First Monday

The Premise and the Promise of a Global Information Infrastructure


Converging Tasks and Technologies
Modes of Communication
Task Independence and Task Dependence
Technology Adoptation and Adaptation
Organizational Adaptation
Creating a Global Information Infrastructure
What Is Infrastructure?
Infrastructure as Public Policy
Information Infrastructure as a Technical Framework
Information Infrastructure as Technology, People, and Content




Let us build a global community in which the people of neighboring countries view each other not as potential enemies, but as potential partners, as members of the same family in the vast, increasingly interconnected human family.
— Vice–President Al Gore (1994a)
The information society has the potential to improve the quality of life of Europe’s citizens, the efficiency of our social and economic organization and to reinforce cohesion.
— Bangemann Report (1994)

The premise of a global information infrastructure is that governments, businesses, communities, and individuals can cooperate to link the world’s telecommunication and computer networks together into a vast constellation capable of carrying digital and analog signals in support of every conceivable information and communication application. The promise is that this constellation of networks will promote an information society that benefits all: peace, friendship, and cooperation through improved interpersonal communications; empowerment through access to information for education, business, and social good; more productive labor through technology–enriched work environments; and stronger economies through open competition in global markets.

The promise is exciting and the premise appears rational. Information technologies are advancing at a rapid pace and becoming ever more ubiquitous. Many scholars, policy makers, technologists, business people, and pundits contend that changes wrought by these new technologies are revolutionary and will result in profound transformations of society. Physical location will cease to matter. More and more human activities in working, learning, conducting commerce, and communicating will take place via information technologies. Online access to information resources will provide a depth and breadth of resources never before possible. Most print publication will cease; electronic publication and distribution will become the norm. Libraries, archives, museums, publishers, bookstores, schools, universities, and other institutions that rely on artifacts in physical form will be transformed radically or will cease to exist. Fundamental changes are predicted in the relationships between these institutions, with authors less dependent on publishers, information seekers less dependent on libraries, and universities less dependent on traditional models of publication to evaluate scholarship. Networks will grease the wheels of commerce, improve education, increase the amount of interpersonal communication, provide unprecedented access to information resources and to human expertise, and lead to greater economic equity.

In contrast, others argue that we are in the process of evolutionary, not revolutionary, social change toward an information-oriented society. People make social choices which lead to the development of desired technologies. Computer networks are continuations of earlier communication technologies such as the telegraph and telephone, radio and television, and similar devices that rely on networked infrastructures. All are dependent on institutions, and these evolve much more slowly than do technologies. Digital and digitized media are extensions of earlier media, and the institutions that manage them will adapt them to their practices as they have adapted many media before them. Electronic publishing will become ever more important, but only for certain materials that serve certain purposes. Print publishing will co–exist with other forms of distribution. Although relationships between institutions will evolve, publishers, libraries, and universities serve gatekeeping functions that will continue to be essential in the future. More activities will be conducted online, with the result that face–to–face relationships will become ever more valued and precious. Telecommuting, distance–independent learning, and electronic commerce will supplement, but not supplant, physical workplaces, classrooms, and shopping malls. Communication technologies often increase, rather than decrease, inequities, and we should be wary of the economic promises of a global information infrastructure.

Which of these scenarios is more likely to occur? Proponents of each offer historical precedent and argue rationally for their cases. Many other scenarios exist, some between those presented above and some at the far ends of the spectrum. The extremes include science–fiction–like scenarios in which technology controls all aspects of daily life, resulting in a police state where every activity is monitored, and survivalist scenarios in which some catastrophe destroys all technology, with the result that new societies are reinvented without it. The science fiction and survivalist scenarios are easily discounted because checks and balances are in place to prevent them. Choosing between the revolutionary, discontinuity scenario and the evolutionary, continuity scenario described above is more problematic. Each has merit and each is the subject of scholarly inquiry and informed public debate.

In view of the undisputed magnitude of some of these developments, it is reasonable to speak of a new world emerging. It is not reasonable, however, to conclude that these changes are absolute, that they will affect all people equally, or that no prior practices or institutions will carry over to a new world. Nor is it reasonable to assume that any individual institutions, whether libraries, archives, museums, universities, schools, governments, or businesses, will survive unscathed and unchanged into the next millennium. Strong claims in either direction are dangerous and misleading, as well as lacking in intellectual rigor. The arguments for these scenarios, the underlying assumptions, and the evidence offered must be examined. Upon close examination, it will often be found that strong claims about the effects of information technologies on society, and vice versa, are based on simplistic assumptions about technology, behavior, organizations, and economics. None of these factors exists in a vacuum; they interact in complex and often unpredictable ways.

I argue throughout this book that the most likely future scenario lies somewhere between the discontinuity and continuity scenarios. Information technology makes possible all sorts of new activities and new ways of doing old activities. But people do not discard all their old habits and practices with the advent of each new technology. Nor are new technologies created without some expectations of how they will be employed. The probable scenario is neither revolution nor evolution, but co–evolution of information technology, human behavior, and organizations. People select and implement technologies that are available and that suit their practices and goals. As they use them, they adapt them to suit their needs, often in ways not anticipated by their designers. Designers develop new technologies on the basis of technological advances, marketing data, available standards, human factors studies, and educated guesses about what will sell. Products evolve in parallel with the uses for which they are employed. To use a simplistic aphorism: Technology pushes, while demand pulls.


Figure 1


The central concern of this book is access to information in a networked world. Information access is among the primary arguments for constructing a global information infrastructure. Information resources are essential for all manner of human affairs, including commerce, education, research, participatory democracy, government policy, and leisure activities. Access to information for all these purposes is at the center of the discontinuity–continuity debates. Some argue that computer networks, digital libraries, electronic publishing, and similar developments will lead to radically different models of information access. The technologies of creation, distribution, and preservation will undergo dramatic transformation, as will information institutions such as libraries, archives, museums, schools, and universities. Relationships among these and other stakeholders, including authors, readers, users, and publishers, will evolve as well. Others argue that stakeholders, relationships, and practices are so firmly entrenched that structural changes will be slow and incremental because most new technologies are variations on those that came before. My view is that some degree of truth exists in each of these statements. These and other arguments are examined throughout the book.

Much has been written about technology, human behavior, and policy regarding access to information. Most of the writing, however, focuses on one of these three aspects with little attention to the other two. In this book I endeavor to bring all three together, drawing on themes, theories, results, and practices from multiple disciplines and perspectives to illustrate the complex challenges that we face in creating a global information infrastructure. Technical issues in digital libraries and information retrieval systems are addressed, but not in the depth provided in recent books by Lesk (1997) and Korfhage (1997). Nor are design issues addressed to the degree covered by Winograd, et al. (1996). Information–related behavior in electronic environments is covered, but in less depth than in Marchionini (1995). Institutional and organizational issues are treated more fully in Bishop and Star (1996), Bowker, et al. (1996), and Sproull and Kiesler (1991). Policy issues of the Internet are addressed in more depth in Branscomb and Kahin (1995), Kahin and Abbate (1995), and Kahin and Keller (1995). In this book I draw on these and many other resources to weave a rich discussion of access to information in a networked world. In view of the early stages of these developments, more questions are raised than yet can be answered. My hope is to provoke informed discussion between the many interested parties around the world.



Converging Tasks and Technologies

People use computer networks for a vast array of activities, such as communicating with other individuals and groups, performing tasks requiring remote resources, exchanging resources, and entertainment (whether with interactive games or passive media such as videos). Among the few common threads in predictions of future technology (see, e.g., “The Next 50 Years”, 1997 and Pontin, 1998) is that we will see more convergence of information and communication technologies, blurring the lines between tasks and activities and between work and play. We will have “ubiquitous computing” (Pontin, 1998) and “pervasive information systems” (Birnbaum, 1997). We will become “intimate with our technology” (Hillis, 1997), and “information overload” (Berghel, 1997a) will be more of a problem than ever.

An underlying theme of such predictions is “digital convergence,” indicating that more and more information products will be created in digital form or will be digitized, allowing applications to be blended more easily. Digital technologies will co–exist with analog and other forms of information technologies yet to be invented. Analog technology is based on continuous flows, rather than the discrete bits of digital technology. Computer and communication networks are an example of the bridge between these technologies. The word “modem” was coined from “modulate” and “demodulate,” which describe the device’s function in converting digital data produced by computers into analog signals that could be sent over telephone lines designed for voice communication and vice versa. Predictions of ubiquitous computing are based on an increasing reliance on small communication devices and embedded systems such as those that control heating and lighting in homes and offices. Future computer networks are expected to link these devices just as they now link personal computers, data storage, printers, and other peripherals (Pontin, 1998).



Modes of Communication

No matter what technologies gird the framework of the global information infrastructure, human activities involving the network will be intertwined. As the editors of Wired [1] put it,

... broader and deeper new interfaces for electronic media are being born. ... What they share are ways to move seamlessly between media you steer (interactive) and media that steer you (passive). ... These new interfaces work with existing media, such as TV, yet they also work on hyper–linked text. But most important, they work on the emerging universe of networked media that are spreading across the telecosm.

Despite the hyperbole, this quotation highlights a useful distinction between “pull” technology (which requires explicit action by the user) and “push” technology (which comes to the user without the user’s explicit action). Some activities are easily categorized by this dichotomy, but others have characteristics of each. Composing and sending an e–mail message and searching a database require explicit “pull” actions, for example. Although both the broadcast mass media and the emerging media services that deliver tailored selections of content to workstations during idle time can be classified as push technologies (editors of Wired, 1997), the latter form also could be considered “pull,” because the user presumably took action to subscribe to the service. Similarly, if composing and sending e–mail is pull technology, then receiving mail can be viewed as a form of “push.” Opening and reading messages requires explicit actions, but users can decide what to read, delete, or ignore. They also can sort desirable and undesirable messages by means of automatic filters. Because subscribing to desirable content and filtering out undesirable content require parallel actions, both can be viewed as forms of push technology if one accepts the Wired definitions of “push” and “pull.”

Push and pull combine in other ways as well. People subscribe to distribution lists, which then send messages at regular or irregular intervals. They also subscribe to services that alert them when new resources are posted on a specific network site, but they must take explicit action to view or retrieve the resources from that site.

Truly interactive forms of communication are difficult to categorize as push or pull. People engage in conversations in “chat rooms,” play roles in MUDS and MOOS, and hold conferences, meetings, and classes online in real time. All require explicit actions, but the characteristics of these two–way or multi–way conversations are far richer than the solo–action pull of searching a database or sending a message. Some of these are the “demassified” communication technologies that Rogers (1986) predicted more, tailored to individual users and to small audiences. However, the “push” technologies of customized desktop news delivery touted by Wired in 1997, in which messages continually scroll across the subscriber’s screen, have yet to become the commercial success that was predicted. Perhaps they were not sufficiently customized or “demassified.” Perhaps people found them too disruptive, preferring “pull” modes in which they could acquire the desired content at their convenience.

The intertwining of communication modes in electronic environments adds new dimensions to information access. Although more study has been devoted to “active” than to “passive” information seeking, even these categories are problematic in this new environment. These are but a few of many communication definitions and concepts being reconsidered in the light of new information technologies.



Task Independence and Task Dependence

The more intertwined tasks and activities become, the more difficult it becomes to isolate any one task for study. In the past, most theory and research presumed that the human activities involved in access to information could be isolated sufficiently to be studied independently. This is particularly true of information–seeking behavior, a process often viewed as beginning when a person recognizes the need for information and ending when the person acquires some information resources that address the need. Such a narrow view of the process of seeking information simplifies the conduct of research. For example, information seekers’ activities can be studied from the time they log onto an information retrieval system until they log off with results in hand. The process can be continued further by following subsequent activities to determine which resources discovered online were used, how, and for what purposes. Another approach is to constrain the scope of study to library–based information seeking. People can be interviewed when they first enter a library building to identify their needs as they understood them at that time. Researchers can follow users around the building (with permission, of course), and can interview the users again before departure to determine what they learned or accomplished.

Narrowly bounded studies such as these provide insights into detailed activities and are useful for evaluating specific systems, services, and buildings. However, their value and validity are declining for the purposes of studying the information environment of today and assessing the needs of the future. In the early days of information retrieval, people might reasonably conduct most or all of their searching on one retrieval system. Only a few systems existed, and each had a limited number of databases. These were complex systems requiring lengthy training. Information seekers, often with the assistance of skilled searchers, would devote considerable effort to constructing, executing, and iterating a search on a single system (Borgman, et al., 1984). A close analysis of user–system interaction could provide a rich record of negotiating a single search query. Even so, such studies provide little insight into the circumstances from which the information need arose or into the relationship between a particular system and the use of other information resources.

In today’s environment, most people have access to a vast array of online resources via the Internet and online resources provided by libraries, archives, universities, businesses, and other organizations with which they are affiliated, as well as print and other hard–copy resources. They are much less dependent on any single system or database. Rather, they are grazing through a vast array of resources, perhaps “berry picking” (Bates, 1989) from multiple sources and systems. Studying any individual system is far less likely to provide a comprehensive view of information–seeking activities than it was in the past. Similarly, people have fewer reasons to spend time in library buildings, now that they can use many library resources from the convenience of home, office, dorm, coffee shop, or anywhere else with network access. And they can do so at hours of day or night when library buildings normally are closed. Thus, time spent in the library building may be for narrower and more specific purposes, and may occur only at critical stages in the search process. The use of library buildings also reflects patterns that are influenced by age, generation, culture, discipline of study, and many other factors. Such research should yield insights into the design of future buildings and services, provided it is set in a larger context of overall information–use patterns.

Future research on access to information must consider the complex relationships between information–related activities and the context of work and leisure practices in which these activities are conducted. Although all scholarship is constrained by the necessity of studying that which can be studied, particular caution is necessary when studying tasks that tend to be interdependent.



Technology Adoption and Adaptation

Underlying the design of any information technology are assumptions about how and why people will use it. The assumptions are sometimes explicit and sometimes only implicit, whether for individual communication devices, for information systems, or for the design of a global information infrastructure. In identifying design criteria, and making implicit assumptions explicit, many methods and perspectives can be applied. We can evaluate which prior technologies were adopted and which were not, the processes by which they were adopted, how similar technologies are used, what features and functions are most popular and most effective, and how their users adapt them to new purposes.

I will highlight three perspectives on assessing how and why people use information technologies. Though many other perspectives and methods exist, these three are applicable to our concerns for access to information.


Of the vast number of information technologies that are invented, only a few make it to the marketplace, and of these, even fewer are successful. The quality of the product is only one determinant of market success. Many products that receive critical acclaim fail to garner large market shares. The Beta video recording technology and the Macintosh computer are the best–known examples. In contrast, many products whose reviews range from skepticism to scorn achieve great market success. Business factors such as timing, marketing, and pricing are determinants of success. Other determinants are social factors involving how and why people choose to adopt any particular innovation. Rogers (1983, 1986) summarizes the results of a large number of adoption studies using a five-stage model. The first stage of adoption is knowledge, or becoming aware of the existence of a new technology that might be useful. This stage is influenced by factors such as previous practices, felt needs or problems, tendencies toward being innovative, and norms of the individual’s social system. The second stage is persuasion, which in turn is influenced by the perceived characteristics of the innovation, how well it might work, how easy it is to try, and how easily the outcome can be observed. In the third stage, the adopter makes a tentative decision to accept or to reject the technology. Acceptance may lead to implementation (fourth stage) and, if the innovation is deemed sufficiently useful, to a confirmation to continue its use (fifth stage). If the innovation is rejected, the individual still may revisit the decision and adopt it later.

Electronic mail (e-mail) provides an instructive example of the adoption process. A person may first become aware of its existence through news reports or through discussions with friends, family, or co–workers. Someone surrounded by e–mail users will hear about it more quickly and frequently than someone whose acquaintances are non–users. Even today, elderly Americans who have minimal contact with computer users may have at most a vague idea of what e–mail is, for example. In countries with minimal telecommunications and computing penetration, only the elite may be aware of e–mail as a potentially useful technology. In the persuasion stage, a person who has many potential e–mail correspondents will find the technology more attractive than a person who knows no one else with an e–mail address. Similarly, a person who already owns a computer with a modem will find it far easier to try e–mail than one who must acquire the technology and the skills to use it. Once they have tried it, some people will find e–mail sufficiently useful, affordable, and worth the time and effort to continue using it. Others will not. Thus, once people become aware of e–mail, only some will consider trying it, a smaller number will make the effort to try it; of these, only some will acquire it and continuing using it, and they may abandon it later. Conversely, some who rejected e–mail at any of these adoption stages may consider it again at some later time.

This adoption pattern also operates in the aggregate. The “early adopters” typically are risk takers who are willing to try unproven techniques, often at great expense. If they adopt the new technology, their successes may convince more risk–averse individuals to try it. Conversely, if the early adopters reject it, others may be more reluctant to try it. By the time the low–risk late adopters decide to implement a technology, the early adopters may have moved on to something yet newer and more innovative. Some technologies reach a critical mass of adoption in a short period of time and are great market successes. Others are unable to find a match with early adopters fast enough, and the entrepreneurs fail before finding their niche in the market. Others fail because they do not fill a perceived need. Yet others succeed because they are good enough, cheap enough, and at the right place at the right time, although not necessarily an optimal design. Though this explanation is a gross simplification of the adoption process, it illustrates a few of the many social variables that influence the success of new information technologies.

Again, e–mail provides a useful case example. E–mail filled a perceived need early in the development of computer networks and reached a critical mass of computer users fairly quickly. Spreadsheets were a similarly attractive technology that contributed to the adoption of personal computers. Early adopters of both technologies were sophisticated computer users who tolerated complex user interfaces, often unreliable software, and minimal functionality because the technology was sufficiently valuable for their purposes. People who are early adopters of one technology tend to be early adopters of others, willing to tolerate immature technologies in return for their benefits, and often enjoy the challenge of working at the “bleeding edge” of technical frontiers.

Conversely, late adopters of one technology tend to be late adopters of others. These people are far less likely to appreciate technology for its own sake, preferring mature, easy–to–use technologies with a high perceived payoff relative to the effort required in learning to use them. They are happy to let others “work the bugs out” before spending the time, effort, and money to adopt them. This distinction between the personality characteristics and social context of early and late adopters is an important one to bear in mind when considering technologies intended for a mass market. If a global information infrastructure is to achieve wide acceptance, it must be attractive to late adopters.


Theories of diffusion and adoption are valuable in understanding the social processes involved in choosing to employ a particular technology. The “diffusion of innovations” theory originated in rural sociology to explain farmers’ choices of agricultural innovations such as farming equipment, hybrid plants, pesticides, and techniques for planting, harvesting, and storing crops. The theory was later extended to study the adoption of a diverse array of innovations including solar energy during a fossil–fuels shortage and family planning methods in developing countries. One weakness of applying the “diffusion of innovations” theory to information technologies is the implicit assumption that the innovation is relatively static. Information technologies tend to be more dynamic and flexible than farming equipment, for example. Any communication device may be short–lived, making it difficult to compare the actions of someone who adopted the first crude implementation to those of someone who adopted a more sophisticated and less expensive version only months later. Moreover, information technologies are more malleable and adaptable to individual purposes than are most other technologies. Thus, we must look not just at the adoption of information technologies as a binary (adopt/not adopt) decision, but also at how technologies, once adopted, are adapted over time.

Books provide an early example of how people adapt information technologies to their purposes. Manuscripts (meaning, literally, hand–written) were the first form of written record. Manuscripts on sheepskin or parchment were easier to create and read than chiseled stone tablets, but still could be read only by one person in one place at a time. Manuscripts could be loaned for manual copying, which enabled duplication, however laborious. Gutenberg’s improvements in movable type in the fifteenth century made multiple copies economically feasible for the first time. Early printed books retained the shape and size of manuscripts, following the earlier technology. Although the distribution of multiple copies meant that more people could own and read a work concurrently, books still were too bulky for portable use, except by the very rich. Greenberg (1998) recounts the oft–told story of Abdul Kassem Ismael, who was said to have had a library of 117,000 books in tenth–century Persia. Not only did he carry his library with him while he traveled, on the backs of 400 camels, he trained the camels to walk in alphabetical order. Later innovations led to publishing books in more portable sizes that fit not only in the saddlebags of yesteryear, but in the backpacks and briefcases of today.

We find similar adaptations in the use of computer networks. The ARPANET, precursor to the Internet, was created for remote access to scarce computing resources. Electronic mail was a feature intended to serve as an ancillary communication function. E–mail proved so useful for general communication that it became the dominant use of the network, much to the surprise of the ARPANET’s designers (Licklider and Vezza, 1978; Quarterman, 1990). E–mail was the “killer application” that attracted most people to the Internet (Anderson, et al., 1995; Quarterman, 1990), and it remains the most important reason for becoming an Internet user (Katz and Aspden, 1997).

E–mail is a far different application today than it was in the early days of the ARPANET, however. Early e–mail consisted of very short plain text messages. Less than a decade ago, messages could take several days to arrive, with delays caused whenever a server in a store–and–forward network went down. E–mail was neither fast enough, reliable enough, nor functional enough to replace most other forms of communication. The technology advanced, as did users’ perceived needs for more capabilities and better services. Today’s e–mail supports long messages of formatted text and is fast, reliable, convenient, and inexpensive (Berghel, 1997b). Increasingly, e–mail software allows people to send and receive file attachments that preserve the integrity of text, images, graphics, and sound. For many purposes, e–mail is a suitable substitute for telephone, fax, post, or express mail.

E–mail now combines the features of word processors, file transfer (ftp), and multimedia file management. It also provides a bridge to the World Wide Web by embedding live links to Web sites. By including a URL (uniform resource locator) address in an e–mail message, a user can click on an address to launch a browser application and link to the Web site. And the reverse is true. Once at the Web site, a user can click on “e–mail” and send a message to the Web site.

E–mail has evolved from a simple application to one that combines a rich array of services. As users realized its value and its constraints, they identified further improvements that could be made. Yet today’s complex e–mail technology has too much functionality to be feasible for some purposes. Thus, we also find evidence of complex applications being stripped down to the bare elements that suit newly identified needs. An example is the convergence of e–mail with pocket pagers, which themselves were initially a simple, single–function technology. Some of today’s more elaborate pagers include a full, albeit tiny, QWERTY keyboard and alphanumeric display, on which people can send and receive terse messages. Other pagers include function keys for common responses to e-mail-type messages: yes, no, time, date, etc. Such devices can convey cryptic but critical messages, such as “When do you arrive?” (answer: "AA 75, 8:44pm LAX"), “Did we win the case?,” “Running late, resched Tu at 3?” (answer: “no. Tu 2pm ok?”), “pls get milk,” or “get KT @ school.”

These are but a few examples of how people adapt information technologies by using them. People sometimes adopt only part of a technology, as illustrated by the example of stripped–down e–mail. Other times they disable or circumvent features of a technology. E–mail file attachments are a case in point. They are extremely useful for exchanging files quickly between team members, co–authors, authors and editors, authors or publishers and readers, or teachers and students. But they are useful only when they work. When exchange partners have identical hardware and software platforms, fast connections, and (better yet) the ability to scan for viruses before receipt, file exchange may be seamless.


Figure 2


System designers, along with as those who send file attachments, often are unaware of the difficulties involved in receiving attachments intact and in a usable form, however. Despite considerable progress, the necessary platform independence and software independence required for reliable exchange of attachments over networks has yet to be achieved. File exchanges between different platforms (e.g., PC and Macintosh) and different operating systems (Windows 95, Windows 98, Windows NT, Macintosh OS 7.5, Macintosh OS 8.0, Unix, etc.) introduce compatibility problems. Files created with widely used word processing software such as Microsoft Word and Corel WordPerfect often fail to transfer intact. Text may transfer but formatting may be corrupted, and the likelihood of accurate transfer decreases with the inclusion of software–specific features such as tables, graphics, and macros. The more recent the version of the software used to create a file, the less likely that earlier versions of the same software or of competing software can open it intact. Exchanging files of graphics or sound is yet more problematic. Adding another layer of concern is the ability of attachments to carry computer viruses that can contaminate the receiver’s computer.

Unsolicited file attachments containing job applications, advertisements, jokes, cartoons, greeting cards, and myriad other materials clog network and modem lines and fill disk space. Owing to problems with technical compatibility, viruses, and bandwidth, many people are making minimal use of file attachments, and some are setting their e–mail parameters to reject them entirely. Local network managers are introducing delays in e–mail delivery to scan all attachments for viruses, adding another layer of complexity. Sending faxes, or mailing paper and disks, can be faster, more reliable, and less labor intensive.

The e–mail examples offer several lessons in the adoption and adaptation of information technologies. One lesson is that early adopters are willing to use an immature technology. As they use it, they will identify problems, recognize new possibilities, and demand improvements. Later adopters will identify yet more problems and more desirable capabilities as they integrate it into their practices, refining the technology further. Another lesson is that one simple technology may spawn so many features that it subdivides into component parts, as e–mail has done. We also see that advanced features that are extremely useful in some situations may result in unintended and undesirable consequences in others, as is the present case with file attachments. When people have positive experiences with a technology, they often are more inclined to adopt another technology. Conversely, when they have negative experiences, they trust the technology less than before, and are less inclined to try something new. All these lessons argue for the importance of studying the use of information technologies in actual working situations. Though laboratory experiments are extremely valuable for improving technologies under ideal conditions, field studies are essential to determine how technologies are adopted and adapted.



Organizational Adaptation

Though some technology adoption and adaptation is attributable to individual choices by individual users, much of it takes place in the context of organizations. Organizations such as businesses, governments, universities, and schools make decisions about what hardware, software, and services to purchase for use by their constituencies. Individuals may have little choice in which computing platform, Internet provider, or services they use. Organizations usually set policies about how services such as e–mail and information resources are used. Even in view of these constraints, individuals often have considerable latitude in how they employ these technologies in their work practices, however.

Sproull and Kiesler (1991) explain the unpredictable effects of introducing technology into organizations from a “two–level perspective.” They argue that most inventors and early adopters of technology think primarily about efficiency of the technology. System designers, as well as early adopters, focus on the instrumental uses to which the technology is put, whether reducing “telephone tag” through the use of electronic mail or lowering secretarial costs by replacing typing with word processing. These are the “first–level effects” of a technology.

Users rarely implement a new technology in precisely the way that designers intend, however. Organizations find it difficult to determine accurate estimates of direct costs, much less to determine the first–level effects of technology on work practices, productivity, or profits. Because technologies interact with routine work practices and policies, implementation leads to “long–term changes in how people work, treat one another, and structure their organizations” [2]. It is these “second–level effects” on the social system of interdependent people, events, and behaviors that are most pervasive and most important for organizations. These effects are also the most difficult to predict.

Again, e–mail offers illustrations of first– and second–level effects of introducing an information technology into organizations. The instrumental uses of e–mail are many: it offers rapid interpersonal communication within the organization and between the organization and the external world, whether clients, suppliers, members, customers, citizens, colleagues, friends, or family. E–mail is convenient and portable. Because it is asynchronous, it can improve time management by enabling people to send and receive messages at their convenience. It serves as a broadcast technology, allowing an organization to deliver the same message to a mass audience of its employees, students, or other groups simultaneously. E–mail has radically increased the speed and volume of communication for most people who use it.

We are finding many second–level effects of e–mail that were not anticipated at the time of its initial development or adoption. E–mail is easily abused, whether by broadcasting messages that are of interest only to a few or by sending rude and inappropriate messages that are unlikely to be communicated by other means. Junk e–mail can proliferate, resulting in inefficient use of staff time to sort through it, rather than the efficiency of communication intended. Once an organization adopts e–mail, usually everyone who is provided access is expected to use it regularly. People are expected to respond to messages, and to do so quickly. As a result, memos and other communications that did not require a response in paper form now result in a flurry of acknowledgments and responses, adding another layer of communication activity.

Communications that once were oral, or confined to one or a few paper copies that were controlled by the individuals involved, are now captured in permanent form on an organization’s e–mail servers. As a result, organizations are faced with a difficult balance between controlling their resources and the rights of individuals to their privacy (Anderson, et al., 1995; Berghel, 1997b). Organizations that read employees’ e–mail may defend this practice on the grounds that e–mail is organizational documentation and that it resides on computers owned by the organization. Individuals, particularly those who have lost jobs over the content of e–mail messages, may contend that e–mail is the equivalent of telephone or other oral communications and is subject to reasonable expectations of privacy.

Conversely, organizations are learning that e–mail can have unexpected and adverse legal consequences. Conversations that once were oral and now are recorded can be treated as legal evidence. Among the evidence that convicted Oliver North in the Iran–Contra affair were e–mail messages that he had deleted; they were recovered from backup storage as part of the legal discovery process. Similarly, e–mail messages internal to the Microsoft Corporation are being used by the U.S. Government as evidence in an antitrust case against the corporation. As a result of these and other cases, many organizations are expanding the scope of their e–mail policies to limit the content of e–mail messages and to minimize the archival storage of e–mail transactions (Harmon, 1998).

These are only a few of many examples of the positive and negative effects that e–mail has had on organizational communication (For more, see Anderson, et al., 1995; Berghel 1997b; Markus, 1994.). People’s experiences with e–mail and their perceptions of its role in an organization combine to determine how they will adapt it to their own practices.

As information technologies are more widely adopted, concern about their second–level effects is increasing. These concerns cross many disciplines, levels of analysis, and research methods. “Social informatics” is an emerging research area that brings together the concerns of information, computer, and social scientists with those in the domains of study (Bishop and Star, 1996; Borgman, et al., 1996; Bowker, et al., 1996). Social informatics scholars are attempting to build upon research in the design and the use of information systems and upon social studies of science and technology. This book brings a social informatics perspective to bear on access to information in digital libraries and in a global information infrastructure, considering first-level effects when these are all that can be known and second–level effects where possible.



Creating a Global Information Infrastructure

The integration, interaction, and interdependence of information–related tasks and activities leads us to think in terms of an information infrastructure. Rather than relying on separate devices for producing text (e.g., typewriters and personal computers), producing images (e.g., personal computers, photocopy machines, drawing pads), communicating with individuals (e.g., telephones, telefacsimile (fax) machines, mailboxes and stamps), and searching for information resources (e.g., personal computers, local servers, print technologies), all these tasks can be accomplished via a personal computer connected to the Internet. Conversely, these tasks can be divided up in many new ways by means of specialized devices such as cell phones, pagers, palmtops, and other “information appliances” that can share information. Computer and communication networks enable the integration of tasks and activities involved in creating, seeking, and using information, increase the interaction between these activities, and make them ever more interdependent.

In considering the premise and the promise of a “global information infrastructure,” we must determine what is meant by this phrase. Already it is used in variety of contexts, with meanings that include a set of technologies, a set of principles for an international computing and communications network, and a loose aggregation of people, technology, and content.



What Is Infrastructure?

Terms such as “national information infrastructure” and “global information infrastructure” are being bandied about with minimal discussion of what is meant by “infrastructure.” Social scientists and historians are beginning to take a research interest in this concept, particularly as it relates to organizational communication and work practices. Star and Ruhleder [3] describe infrastructure as follows:

It is both engine and barrier for change; both customizable and rigid; both inside and outside organizational practices. It is product and process. ... With the rise of decentralized technologies used across wide geographical distance, both the need for common standards and the need for situated, tailorable and flexible technologies grow stronger.

Star and Ruhleder are among the first to describe infrastructure as a social and technical construct. Their eight dimensions [4] can be paraphrased as follows: An infrastructure is embedded in other structures, social arrangements, and technologies. It is transparent, in that it invisibly supports tasks. Its reach or scope may be spatial or temporal, in that it reaches beyond a single event or a single site of practice. Infrastructure is learned as part of membership of an organization or group. It is linked with conventions of practice of day–to–day work. Infrastructure is the embodiment of standards, so that other tools and infrastructures can interconnect in a standardized way. It builds upon an installed base, inheriting both strengths and limitations from that base. And infrastructure becomes visible upon breakdown, in that we are most aware of it when it fails to work — when the server is down, the electrical power grid fails, or the highway bridge collapses.

As a means to explore the technical and public policy implications of information infrastructure, the Corporation for National Research Initiatives has sponsored a series of studies that address historical examples of large–scale infrastructure. These include studies of the growth of railroads, telephony and telegraphy, electricity and light, and banking (Friedlander 1995a, 1995b; 1996a, 1996b). In each case, the technologies involved took some time to be adopted, to stabilize, and to achieve the critical mass necessary to form an infrastructure. Railroads, telephones, power companies, and banks all provided local services for years, or even decades, before reaching nationwide connectivity. Each developed with some combination of public and private investment and government regulation. The means by which an integrated infrastructure evolved varied, and each involved experimentation with different forms of technology, regulation, and social arrangements.


Figure 3


Models of infrastructure for railroads, telephones, energy, and banking could have taken far different forms than they did. Indeed, with the possible exception of railroads, each of these infrastructures is still evolving actively. Telephony underwent extensive restructuring in the United States during the 1980s and the 1990s due to changes in regulatory structure, mergers and acquisitions, and technological advances. Similar regulatory restructuring is now underway in Europe and elsewhere. Meanwhile, technology advances and mergers and acquisitions continue apace. On the energy front, models for service provision are changing as energy companies are privatized and global power relationships shift with variations in supplies and prices of fossil fuels. On the financial front, models for banking infrastructure are under scrutiny as markets for stocks, commodities, currencies, and other financial instruments are becoming much more tightly coupled.

Each of these infrastructures is deeply embedded in our social fabric, relies on technical standards, and builds upon an installed base within the scope of its own and other infrastructures. A corollary to the notion that infrastructure becomes visible upon breakdown is that we rarely are aware of it when it is functioning adequately. We often fail to recognize these as essential infrastructures until telephone service becomes more complex and expensive, energy services change in cost and character, or the stock market takes a precipitous fall in value. And, although Americans make minimal use of railroads, railroads are an essential form of transportation in much of the world, where people are very much aware of changes in schedules, routes, prices, and services.

Star and Ruhleder’s (1996) set of eight infrastructure dimensions highlights the complex interaction of technology, social and work practices, and standards. They also emphasize social context by noting that infrastructure builds upon an installed base. An information infrastructure is built upon an installed base of telecommunications lines, electrical power grids, and computing technology, as well as on available information resources, organizational arrangements, and people’s practices in using all these aspects. An installed base establishes a set of capabilities and a set of constraints that influence future developments. For example, mobile telecommunications must interoperate with land–based networks, and new computers should be able to read files that were created on the preceding generation of technology.

The concepts of embeddedness, transparency, and visibility are especially relevant to a discussion of a global information infrastructure. To be effective, a GII must be embedded in the technical and social infrastructure of the many nations and cultures it reaches — so much so that the infrastructure is invisible most of the time. Whether this degree of embeddedness is possible across countries and cultures is examined throughout this book. When an information infrastructure works well, people depend on it for critical work, education, and leisure tasks, taking its reliability for granted. When it breaks down (for example, when e–mail cannot be sent or received, when transferred files cannot be read, or when online information stores cannot be reached), then the information infrastructure becomes very visible. People may resort to alternative means to complete the task, if those means exist; they may create redundant systems at considerable effort and expense; and they will trust the infrastructure a bit less each time it breaks down.



Infrastructure as Public Policy

Infrastructures of many kinds are subject to public policy. For example, the Clinton Administration (1997, 1998) set forth a policy on “critical infrastructure protection” that is noteworthy for our concerns. The white paper on Presidential Decision Directive (PDD) 63 (Clinton Administration, 1998) defines “critical infrastructures” as “those physical and cyber–based systems essential to the minimum operations of the economy and government. They include, but are not limited to, telecommunications, energy, banking and finance, transportation, water systems, and emergency services, both governmental and private.” In the past, these infrastructures were physically and functionally separate. However, with advances in information technology these systems are increasingly linked and interdependent. The significance of this interdependence is that critical systems are ever more vulnerable to “equipment failures, human error, weather and other natural causes, and physical and cyber attacks.” PDD 63 has the goal of protecting critical infrastructure from intentional attack and minimizing service disruptions due to any other form of failure.

Information technologies link these critical infrastructures, making them interdependent, and thus all information technologies could be considered parts of an information infrastructure. Information infrastructure usually is more narrowly defined in public policy documents, however. Typically the scope includes computing and communications networks, associated information resources, and perhaps a set of regulations and policies governing use.

Metaphors for Information Infrastructure

Clever metaphors for information infrastructure have helped to capture public attention. The concept of information infrastructure is best known in common parlance as the “information superhighway” (Gore, 1994b), or sometimes as the “I–way” or the “Infobahn.” These metaphors for information infrastructure emphasize the roads or pipes over which data flow, whether telecommunications, broadcast, cable, or other channels. The highway metaphor captures only a narrow sense of infrastructure, as it does not encompass information content, communication processes, or the larger social, political, and economic context. The superhighway metaphor is misleading both because it skews public understanding toward a low–level infrastructure and because it suggests that the government would pay the direct costs of the highway’s construction. The Internet was constructed with a combination of government and private funds. Current public policy, especially in the United States, is oriented toward private funding for further expansion (Branscomb and Kahin, 1995; Kahin and Abbate, 1995; Kahin and Keller, 1995).

Though metaphors such as the information superhighway have been extremely effective in marshalling support for information infrastructure development, far more is involved than laying roads over which information will travel.

National and International Policies

Individual countries began plans for national information infrastructures in the early 1990s (see, e.g., Information Infrastructure Program, 1992; Karnitas, 1996). In the United States, there was the National Information Infrastructure Act of 1993. In Europe, there was the European Union’s proposal for a European Information Infrastructure (Bangemann Report, 1994). The installed base of technology on which these plans are predicated includes the Internet, which began in the late 1960s with the ARPANET (National Research Council, 1994; Quarterman, 1990), the “intelligent network” of telecommunications that followed the deregulation of telephony (Mansell, 1993), and related technologies such as cable and satellite television networks.

In the mid 1990s, national information infrastructure plans began to converge. In 1994 the United States proposed formal principles for a global information infrastructure. The following principles were incorporated into the International Telecommunication Union’s “Buenos Aires Declaration on Global Telecommunication Development for the 21st Century” (1994) and the United States’ “Global Information Infrastructure: Agenda for Cooperation” (Brown, et al., 1995):

A few months later, the Group of Seven [5] (seven leading industrialized nations, known as “G–7”) met to discuss these principles and agreed to collaborate “to realize their common vision of the Global Information Society” and to work cooperatively to construct a global information infrastructure [6]. These principles emerged from the 1995 G–7 meeting:


The G–7 document also included the following.

These principles will apply to the Global Information Infrastructure by means of:

  • promotion of interconnectivity and interoperability
  • developing global markets for networks, services, and applications
  • ensuring privacy and data security
  • protecting intellectual property rights
  • cooperating in R&D and in the development of new applications
  • monitoring the social and societal implications of the information society.

The Buenos Aires and G–7 statements have much in common: they are concerned with technical capabilities (“interconnectivity,” “interoperability,” “open access”), promises of rights to provide network services (“open competition,” “dynamic competition”), guarantees of network services (“universal service,” “equality of opportunity”), a means of funding network development (“encouraging private investment”), and a means of regulating various aspects of its development and use (“flexible regulatory environment,” “adaptable regulatory framework”). However, they vary on their treatment of content: the G–7 principles promote diversity of content and offer some general protections (“privacy,” “data security,” “intellectual property”), while the telecommunications principles do not mention content, addressing only the development and regulation of communication channels.

Implementing Global Policy

Statements by the G–7 and other multinational bodies such as the United Nations promote policy agendas of the countries involved, but they lack the force of law and they provide little if any funding for implementation. Some of the language offers more platitudes than policy, such as the claim in the European Information Infrastructure plan that, “as a strategic creation for the whole Union,” it will lead to “a more caring European society with a significantly higher quality of life” (Bangemann Report, 1994).

The G–7 policy statements that frame a global information infrastructure have raised considerable concern about human rights and social protections from adverse consequences of its use. Though the G–7 principles include a general statement about privacy and comment on the need to monitor the social implications of the information society, they do not ensure legal protection of rights such as privacy, free expression, and access to information. Despite requests by human rights groups, the G–7 principles omit references to assurances in the United Nations Declaration of Human Rights that were approved in 1948 (see United Nations, 1998). Particularly relevant are Articles 12 and 19:

Article 12: No one shall be subjected to arbitrary interference with his privacy, family, home or correspondence, nor to attacks upon his honor and reputation. Everyone has the right to the protection of the law against such interference or attacks.

Article 19: Everyone has the right to freedom of opinion and expression; this right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers.

These principles are receiving renewed attention upon the fiftieth anniversary of their adoption (United Nations, 1998). Computer networks offer unanticipated capabilities for free speech and access to information. Because transactions and interactions are easily trackable, computer networks also can create unanticipated intrusions into privacy (Kang, 1998). Many privacy advocates promote an alternative design model, known as “privacy–enhancing technologies” (Burkert, 1997), in which individuals can acquire access to most information services without revealing their identity if they so choose. Privacy, freedom of speech, and freedom of access to information are tenets of democracy (Dervin, 1994; Lievrouw, 1994a, 1994b). People cannot speak freely or seek information freely if their movements are being tracked and if they cannot protect and control data about themselves (Agre and Rotenberg, 1997; Diffie and Landau, 1998; Information Freedom and Censorship, 1988, 1991).

These are contentious issues in the United States. One example is that the federal policy on critical infrastructure protection, discussed above, is being challenged on the basis of its potential to erode civil liberties (Electronic Privacy Information Center, 1998). Public policy on social aspects of information infrastructure is subject to the laws, the norms, and the practices of individual countries and jurisdictions, despite the global reach of computer networks. When local activities took place only locally, variances in policy and regulation were less apparent and jurisdiction was rarely an issue. Now that individual communications and information resources flow quickly and in vast quantities across borders, variances in policy and regulation can be highly visible and jurisdiction can be highly contentious. Privacy rights and regulations have become an international battlefield where many of these issues are being played out.

The European Union Data Directive, which took effect in late 1998, highlights fundamental differences in policy approaches to privacy protection. The United States long has taken a “sector approach,” with specific laws governing credit reports, library borrowing records, videotape rentals, federal government databases, etc. In the new arena of computer networks, U.S. policy has favored self–regulation by the private sector over government–imposed regulation. In contrast, European countries have favored generalized policies over the control of personal data, assigning stronger rights to individuals to control information about themselves than to organizations that collect and manage personal data. The EU Data Directive consolidates the policies of individual countries and regulates privacy protections throughout the European Union. In view of the extensive commerce between the United States and the European Union and the volumes of data about personnel, customers, clients, and suppliers that are subject to regulation, the policies of these jurisdictions often are in conflict.

For overviews of the rapidly evolving landscape of electronic privacy, see Agre and Rotenberg (1997), Diffie and Landau (1998), Kang (1998), Rotenberg (1998), and Schneier and Banisar (1997). Updates, including pointers to government documents and other primary sources, can be found at and at



Information Infrastructure as a Technical Framework

“Information infrastructure” can refer to a technical framework rather than to a public policy. As defined by the U.S. National Research Council [7], an information infrastructure is “a framework in which communications networks support higher–level services for human communication and access to information. Such an infrastructure has an architectural aspect — a structure and design — that is manifested in standard interfaces and in standard objects (voice, video, files, email, and so on) transmitted over the interfaces.”

One of the key components in defining an information infrastructure as a technical framework is for it to have an open architecture that will enable all parties to interconnect electronically and to exchange data. The “Open Data Network” concept (National Research Council, 1994) follows both from the Internet (a successful open architecture for computing) and from established telecommunications policy principles (Mansell, 1993; National Research Council, 1994). Under the G–7 principles, closed networks can interconnect with the open network; closed service networks such as cable television are allowed under other communications regulations as well. As we move toward ubiquitous computing, a wider array of devices must interconnect; this makes open systems and interoperability much more essential.

The emerging global network that interconnects a wide variety of computing devices located around the world offers great utility for communication between individuals and organizations, whether for education, work, leisure, or commerce. The technical framework for such an information infrastructure is now expected to support a range of tasks and activities far wider than that for which it was originally designed, however. The original ARPANET and the early generations of the Internet were constructed by and for the research, development, and education communities (Quarterman, 1990). Benign uses by a collegial community were presumed when its technical architecture was designed (Oppliger, 1997).

Substantial enhancements are being made to the technical architecture of the Internet to support a vastly larger volume and variety of users, capabilities, and services than was anticipated in the original design. Two new network services illustrate the scope of the improvements that are under way (Lawton, 1998; Lynch, 1998). One is “quality of service”: the ability to reserve a set amount of bandwidth, at a predetermined level of quality, in advance. Rather than the current model, which is largely “first come, first served” for bandwidth usage, mostly at flat pricing, the new model supports differential pricing for differential services. Many organizations are willing to pay a premium to guarantee adequate bandwidth at a specified time (for a teleconference or a distance–education course, for example). Conversely, many individuals are willing to tolerate delays in e—mail delivery or Web access in return for lower costs. In view of the complexity of Internet architecture and the number of political and service—provider boundaries crossed by an individual transmission, guaranteeing quality of service will not be a simple accomplishment. Though quality of service is considered an essential capability of an information infrastructure, precise assessments of what can be guaranteed and how it can be measured have yet to be established (Lynch, 1998).

Multicasting is another long–awaited service improvement for the technical framework of a global information infrastructure. At present, most communications are point–to–point (“unicasting”): copies of a message are sent individually to each intended recipient. The alternative is broadcasting, in which one message is sent to all users of the network, whether they want it or not. An intermediate model is “multicasting”: one message is sent to a selected group of recipients, reducing the amount of bandwidth required. Technically, under multicasting, the originating server sends one message to each network router on which intended recipients are located and that router re–sends to its local subscribers (Lawton, 1998). As with quality of service, the number of providers involved makes multicasting a complex process, but one that is necessary for efficient use of bandwidth on a global information infrastructure (Lynch, 1998). A variety of economic and technical models for network service provision are under consideration for the next generation of network architecture (Shapiro and Varian, 1999).

The Internet is already a “network of networks.” A global information infrastructure will be even more so. Though we speak metaphorically of a single open network, in actuality the Internet links many layers of networks within organizations, within local geographic areas, within countries, and within larger geographical regions. These go by various names, including intranets, extranets, local–area networks (LANs), metropolitan–area networks (MANs), and even tiny–area networks (TANs). Suffice it to say that the information infrastructure topography is becoming increasingly complex, linking together internal organizational networks, closed networks such as cable TV, and the international Internet.


Figure 4


The boundaries of individual networks can be controlled to varying degrees. A common technique is to protect organizational or even national networks with “firewalls” that limit the abilities of authorized users to exit and of outsiders to enter. Some internal resources can be publicly accessible while others are restricted to internal use, for example. Similarly, firewalls and filtering techniques can be used to limit external sites that can be reached. Parents can limit their children’s ability to connect to sites known to contain pornography or other undesirable material. The definition of “undesirable” varies by context. Companies can limit access to known sites containing games. Countries can limit access to sites known to provide undesirable political views. China, for example, currently attempts to control access to sites outside the country through a single gateway, so that specific sites deemed objectionable can be blocked. Chinese Internet users are required to register with the police to gain access to the network (Tan, et al., 1997). A key phrase here is “known sites.” As the Internet proliferates, new sites appear daily, and sites change names, location, and content frequently. Reliable filtering software that can distinguish between acceptable and unacceptable materials is not yet feasible, and may never be.

For most businesses and governments, security and risk management are far greater concerns than is pornography. After connectivity, the most important enabling technology for electronic commerce is security (Dam and Lin, 1996; Geer, 1998; Oppliger, 1997). One model being studied and implemented is “trust management,” in which mechanisms such as cryptography are employed to verify the identities of all parties involved in electronic transactions. Such transactions include buying and selling goods or services, transferring secure data (such as financial transactions between banks and stock markets), and proprietary communications within organizations or between organizations and their clients, customers, and suppliers. Both retail transactions between individuals and companies and wholesale transactions between companies can be accommodated. An alternative model is “risk management,” which focuses on the likelihood of losses and the size of potential losses from electronic commerce. Rather than assume that trust can be guaranteed in all transactions, parties try to determine the degree of risk exposure and to insure against it. Cryptography is essential to both models as a means of assuring the authenticity of transactions to the extent possible. The frontiers of electronic commerce are being tested in the financial markets today. In view of the size and volume of transactions among banks, stock markets, investors, and other parties, many technical and policy aspects of information infrastructure are likely to be tested first in this arena (Geer, 1998).



Information Infrastructure as Technology, People, and Content

Among the broadest conceptualizations of an information infrastructure is that presented in National Information Infrastructure: Agenda for Action (1993), where an NII is defined as encompassing a nation’s networks, computers, software, information resources, developers, and producers. This definition comes closer to capturing the larger sense of infrastructure as a complex set of interactions between people and technology than do most other public policy statements, technical definitions, or metaphors.

The above definition is compelling, if vague, because it recognizes that we are creating something new and something that is more than a sum of its parts. The information infrastructure is not a substitute for telephone, broadcast, or cable networks, for computer systems, for libraries, archives, or museums, for schools and universities, for banks, or for governments. Rather, it is a new entity that incorporates and supplements all these technologies and institutions but is not likely to replace any of them. However, a GII is likely to change each of these institutions, and how people use them, in profound ways.

The term “global information infrastructure” is used in this broad sense throughout the present book. A GII consists of a technical framework of computing and communications technologies, information content, services, and people, all of which interact in complex and often unpredictable ways. No single entity owns, manages, or controls the technical framework of a GII, although many governments, vast numbers of public and private organizations, and millions of people contribute to it and use it. The GII is perhaps best understood by the metaphor of the elephant being examined by a group of blind people — each one touches a different part of the beast, and thus senses a different entity. From this perspective, a global information infrastructure is a means for access to information. However, it can be viewed from many complementary perspectives that also are valid.




These are exciting times. Information technologies are increasing in speed, power, and sophistication, and they now can link together a vast array of devices into a network that spans the globe. They offer new ways of learning, working, and playing, as well as conducting global commerce. Some contend that these changes are revolutionary and will change the world; others argue that the changes are evolutionary, and that individuals and organizations will incorporate networked information technologies into their practices just as they incorporated many earlier media and technologies. In this book I take the view that these changes are neither revolutionary nor evolutionary but somewhere between: that they are co–evolutionary. New technologies are based on perceived needs and available capabilities. People adopt these new technologies if and when they deem the technologies useful and when they deem the effort and the costs appropriate. Sometimes individuals make these decisions; sometimes organizations make them. The result is that some technologies are adopted by some of the people some of the time. No matter how voluntary or involuntary the adoption process, individuals and organizations adapt technologies to their interests and practices, often in ways not anticipated by the designers of those technologies. Information technologies are more flexible and malleable to individual practices than are most other innovations, and this makes them especially adaptable. They also evolve more quickly than most other innovations, with new and improved versions appearing at a dizzying rate.

Adoption and adaptation of technology are difficult to predict, owing to the complex interactions between characteristics of information technologies, practices of individuals and organizations, economics, public policy, local cultures, and a host of other factors. Organizations acquiring new technologies find that estimates of first–level effects, such as those on productivity and profits, are unreliable. Reliable predictions of longer–term, second–level effects, such as those on organizational communication and structure, are nearly impossible. One reason is that external factors, such as changes in the legal status of electronic communications, can have profound effects on how individuals and organizations use information technologies.

We are in the process of creating a global information infrastructure that will interconnect computer networks and various forms of information technologies around the world. After a review of some of the many meanings of “information infrastructure,” it was determined that the concept incorporates people, technology, and content and the interactions between them. This broad definition incorporates definitions of information infrastructure as a set of public policies and as a technical framework. The broader definition is best suited to studying the co-evolution of technology and behavior as related to access to information, which is the primary concern of this book. An information infrastructure is only one of several infrastructures that are essential to a well–functioning society. Others include energy, transportation, telecommunications, banking and finance, transportation, water systems, and emergency services. Because each of these infrastructures is increasingly reliant on information technologies, they are more interconnected and interdependent. Their interdependence means that more and more aspects of daily life depend on the emerging global information infrastructure. End of article


About the author

Christine L. Borgman is Professor and Presidential Chair in Information Studies at the University of California, Los Angeles, and Visiting Professor at Loughborough University, England.
E–mail: cborgman [at] ucla [dot] edu



1. Editors of Wired (1997), p. 14.

2. Sproull and Kiesler (1991), p. 1.

3. Star and Ruhleder (1996), pp. 111–112.

4. Ibid., p. 113.

5. The Group of Seven nations are Canada, France, Germany, Italy, Japan, the United States, and the United Kingdom. Russia has participated in some recent meetings. When Russia is involved, the press sometimes refers to these as meetings of “the G–7 plus Russia” or “the G–8.”

6. G–7 Ministerial Conference on the Information Society (1995a), pp. 1–2.

7. U.S. National Research Council (1994), p. 22.



P.E. Agre and M. Rotenberg (editors), 1997. Technology and Privacy: The New Landscape. Cambridge, Mass.: MIT Press.

R.H. Anderson, T. Bikson, S.A. Law, and B.M. Mitchell, 1995. Universal Access to E–mail: Feasibility and Societal Implications. Santa Monica, Calif.: Rand, at

Bangemann Report, 1994. Europe and the Global Information Society: Recommendations to the European Council. Brussels: European Council.

M.J. Bates, 1989. “The design of browsing and berry–picking techniques for the online search interface,” Online Review, volume 13, number 5, pp. 407–424.

H. Berghel, 1997a. “Cyberspace 2000: Dealing with information overload,” Communications of the ACM, volume 40, number 2, pp. 19–24.

H. Berghel, 1997b. “E–mail — the good, the bad, and the ugly,” Communications of the ACM, volume 40, number 4, pp. 11–15.

J. Birnbaum, 1997. “Pervasive information systems (The Next 50 Years: Our Hopes, Our Visions, Our Plans),” Communications of the ACM, volume 40, number 2, pp. 40–41.

A.P. Bishop and S.L. Star, 1996. "Social informatics for digital library use and infrastructure," In: M.E. Williams (editor). Annual Review of Information Science and Technology, volume 31. Medford, N.J.: Information Today, pp. 301-401.

C.L. Borgman, D. Moghdam, and P.K. Corbett, 1984. Effective Online Searching: A Basic Text. New York: Dekker.

C.L. Borgman, S.G. Hirsh, and J. Hiller, 1996. “Rethinking online monitoring methods for information retrieval systems: From search product to search process,” Journal of the American Society of Information Science (ASIS), volume 47, number 7, pp. 568–583.<568::AID-ASI8>3.0.CO;2-S

G. Bowker, S.L. Star, W. Turner, and L. Gasser (editors), 1996. Social Science, Technical Systems, and Cooperative Work: Beyond the Great Divide. Hillsdale, N.J.: Lawrence Erlbaum.

L.M. Branscomb and B. Kahin, 1995. “Standards processes and objectives for the national information infrastructure,” In: B. Kahin and J. Abbate (editors). Standards Policy for Information Infrastructure. Cambridge, Mass.: MIT Press, pp. 3–31.

R.H. Brown, L. Irving, A. Prabhakar, and S. Katzen, 1995. The Global Information Infrastructure: Agenda for Cooperation (March), at

H. Burkert, 1997. “Privacy–enhancing technologies: Typology, critique, vision,” In: P.E. Agre and M. Rotenberg (editors), 1997. Technology and Privacy: The New Landscape. Cambridge, Mass.: MIT Press, pp. 125–142.

Clinton Administration, 1997. President’s Commission on Critical Infrastructure Protection: Executive Order 13010 (10 July 10), at

Clinton Administration, 1998. White Paper: The Clinton Administration’s Policy on Critical Infrastructure Protection: Presidential Decision Directive 63 (22 May), at Reprinted in Electronic Privacy Information Center, 1998. Critical Infrastructure Protection and the Endangerment of Civil Liberties. Washington, D.C.: Electronic Privacy Information Center, pp. 36–49.

K.W. Dam and H.S. Lin (editors), 1996. Cryptography’s Role in Securing the Information Society. Computer Science and Telecommunications Board; Commission on Physical Sciences, Mathematics, and Applications; National Research Council. Washington, D.C.: National Academy Press.

B. Dervin, 1994. “Information — democracy: An examination of underlying assumptions,” Journal of the American Society of Information Science (ASIS), volume 44, number 8, pp. 480–491.

W. Diffie and S. Landau, 1998. Privacy on the Line: The Politics of Wiretapping and Encryption. Cambridge, Mass.: MIT Press.

Electronic Privacy Information Center, 1998. Critical Infrastructure Protection and the Endangerment of Civil Liberties: An Assessment of the President’s Commission on Critical Infrastructure Protection (PCCIP). Washington, D.C.: Electronic Privacy Information Center, at

A. Friedlander 1995a. Emerging Infrastructure: The Growth of Railroads. Reston, Va.: Corporation for National Research Initiatives,

A. Friedlander 1995b. National Monopoly and Universal Service: Telephones and Telegraphs in the U.S. Communications Infrastructure, 1837–1940. Reston, Va.: Corporation for National Research Initiatives,

A. Friedlander 1996a. Power and Light: Electricity in the U.S. Energy Infrastructure, 1870–1940. Reston, Va.: Corporation for National Research Initiatives,

A. Friedlander 1996b. “In God We Trust”: All Others Pay Cash: Banking as an American Infrastructure, 1800–1935. Reston, Va.: Corporation for National Research Initiatives,

D.E. Geer, 1998. “Risk management is where the money is,” Forum on Risks to the Public in Computers and Related Systems, ACM Committee on Computers and Public Policy, Risk–Forum Digest, volume 20, number 6 (12 November),

A. Gore, 1994a. “Al Gore, Speech on U.S. Vision for the Global Information Infrastructure,” World Telecommunications Development Conference, Buenos Aires (March). Quoted in R.H. Brown, L. Irving, A. Prabhakar, and S. Katzen, 1995. The Global Information Infrastructure: Agenda for Cooperation (March), at

A. Gore 1994b. “The Information Superhighway,” Speech, The Superhighway Summit, University of California, Los Angeles (11 January).

A. Harmon, 1998. “Corporate delete keys busy as e–mail turns up in court,” New York Times (11 November), pp. A1, C2 (national edition).

D. Hillis, 1997. “A time of transition,” Communications of the ACM, volume 40, number 2, pp. 37–39.

Information Freedom and Censorship, 1988. The Article 19 World Report 1988. Prepared by the staff of Article 19. Harlow: Longman.

Information Freedom and Censorship, 1991. The Article 19 World Report 1991. Prepared by the staff of Article 19. London: Library Association Publishing.

Information Infrastructure Program/IIF/, 1992. Budapest: Hungarian Academy of Sciences, Computer Science Institute (Sztaki).

B. Kahin and J. Abbate (editors), 1995. Standards Policy for Information Infrastructure. Cambridge, Mass.: MIT Press.

B. Kahin and J. Keller, 1995. Public Access to the Internet. Cambridge, Mass.: MIT Press.

J. Kang, 1998. “Information privacy in cyberspace transactions,” Stanford Law Review, volume 50, number 4, pp. 1193—1294.

E. Karnitas, 1996. Latvian Infrastructure Infrastructure: The Present Stage and Plans for Development. Riga: Latvian Academy of Sciences.

J. Katz and P. Aspden, 1997. “Motives, hurdles, and dropouts,” Communications of the ACM, volume 40, number 4, pp. 97–102.

R.R. Korfhage, 1997. Information Storage and Retrieval. New York: Wiley.

G. Lawton, 1998. “Multicasting: Will it transform the Internet?” IEEE Computing, volume 31, number 7, pp. 13–15.

M.E. Lesk, 1997. Practical Digital Libraries: Books, Bytes, and Bucks. San Francisco: Morgan Kaufman.

J.C.R. Licklider and A. Vezza, 1978. “Applications of information networks,” IEEE Proceedings, volume 66, pp. 1330–1346.

L.A. Lievrouw (editor), 1994a. “Special topic issue: Information resources and democracy,” Journal of the American Society of Information Science (ASIS), volume 44, number 8, pp. 350–421.<350::AID-ASI1>3.0.CO;2-2

L.A. Lievrouw, 1994b. “Information resources and democracy: Understanding the paradox,” Journal of the American Society of Information Science (ASIS), volume 44, number 8, pp. 350–357.<350::AID-ASI1>3.0.CO;2-2

C. Lynch, 1998. “The evolving Internet: Applications and network service infrastructure,” Journal of the American Society of Information Science (ASIS), volume 49, number 11, pp. 961–972.<961::AID-ASI2>3.0.CO;2-#

R. Mansell, 1993. The New Telecommunications: A Political Economy of Network Evolution. London: Sage.

G. Marchionini, 1995. Information Seeking in Electronic Environments. New York: Cambridge University Press.

M.L. Markus, 1994. “Finding a happy medium: Explaining the negative effects of electronic communication on social life at work,” ACM Transactions on Information Systems, volume 12, number 2, pp. 119–149.

National Information Infrastructure: Agenda for Action, 1993.

National Research Council, 1994. Realizing the Information Future: The Internet and Beyond. Washington: D.C.: National Academy Press.

“The Next 50 Years”, 1997. Communications of the ACM, volume 40, number 2, pp. 3–102, passim. Special 50th anniversary of ACM issue.

R. Oppliger, 1997. “Internet security: Firewalls and beyond,” Communications of the ACM, volume 40, number 5, pp. 92–102.

J. Pontin, 1998. “The post–PC world. The new era of ubiquitous computing,” Red Herring, number 61 (December), pp. 50–66, passim.

J.S. Quarterman, 1990. The Matrix: Computer Networks and Conferencing Systems Worldwide. Bedford, Mass.: Digital Press.

E.M. Rogers, 1983. Diffusions of Innovations. Third edition. New York: Free Press.

E.M. Rogers, 1986. Communication Technology: The New Media in Society. New York: Free Press.

M. Rotenberg, 1998. The Privacy Law Sourcebook: United States Law, International Law, and Recent Developments. Washington, D.C.: Electronic Privacy Information Center, at

B. Schneier and D. Banisar, 1997. The Electronic Privacy Papers: Documents on the Battle for Privacy in an Age of Surveillance. New York: Wiley.

C. Shapiro and H.R. Varian, 1999. Information Rules: A Strategic Guide to the Network Economy. Boston: Harvard Business School Press.

L. Sproull and S. Kiesler, 1991. Connections: New Ways of Working in the Networked Organization. Cambridge, Mass.: MIT Press.

S.L. Star and K. Ruhleder, 1996. “Steps towards an ecology of infrastructure: Design and access for large information spaces,” Information Systems Research, volume 7, number 1, pp. 111–134.

Z. Tan, M. Mueller, and W. Foster, 1997. “China’s new Internet regulations: Two steps forward, one step back,” Communications of the ACM, volume 40, number 12, pp. 11–16.

United Nations, 1998. Universal Declaration of Human Rights. Fiftieth Anniversary. Adopted and proclaimed by General Assembly Resolution 217 A (III) of 10 December 1948. High Commissioner for Human Rights.

T. Winograd, J. Bennett, L. De Young, and B. Hartfield (editors), 1996. Bringing Design to Software. Reading, Mass.: Addison–Wesley.

Wired editors, 1997. “‘Kiss your browser goodbye’ The radical future of media beyond the Web,” Wired, volume 5, number 3 (March), pp. 12–23.



This text represents the first chapter of From Gutenberg to the Global Information Infrastructure: Access to Information on the Networked World by Christine L. Borgman, published in May 2000 by MIT Press. Reprinted by permission of MIT Press and Christine L. Borgman. Copyright © 2000 Christine L. Borgman; All Rights Reserved.

From Gutenberg to the Global Information Infrastructure is available from MIT Press directly, fine bookstores everywhere, and other sources such as


Editorial history

Paper received 5 July 2000; accepted 24 July 2000.

Copyright © 2000, First Monday

The Premise and the Promise of a Global Information Infrastructure
by Christine L. Borgman
First Monday, Volume 5, Number 8 - 7 August 2000