinions of the members of the community, on putting the

people in touch with their community through computer-mediated many-to-many communication.

Case Western Reserve University (CWRU) started a small BBS in 1984 to test the feasibility of
disseminating medical information electronically. The system allowed members of the Cleveland
community to leave medical questions and retrieve the answer within 24 hours. It was so successfirl that it
was moved to a larger system, and the areas were expanded to include law, education, arts, sciences and
government, as well as interpersonal electronic mail. The system, called the Cleveland Free-Net, is
available for fiee, 24 hours a day to anyone with a computer, modem and phone hae. It became the model
for community network pioneers, and the production and distribution of it operating system was spun off
as the National Public Telecomputing Network (NPTN) in September of 1989. NPTN is a non-profit
corporation that sell a turnkey community network solution, including software, operating guidelines, and

a funding structure (“Community Computing” 2, 6-7).

Today, the nearly 50 Free-Nets and 30 plus other community networks consider unstructured
interpersonal messaging and expert assistance to be two of their most valuable services. Conferences on a
range of community issues use the ftmctions of both messaging and expert advice. Most community
networks have databases of local, regional, and national social resources available, and provide some type
of Internet access. Democracy is emerging as an important apphcation of community networks, and

exemphfies the apphcation of computer networking to social networking: people use the systems to

 

5Much of the information for this chapter was culled from a series of electronic survey responses that
are available through the Internet: ftp/l

73
become more involved in their governments. The Santa Monica Pubhc Electronic Network (PEN) was set

up by the City of Santa Monica in order to remove some of the barriers to communication that separated
the local government fi'om its citizens. The use of community networking to formahze pohtical processes
is a realization of social networking, and seems to be reasonably successful. On the other hand, the
systems are not deemed secure enough for the next logical step in electronic democracy: on-line voting. At

least one system, Santa Cruz County WAN (scruznet), exists solely to tie together local BBSs.

The demographics of community networks have proven to be somewhat more diverse than other on-
line systems. Most of the networks make it part of their mission to make information resources available
to all, and many have active out-reach programs. Some systems, such as Montana’s Big Sky Telegraph,
focus on providing hnkage to rural peoples. At least one system provides services primarily to Native
Americans. Notable successes in serving broad constituencies have resulted fi'om having pubhcly
accessible terminals. PEN has been successful in giving homeless people a human presence in the
community and provided resources that help the homeless find shelter and work. Professionals and school
children have found places on community networks. One of the most consistent features of community
networks is that their target audiences include non-profit agencies and schools. The New Mexico
Community Development Network (NMCDM) was created to provide a hnk between social welfare
organizations throughout the state. Many of the networks are actively supported by local school systems,

with the explicit goal of bringing together school children, teachers, and community at large.

Some systems have managed to build up an impressive list of supporting agencies. The Heartland
Free-Net in Peoria, Illinois, for example, boasts a hst of over 40 affiliated organizations including the Red
Cross, the Audubon Society, the Boy Scouts, HAM Radio of Central Ilhnois, Goodwill Industries, and
numerous local schools and universities. One constituency that is relatively lacking in community
networks is the business community. Roughly half the systems include their local Chamber of Commerce,
small business development organizations, and economic development agencies as key participants. Few,
if any, have a ratio of commercial to private or public participants that would reflect what exists in the

community at large. Also, many of the systems do not allow commercial activities.

74
All of the systems have made it part of their mission to raise the awareness of and comfort levels

with computer technologies for members of their communities. In general they seem to have met with

success. While the consensus is that the network apphcatioas—on terminals or cheats—are not as simple
and easy to use as they should be, the systems have proven to be useable by computer novices. Often times
new users have expressed surprise at how easy a system is to use, and that they derived some benefit from

using the system that made up for any difficulties they experienced.

The average number of total users on community networks averages about 400, running the gamut
from a few to several thousand. The National Capital Free-Net in Ottawa, Canada, for example reports
having 4100 user accounts in April of 1993, with some 75 new accounts every business day, and
averaging 2000 logins per day at that time. Big Sky Telegraph, on the other hand, is one of the best
known community networks and it had 2000 members and was averaging just over 100 logins in the

Spring of 1993.

All community networks have some form of organization that is responsible for deployment and
operation. Typically, it consists of a board of directors, a small paid staff, and as many volunteers as
possible. As with other specifies, the operational details of community networks varies greatly. The Santa
Cruz County WAN is a self-described “anarchist co-op.” Big Sky Telegraph, the Blacksburg Electronic
Village, and numerous others are nm by universities. Free-Nets are somewhat more consistent in
organization, at least in part because of the guidance of NPTN. NPTN strongly encourages its members to
be independent, non-profit corporations similar to pubhc broadcasting stations, with the commensurate
corporate structure and operations stafi‘. A few community networks are run by commercial
organizations—typically BBS or Internet access providers—as a pubic service. Chattanooga’s nascent
community networking organization is currently in the organizing stage and consists primarily of
technologist working in teams. The Community Link project, as it is called, has Marketing, Technology,
and Education teams consisting of two to five peOple. The system as it exists today is operated by the local

Internet access provider.

 

75
All of the community networks depend on their users and afiiliated organizations to provide

information, but only a couple rely on third parties to provide services. Under the network model, roles in
a community network should be divisible between users, access providers, service providers, information
providers, and network operations. In almost all cases there is extensive vertical integration of access,
service, and operation roles. The notable exceptions are NMCDM and scruznet, both of which exist to
provide a point of interconnection for their constituents—NMCDM for non-profit agencies, and scruznet

for BBS operators—and only fill the roles of access provider and network operations.

Community networks have financing methods that are just as varied as their organizational
approaches. The Free-Net system is based on voluntary donations in the form of memberships, and
corporate sponsorship. Free-Nets are dedicated to creating free pubhc information, and do not charge user
fees. Many of the other system do charge modest fees. Most of these systems provide access for fiee,
charging for services such as e-mail and Internet access. Community networks almost universally support
their operations with donated equipment and time. Technical expertise could be a major expense, but
community networks have managed to find a large base of computer and networking professionals willing
to donate their services for network operations. Many of the systems have federal tax exemption status that
allows donors to receive a tax deduction for donations. The other major source of filnding is from grants.
Several of the systems benefit from private foundation grants. Both the Corporation for Public
Broadcasting (CPB) and the National Telecommunication and Information Administration (NTIA) had
major grant programs in 1994. Through the Telecommunications and Information Infrastructure
Assistance Program (TIIAP), NTIA competitively awarded twenty-six milhon dollars that Congress had
appropriated "for the planning and construction of telecommunications networks for the provision of
educational, cultural, health care, hbrary and pubhc information, pubhc safety or other social services” in
fiscal year 1994. Much of this money went to seed funds to finance the start-up of new community
networks. The CPB has not only run a competitive grant programs aimed at community information
services, but its subsidiary, the Pubhc Broadcasting System (PBS) is actively providing dial-up computer

network components to its member stations.

 

76
There are two technologies that are overwhelmingly used by community networks. The first is

NP’I'N’s proprietary FreePort software. The other is a combination of Unix and IP. NPTN software runs
across multiple processors, but appears to its users as a single program. Services are accessed via a
terminal or terminal emulation program. The system doesn’t support a GUI or pointing device and can
only display text and simple graphics. While FreePort runs on Unix and supports most of the IP

apphcation protocols, it does not provide a packet-based connections or workstation-based cheat software.

Most of the non-Free-Net community networks provide packet-based (PPP) connections to IP
services and character-based connections to a hypertext host. Users of the system may purchase
commercially available cheat software, or the systems will provide shareware/freeware clients. A hand-
full of community networks use BBS software such as FirstClass or Wildcat. The San Jose Mercury
Center, which is technically an on-hne newspaper run by the San Jose Mercury News, uses a system that
was developed in conjunction with America On-Line, a major consumer information service that is owned
by the newspaper’s parent company, Knight-Ridder. FirstClass is being used by several systems, including
the Cupertino City Net, and has been chosen by PBS for use by its affiliates. While the Mercury Center
and FirstClass both have attractive point-and-click interfaces, they are terminal/host systems with hmited
support for network firnctionahty, e.g. dynamically changing or transparently interconnecting multiple

services.

All of the active community networks provide access via phone hnes and modems, and most of them
allow for character-based telnet connections through the Internet. Several systems support high-speed
digital connections to their larger members (universities, local governments, hbraries, etc.), as well as to
the Internet. One system, Minneapohs Telecommunications Network (MTN), also provides non-profit
pubhc cable TV services, but does not offer any kind of network access through the cable infrastructure.
The most unique and progressive access is offered by the Blacksburg Electronic Village, not only does it
support connections through high-speed digital phone hnes, but it can be accessed from LAN s that have

been installed in several apartment complexes.

 

CONCLUSIONS, OPPORTUNITIES, AND RECOMMENDATIONS

We have seen that individuals, families, organizations, and institutions nwd community in order to
grow and prosper, that it is a social meta-system in which other social systems flmction. Community is a
mechanism for personal, social and economic fulfillment, providing its constituents with the means for
adaption, goal-attainment, integration, and tension management (Clark 47). These functions are closely
related to the various relational systems, their associated systems of customs and values, and the primary
processes of a community: recruitment, sociahzation, and social control. The processes and components of
a community give it the form of an extensive, complex social network. The stronger the links in these
social networks, the more effectively they support community processes. Unfortunately, strengthening

many social ties using traditional means would be either very inefficient or inequitable.

Large corporations found themselves with much the same problem at the end of the industrial
revolution. They were beset by a crisis of control, in which their operations were becoming so complex
and fragmented that they had a problems monitoring, let alone optimizing, processes. Big businesses were
having trouble keeping track of their far flung ventures, finding problems and opporttmities difficult to
handle. Large, centralized information systems made large corporations more manageable. They also gave
them such structured operations that they couldn’t respond to chaotic market dynamics. Small, innovative
firms found that by being flexible, well-coordinated, and diverse they could provide customized,
inexpensive products, and steal markets right out from under corporate giants. Businesses began to
downsize their information systems, and to re-engineer their business processes. Deeper down in the org
chart, people were hooking their computers up to share printers and files. When middle management
downsizing met grassroots networking, executives reahzed they could strengthen their enterprises without
setting their processes in stone, that it could be done incrementally, and that costs could be easily assigned
to specific departments or projects. In general, businesses found that computer networking allowed them
to:

- Control costs
- Improve productivity

77

78

- Innovate

Can communities adapt computer network technologies to their social processes? Individuals seem to
benefit from interactions through community computer networks. What is not clear is whether such
systems can be attractive enough to achieve critical mass, to become valuable, persistent social
innovations. Community networks have done an excellent job of automating specific processes—looking
for hbrary books, ordering flowers, writing a letter to the mayor. But, in the sequence of events that
networked organizations go through, automation is just the first step:

1. Automating existing jobs. . .typically to boost productivity. . . [l]ittle changes but the number of
people and the capital costs of doing business.

2. Electronic infrastructure. Islands of functional automation are hnked together. . . . [but] without
change there is usually httle economic incentive to overcome the inevitable technical
incompatibihties and to battle over who does what.

3. Business-process redesign. . . .computers enable things to be done in new and more efficient
ways.

4. Business-network redesign. Creating links with suppliers and customers not only creates new
opportunities for changing business processes, it also changes the balance of competition.

5. Business-scope redesign. As part of the process of self-improvement, information technology
enables some companies to move into new businesses. (Browning 7-8)

When we place community networks in this sequence, we see that they have only gone halfway. For
networking to empower people to restructure community conflicts, to find opportunities instead of
problems, to not only work together, but work better and smarter together, it has to take them all the way
to “community-scope” redesign. It should distribute authority more evenly, take shorter paths to quality
decisions, and “informationahzation” the community. Luckily, communities are organizationally well-

suited to inforrnationization; the only thing they have to do lose is a couple of layers of bureaucracy.

The challenge for community aetworkers is to push architectural changes in their own systems even
as they are facihtating architectural changes in their communities. Enterprises have become more
efficient, effective, and equitable using networked computer systems. In order to capture similar benefits,
community networks will have to continue on the virtuous spiral of network evolution:

1. Build flexibihty and scalabihty into the network infrastructure,

2. Reduce overall costs with hardware, software, and telecommunication improvements,

3. Get more members of the community into the system,

79
4. Improve productivity through user education, and

5. Distribute control by distributing capacity and costs to the periphery of the network.

A flexible, scalable network infiastructure requires a transition from terminal/server systems and
character—based connections, to a cheat/server or peer-to-peer architecture operating over digital packet
connections. This architecture provides for better performance and rehabihty, makes the system more

widely apphcable, allows improvements to be made incrementally, and leverage investments in existing

equipment.

Information technology markets are extremely competitive. Vendors are constantly introducing new,
improved, exciting technologies. Unfortunately, many of the leading network technology firms have no
presence community networks. Community networking is currently a tiny market compared to enterprise
networking. If the network model were to be adopted and used by even a small portion of the general
population, it could turn the market inside out. Several of the computer industry’s most innovative
products are quite well suited to community networking. Dial-in and branch omce routers, groupware,
wireless communication, and commercial NOSs are all conspicuously absent from community networks.
Hardware and software manufacturers would be ill-advised to ignore community networking. Residential
and home office computer sales have outstripped commercial sales. The more computers there are in

peoples’ home, the more beneficial it will be to interconnect them—such is the nature network economy.

Utility companies, telephone companies in particular, should also pay close attention to community
networking. With telecommunications deregulation reaching into the local exchange, competitors will
soon looking for the ways to differentiate themselves. Fast, inexpensive, and easy to use digital transport
might be just the thing. The synergy between a powerful distribution system, burgeoning consumer
computer sales, the inherent richness of community networks, and interest in the Internet could make for

explosive grth in the telecommunications market.

Building membership is a perennial problem for most all non-profit organizations. For community
networking, perseverance should be the primary strategy. Computer systems are getting cheaper, more

powerful, and more useable. And the old equipment isn’t going away. Be proactive. Consider working to

80
get a computer, modem and software into the hands of anyone who has the faintest interest in community

networking. If the network model holds true just having information tools available will encourage their
use. Also, form partnerships with vendors, resellers, and service providers to get your users “turnkey”

systems for good prices.

Choose tools that are easy to use and intuitive, technologies such as electronic documents and
interactive multimedia. This may sound blatantly obvious, but there are numerous excellent system that
are used extensively in business, but do not show up in community networks. But more than anything else,
wemust make surethe membersofourcommunitiesareeducatedaboutnetworking. Abasic reahtyof
networking is that each person captures network economies in different ways, because the value of
networking comes from interaction, and each interaction is unique. No matter how powerful and

functionalthetoolsmaybe,theworkerstillhastoknowhowtousethem.

No matter how easy to use or powerful a system might be, people will still need compelhng and
engaging content to get them on-hne. There are three tried and true ways to get people’s attention:
conflict, firn and money. The first item is especially important. People love conflict and will go out of their
waytoseeafightorargument. Butmorbidfascinationisnottheprimaryreasontoputconflictonthe
network. Community aetworkers should seek out conflict within their and get it on their systems, because
that’s what networks are good for. Look for groups that disagree, contentious issues, and high-profile
problems, and put them on-line as a means of resolving and setthng things. Networks are excellent

environments for expanding perspectives, finding synergies, and redefining problems as opportunities.

Business and recreation should also become more important objectives of community networking.
Sodalacuvismisadmimbleandnwessary,hrtmkingmoneyaadplayingammommnpossiblymore
fulfilling, and definitely excellent ways to build rich relationships. Putting recreation and commerce on-
hne don’t mean simply automating them. Automated gaming and buying has been attempted without
notable success on every on-hne system that has ever been set up. As we already noted, automating
processes doesn’t even go halfway towards realizing the potential benefits of networking. The community

network must provide an innovative environment for work and play. Think telecommuting, electronic

81

commerce, collaborative engineering, virtual corporations, and a host of other business innovations.
Doing business on-hne is quick, efficient and doesn’t create hazardous byproducts. Recreational activities
could also make good use of the network, whether for multiplayer games, results and statistics, sports
images, whatever. Networking provides the greatest value when it enhances face-to-face interaction, when

it allows us to find and develop rewarding relationships.

Some may respond negatively to these last recommendations. They may fear that bringing commerce
and games to the network will lower the quahty of interactions, attract people who just want to goof ofi or
make a fast buck. Well, for better or worse, those people are your neighbors and members of your
community. Community is at least partially about social control. Getting someone into a virtual space can
be an excellent way to expand their intellectual horizons without threatening them. If someone logs-in to
play NetTrek but stays on for the discussion of clear-cutting then the network and community are better

for it.

Another important way to get people on-hne is with a compelling theme or idea. In Chattanooga,
where environmental and economic development are both important issues, members of the Community
Link project are using interest in sustainable development as an issue to get people interested in
community networking. The hnkage is not a superficial one. Knowledge is the only resource we have that

is practically inexhaustible, and it may well be impossible to overdevelop cyberspace.

lastly, we must work to make community networks more accessible. This may be repeating what has
already been said, but it is a point that bears repeating: the more accessible the system is, the more it will
get used. Think of ways to get it into peoples hves and minds. Think of putting the network into places
where it shouldn’t be. Think of it as a ubiquitous community resource. And accessibility is more than just
physical, it’s intellectual, aesthetic. Possibly the first constituents we should be getting on-hne are artists,
authors and musicians. What would happen if Chattanooga gave its arts community first shot at a blank,
interactive, digital canvas? Who knows what they would come up with. Whatever it might be, it would

probably be unique, challenging and a true synthesis of the community and network technology.

82
Community is a commodity that is all to rare in our society. While we can probably never recapture

small town or neighborhood intimacy, if it ever existed at all, it is possible to use information and
communication technologies to support recruitement, socialization and social control. Enterprises have
demonstrated to power of these technologies, whilte existing community networks have given us a ghmpse
of their potential for communities. The key to community networking is that the network must be as
unique as the community it operates within, and must indeed become an organ of the community, in order
to be successful. While enterprise networks use highly standardized technologies, they are as diverse as
the enterprises. This is because each enterprise is made up of very diflerent people, with different
strengths, weaknesses, values, and customs. There is no pat solution or simple formula for community
networking. There are a range of social, economic and pohtical issues unique to each community that
must be addressed along the virtuous spiral to a functional community network. Not the least of the issues
is the very redefinition of community, for that may be the greatest lesson of enterprise networking:
computer networking changes the way members of an organization find and use information, the way the
interact with each other and with other organizations, often radically changing the organization itself.
Networking provides its greatest benefit when it is used as a strategic resource, to manage these changes
for the greatest overall good. The challenge for communities is to develop a pervasive awareness of
innovation among their members and a wilhngness to use networks as a means of improving their
community, their hfestyle and themselves. It is only through this prediliction that a computer network

could foster social interaction and build community.

BIBLIOGRAPHY

Arensberg, Conrad Maynadier. Introducing Social Change; A Manual for Commqu Development.
Chicago: Aldine-Atherton, 1971

Barlow, John Perry. “Is There a There in Cyberspace?” The Utne Reader March-April 1994: 53-59.

Bell, Daniel. Coming of the Post-Industrial Socieg: a Venture in Social Forecasting. New York: Basic
Books, 1976.

Beniger, James R The Control Revolution: Technological and Economic Origw' of the Information
Society. Cambridge, Mass: Harvard University Press, 1986.

Bernstein, Paul. “Networking computers to manage work flow.” Trial 1994: 103-4.

Blau, Frank. “Local area networks can be difficult to operate; proceed with caution.” The National Law
Journal Dec. 1990: 28.

Bozman, Jean S. “Bank of America tests downsizing: moves host apps to PC LAN 5 across 11 states.”
Computerworld August 1992: 77.

Bozman, Jean S. “Mainframe-to-PC LAN shift taking hold.” Computerworld Dec. 1992: 4.
Brody, Herb “Computers invade the executive suite.” High Technology Business June 1988: 41-6.
Browning, John. “Information Technology,” The Economist 16 June 1990: S5-20.

Bruss, Lois R and Helene T. Roos. “Operations, Readiness, and Culture: Don’t Reengineer Without
Considering Them,” Inform April 1993: 57-64.

Buckley, Walter. Sociology and Modern Systems Theory. Englewood Cliffs, New Jersey: Prentice-Hall,
1967.

Cameron, Kim. (1980) “Critical Questions in Assessing Organizational Effectiveness,” Orgm'zational
Theory: Readings and Cases. Ed. Penny L. Wright and Stephen P. Robbins. Englewood Cliffs, New
Jersey: Prentice-Hall, 1987. 16-31.

Carmody, Deirdre. “Magazines create on-hne salons.” The Nfl York Times 20 Dec. 1993: D6.

Carroll, Daniel T. (1983) “A Disappointing Search for Excellence,” Orgam_z_a' tional Theog: Readings and
Cases. Ed. Penny L. Wright and Stephen P. Robbins. Englewood Chffs, New Jersey: Prentice-Hall,
1987. 245-54.

Crandal, Robert W. and Kenneth Flamm, ed. Changrn' g the Rules. Washington: The Brookings Institute,
1989.

83

84
Cirese, Sarah guest: A Search for Self. New York: Holt, Rinehardt and Winston, 1985.

Clark, Terry N. Comme structure and decision-making: commtive analysis. San Francisco:
Chandle,. 1968.

Contractor, Noshir S. and Eric M. Eisenberg. “Communication Networks and New Media in
Organizations,” Orga_m'zations and Communication Technology. Ed. Charles Steinfield and Janet
Fulk. Newbury Park, California: Sage Pubhcations, 1990. 143-72.

Dance, Frank E. X. “Towards a Theory of Human Communication,” Human Communication Theog:
Origig Essay. Ed. Frank E. X. Dance. New York, Chicago, San Francisco, Toronto, London: Holt,
Rinehart and Winston, Inc., 1967.

Davis, Stan and Bill Davidson. “You Can’t Stay 39 Forever,” Modern Ofl'rce Technology July 1991: 12-4.

Delcomyn, W. E. “Connecting islands in a sea of data.” Electrical World April 1990: 37-41.

Drucker, Peter F. “The New Society of Organizations,” Harvard Business Review Sept-Oct. 1992: 95-
104.

Elkin, Stephen L. City and Regime in the American Mubhc. Chicago: University of Chicago Press.
1987.

F arace, Richard V., Peter R Monge and Hamish M. Russell. Communicag g and Orgagizing. Reading,
Massachusetts: Addison-Wesley. 1977.

Fischer, Claude S., et al. Networks and Places, Social Relations in the Urban Setting. New York:
Macmillan Publishing. 1977.

Fitzgerald, Michael. “Tea maker has new network in the bag.” Commterworld 26 August 1991: 41-3.

Flyml, Laurie. “The Year of the LAN arrived and it was nothing hke I expected.” InfoWorld 12 August
1991: 41.

Freeman, Christopher. “Technical Change and Productivity,” Finance & Develgpment Sept. 1989: 46-9.

Frude, Neil. “InformationTechnology in the Home: Promises as yet Unreahzed,” Desigm'gg for the future:
Information technologies and Pe_ople. Ed. Frank Blackler and David Osborne. The British
Psychological Society, 1987.

Galbraith, Jay R “Organization Design: An Information Processing View,” Organg’ tional Theosy:
Readings and Cases. Ed. Penny L. Wright and Stephen P. Robbins. Englewood Cliffs, New Jersey:
Prentice-Hall, 1987. 179-89.

Gantz, John. “Dinosaurs should beware: the future is in downsizing.” InfoWorld 23 Sept 1991: 101.

Goist, Park Dixon. From Main Steet to State Street. Port Washington, NY, London: Kennikat Press Corp.,
1977.

Guerin-Calvert, Maragret E. and Steven S. Wildman, ed Electronic Services Networks: A Business and
Public Pohgy Challenge. New York: Praeger Pubhshers, 1991.

Guthrie, K. Kendall. Communig Information Smems: Lessons for a Redefinition of Universal Service.
working paper: 1990-91 Universal Service Project, University of Texas at Austin, 1991.

85

Hall, D. M. The Management of Human Systems. Cleveland, Ohio: Association for Systems Management,
1971.

Hall, Richard H. Orgfl'zations: Structures and Process. Englewood Chfl‘s, New Jersey: Prentice-Hall,
1972.

Huber, George P. “A Theory of the Efl‘ects of Advanced Information Technologies on Organizational
Design, Intelligence, and Decision Making,” Orgm' tions and Communication Technology. Ed.
Charles Steinfield and Janet Fulk. Newbury Park, California: Sage Pubhcations, 1990. 237-74.

Hughes, David “Industry rethinking computer strategy.” Aviation Week & Smce Technology 4 March
14, 1994: 92-5.

Johnson, Kirk. “Easing a hospital's suburban trek; St Joseph's uses computers to expand into land of
malls.” 13 New flak Times 23 July 1994): 21.

Johnson, William R, Jr. “Anything, Anytime, Anywhere: The Future of Networking,” Technology 2001.
Ed. Dereck Leebaert. Cambridge, Massachusetts: The MIT Press. 1991. 150-75.

Jones, G. Hunter. “Networking: it works at large firms; LANs for PCs.” The National Law Journal 20
June 1988): 21.

Keen, Peter G. W. Comggng in Time: Using Telecommunications for Comgitive Advantage Ballinger,
1988.

Kustoff, Marc. “Is it time to network your PCs?” Personnel Journal July 1990: 46-54.
LaPlante, Ahce. “LANs, mainframes speak same language.” InfoWorld 13 April 1992: S60.

Latamore, G. Berton “Why everyone works smarter when they're connected.” Working Woman Oct. 1988:
72-77 .

Lauristoa, Robert. “Wireless LANs; Your Next Network?” PC World Sept 1992: 225-32.

Levy, Michael S. “Computer system for small court; LAN for jurists.” The Natignal Law Journal 26 June
1989: 19.

Lias, Edward J. Future Mind: The Microcompgter—New Medim, New Mental Envirnoment. Boston,
Toronto: Little, Brown and Company, 1982.

Lindquist, Christopher. “County thinks small, dumps 4381 for LANs.” Computerworld 6 April 1992: 1-3.
Little, Donald J. “Linking PCs in firms of all sizes.” The National Law Journal 21 Nov. 1988: 9.

Littlejohn, Stephen W. Theories of Hm ng’cation Belmont, Cahfornia: Wadsworth Pubhshing
Company, 1989.

Machlup, Fritz. The Production and Dist_ribution of Knowledge in Qe United States. Princeton: Princeton
University Press, 1962.

Machrone, Bill. “Can LANs replace minis?” PC Magm’ e 25 April 1989: 65-67.

Morley, David, Stuart Proudfoot, and Thomas Burns, eds. Making Cities Work. Boulder, CO: Westview
Press, 1980.

86

Manning, Peter K. Orgm’ tional Communication. New York: Aldine De Gruyter, 1992.

Markus, Lynne M. “Toward a ‘Critical Mass’ Theory of Interactive Media,” Orgm' tions and
Communication Technology. Ed. Charles Steinfield and Janet Fulk. Newbury Park, California: Sage
Pubhcations, 1990. 194-218.

Maslow, A. H. Toward a Psychology of Being. New York: Van Nostrand Reinhold, 1968.

McLuhan, Marshall. Understanding media: the extensions of man. New York: McGraw-Hill, 1964.

Miley, Michael. “Putting the GROUP into groupware.” InfoWorld 14 Oct . 1991: 53-56.

Morely, David, Stuart Proudfoot, and Thomas Burns. “Urban Innovation: an Introduction.” Making Cities
Work: The wages of Urban Innovation. Westview Press: New York, 1980.

Morgan, Hugh. “Doing it in Dayton.” The gall May 1994: 29-34.

Musgrave, Bill. “Stacking up LANs.” Datamation 15 Feb 1988: 56-64.

Naisbitt, John and Patricia Aburdene. Re-inventing the Comgration. New York: Warner Books, 1985.
Nash, Kim S. “Law firm migrates from Wang.” Commterworld 16 Nov 1992: 53.

“New cost-to-use analysis key to evaluating networked enterprises.” Computerworld 25 Oct. 1993: S1-3.
Newman, Denis. “School networks: dehvery or access.” Communications of the ACM May 1993: 49-52.

Nordheimer, Jon “Computers and telephones pave a new path to a college degree.” The New York Times
21 Feb. 1994: B1.

Nykamp, Susan R “LANs help Allen-Bradley beat the clock” Computerworld 1 April 1991: 71.

O'Malley, Kathleen. “LAN helps K mart loosen support lolam.” PC Week 17 April 1989: 77-79.

Perrow, Charles. (1974) “Is Business Really Changing?” Orgasm 'onal Theosy: Readings ad Cases. Ed.
Penny L. Wright and Stephen P. Robbins. Englewood Chfls, New Jersey: Prentice-Hall, 1987. 277-
90.

Peters, Thomas J. (1980) “Putting Excellence into Management,” Orgsmg’ tional Theog: Readings and
Cases. Ed. Penny L. Wright and Stephen P. Robbins. Englewood Cliffs, New Jersey: Prentice-Hall,
1987. 237-44.

Pinella, Paul. “Organizational computing arrives.” Datamation 15 Nov 1990: 42-47.

Quinn, Robert E. and Richard H. Hall. (1983) “Organizational Forms,” Organizational Theory: Readings
and Cases. Ed. Penny L. Wright and Stephen P. Robbins. Englewood Chffs, New Jersey: Prentice-
Hall. 1987.

Quittner, Josh. “Johnny Manhattan Meets the FurryMuckers.” m March 1994: 92-97, 138.

Radosevich, Lynda. “Searching for the perfect LAN.” W 20 Dec 1993: 57-59.

Rheingold, Howard. “The Virtual Community.” The Utne Reader March-April 1994: 60-64.

87

Rice, Ronald E. (1987) “Computer-Mediated Communication and Organizational Innovation,” Journal of
Communication Autumn 1987: 65-94.

Rogers, Everett M. and Rekha Agarwala-Rogers. Communication in Organizations. New York: The Free
Press, 1976.

Rogers, Everett M. Communication Networks : Toward a New Paradigm for Research. New York: Free
Press; London: Collier Macmillan, 1981.

Sanders, Irwin T. The Communig. 3rd ed. New York: Ronald Press, 1975.

Schaffer, Carolyn and Kristin Anundsen. Creating Commqu Anmhere: Finding Sumrt and
Connection in a F ragemented World. New York: Jeremy P. Thatcher/Perigee Books, 1993.

Schlack, Mark. “Enterprise networks: one step at a time.” Datamation 15 April 1991: 60-62.

Schwartz, Evan I. “The dinosaur that cost Merrill Lynch a million a year.” Business Week 26 Nov. 1990:
122.

Scott, Richard W. Orgm'zations: Rational, Natural and ngn Systems. Prentice Hall, 1981.
Sedacca, Boris. “Lloyd's ups the ante for claims processing.” Datamation 1 June 1990: 115-6.

Sehg, Franz F. and Jack T. Nipper. “Strategic Impact of Information Technology,” Industrial
Management and Data Systems June 1990: 3-4.

Seymour, Jim. “The “D“ word.” PC Magg’ne 31 Jan 1989: 77-79.

Shockley-Zalabak, Pamela. Fundamentals of Orggljzational Communicaton: Knowledge, Sensitivig,
Skills Values. New York: Longman, 1991.

 

Spencer, Cheryl. “RS-232 LANs; inexpensive omce connections.” Personal Compgting. June 1987, VII,
p139(7).

Sproull, Lee and Sara Kiesler. Connections, New WM of Working in Networked Orgfl'zations.
Tanenbaum, Andrew S. Comptuter Networks. Englewood Cliffs, New Jersy: Prentice-Hall. 1989.

Stuart, Reginald. “High-Tech Redlining.” The Utne Reader March-April 1994: 73.

Thayer, Lee. Communication and Communication SMms. Homewood, Ilhnois: Richard D. Irwin, 1968.
Vincent, Joe. “Junk my mainfiame? No way!” Computerworlg 30 May 1994: 93-5.

Walford, Robert B. Information Smms and Business mes. Reading, Massachusetts: Addison-
Wesley, 1990.

Wright, Robert G. The Nature of Organizations. Encino, California: Dickenson Pubhshing Company,
1977.

Zmud, Robert W. “Opportunities for Strategic Information Manipulation Through New Information
Technology,” Orgfl'zations and Communication Technology. Ed. Charles Steinfield and Janet Fulk.
Newbury Park, California: Sage Pubhcations, 1990. 95-116.

   

"‘llillllililillil“