Why Communication Researchers Should Study the Internet: A Dialogue

Authors


Newhagen

Why should communication researchers study the Internet?

Rafaeli

To answer your question, we need a shorthand map of communication-related phenomena the Net represents. One could begin by saying that the Net is new, loaded with content, crowded, and seemingly a great business prospect. All four conditions are true. The Net and modes of communicating on and through it are a novelty, and the newness seems to be reborn all the time. New gadgets, new gimmicks every day. Content of all sorts abounds, from pornography to timely news, from scientific journals to mass entertainment to high-brow culture, and from th e very public to the very intimate. All forms of content are congregating on line. So are people. The rates of growth, conservatively estimated at over 5% a month several years going, are astounding. Big business, as well as garden variety entrepreneurs, are not blind to this. None of these-novelty, volume, head counts, or prospects of future profits-are distinct communication phenomena, nor do they serve to set the Net apart from communication as we have known it. Instead, we should search for the communication dimensions in our inquiry, with an eye to finding less ephemeral, more robust concepts. Those, I believe, are the things we would want to study.

Therefore, I propose focusing on five defining qualities of communication on the Net: multimedia, hypertextuality, packet switching, synchronicity, and interactivity. These qualities capture what is, or can be, differentabout Net-based communication. None of these qualities is necessarily realized by any one instance of behavior or application on the Net, and none of these qualities is necessarily good or bad, nor by the Net. However, Net-based communication highlights these dimensions.

The Sensory Appeal

Text, voice, pictures, animation, video, virtual-reality motion codes, even smell, are all already being conveyed on the Net. What journalists used to call the news hole has simply expanded into an unprecedented sensory vastness. It is embarrassing to be a designer on the Web nowadays.Anything goes. Much does, and, by the time it does, it could be overshadowed by new capabilities. The Net's capacity for addressing senses far surpasses that of any other medium. In a sense, this indicates that the medium serves less than ever before in a constraining, guiding role.

Is Communication Linear?

The “shackles” of linearity are being shed. These shackles traditionally have bound communication into a procrustean bed of predetermined order and a tyranny of writer over reader. Neither the authors nor the structure holds that much authority any more. The reader-audience member-receiver shoulders a lot more responsibility now. The World Wide Web is hypertextual, as is the special structure of threads that organize Usenet. Indeed, the hypertext sermon is now being heeded elsewhere, in everything from books for youth to classroom curriculum. The hypertext idea has been with us since before print, but it has come so much closer to becoming the dominant data structure that it deserves to become a focus of communication research.

The Topology of Communication

The organizing principle for routing traffic (“switching”) has always been a focus of communication research. Switching is, for example, a cornerstone of telephony. In mass media it was called “gatekeeping”; in interpersonal communication research, much is made of “turn taking.”

On the Net, due to historical reasons perpetuated by the discovery of other functions, the organizing principle is to have no organization, or deliberate, orderly, anarchy. The message keeps its own gate, carries its own homing device. The Net treats censorship as noise and is designed to work around it. Most other communication is either route directed (I talk to you, she broadcasts to us) or (at least) misdirected (he eavesdrops, they try to block transmission).

By contrast, much of the Net is designed to be route oblivious. Instead, it is packet switched. This technical fact has content, social, and sender-and-receiver results. It affects the kind of legislation that can be imposed on the Net, determines much of the pricing of information, and plays a big role in restructuring the social relations of “Netizens,” or citizens of the Net.

The Elasticity of Synchronicity

All communication is temporally sensitive. How fresh is communication when it is “served”? How quickly does it go stale? Are we becoming an impatient species? Do we have preferences for the old, tried, and true? Naturally, communication is synchronous, but how likely are we to want to communicate asynchronously? The phone has its answering machine. Theater has film and video. TV has VCR. Even interpersonal communication has memories.

The Net stretches the edges of the synchronicity continuum. Communication on the Net travels at unprecedented speed. It can also be consumed at unprecedented delays. Messages have time stamps, accurate to one hundredth of a second. A whole generation now knows what “GMT” stands for. Bandwidth has become the new gold. Archives confront us with the possibility and threat of longer memories (remember the “Irangate” electronic mail records? Witness the use of Usenet logs in various court cases). The Net seems to stretch the temporal overlaps and discrepancies between time zones. Interpersonal communication, once either face-to-face or time delayed, can now be both at once (for instance, the time warps introduced by discussion groups, videoconferencing, IRC, MUD/MOOs, etc.). All of this makes synchronicity both a process and an effects concern.

Interactivity

By interactivity we mean the extent to which communication reflects back on itself, feeds on and responds to the past. Communication on the Net serves to highlight the role of interactivity. It can be consciously programmed in or kept out. Interactivity is behind the issues of moderated or unmoderated Computer Mediated Communication (CMC) groups. Interactivity is an operating force in the allure of IRC and MUD. Good choices and crafty implementation of interactivity are often the difference between successful and failing Web sites.

Three concluding observations. First, all five proposed qualities (multimedia, hypertextuality, packet switching, synchronicity, interactivity) have an engineering nomenclature “sound.” Engineering, indeed, is their origin. Not surprisingly, much of “Net culture,” perhaps more than other cultures, is influenced by engineering. Maybe this is an echo of the (virtual?) reality that redredged McLuhanistic rhetoric, but the sense in which I use each of these terms is removed from the original, engineering-technological intent. This is one way to conceptualize the role of communication research with the Net: fortify unidimensional engineering concepts with social science thinking and study.

Second, and perhaps related to this, is the inviting empiricism inherent in Net behavior. Not only does it occur on a computer, communication on the Net leaves tracks to an extent unmatched by that in any other context-the content is easily observable, recorded, and copied. Participant demography and behaviors of consumption, choice, attention, reaction, learning, and so forth, are widely captured and logged. Anyone who has an opportunity to watch logs of WWW servers, and who is even a little bit of a social scientist, cannot help but marvel at the research opportunities these logs open.

Third, what aspects of the Net can we study? Almost everything. The same things we studied before and more. Each of the five qualities outlined above deserves descriptive scrutiny to establish to what extent it is present. As dependent variables, we would be interested in what brings about optimal levels of each (e.g., how do you get the most interactive group discussion, what causes a preference for synchronicity, etc.). How does each perform as an independent or intervening variable?

Newhagen

I agree with the main thrust of your argument, that defining the Internet is largely a function of the constraints we name when we conceptualize it. A “good” list will naturally point us toward theories that help us understand the importance of this revol utionary communication technology.

You point to the hype surrounding the phenomenon right now, and research focusing solely on its novelty will inevitably have a short shelf life. The hyperbole has become so hot that we are currently suffering a backlash against the Net in the popular press and elsewhere. However, I think we have to be careful not to use the current neo-Luddite vogue as cover to discount the Internet's importance as a communication technology.

Whereas the general question of why we must study the Internet is simple enough on its surface, a closer look reveals the topic to be very difficult to constrain conceptually. For example, in this symposium, December says times were simpler back in the heyday of television. Mass media can generally be conceptualized for their singular capacity to generate a number of replicated messages from one template and distribute them to large audiences. Television messages might be more complex than newspaper messages, to the degree that they include both nonverbal and verbal information across sensory modality. Beyond that, however, the linear architectures of the two communication systems are remarkably similar. Moreover, in their symposium article, Morris and Ogan challenge effects-researchers to move ahead in imagining how and why the distributed, nonlinear character of a computer network differs from mass media.

For me, the two things that set the Net apart from earlier communication technologies are this architectural difference and the digitization of information in the message stream. Notice that neither of these features is manifestly apparent or obvious to the user. I do not think there was much confusion about how and why television differed from newspapers when it diffused into the population 40 years ago. We lack that intuitive advantage with the Net, however. Computers as appliances look more and more like television receivers and vice versa.

This problem of having the critical dimensions of a new technology hidden from view also can be seen in the current trend among publishers to dump text into a computer network and call it an “electronic newspaper.” If journalism does not come to grips with the impact these architectural differences have on the way people use information, it may have trouble finding a home on the Net.

The digitization of information gives the computer programmer the ability to create highly abstract, hierarchical, object-oriented data structures. Digitization democratizes information and yields all binary units equal, either as 0 or 1, regardless of the higher-order symbol systems they represent, such as text, speech, or icon. The programmer can then create data classes, where an image might inherit characteristics from text. This technique makes the manipulation of data highly efficient.

Similarly, programmers and engineers have learned to make networks more efficient by breaking up the bottleneck at the central processor and distributing control to peripheral nodes. The metaphor of the neural network in computer programing is compelling, and communication technology may be on the threshold of extending the human central nervous system outside our skins, as McLuhan envisioned.

You point out that much of the discussion of the Net is couched in the jargon of engineering. Surely, some not versed in it might chafe at its emphasis and wonder what value understanding the engineering principles of computer networks has to do with human communication. I think the short-term, practical and the long-term, theoretical implications that come from having a firm grasp of the engineering principles behind this technology justify the emphasis on understanding engineering principles.

A colleague of mine likened the role of communication research to that of judges at a gymnastics competition, where the engineers roll out a new technology, and we hold up numbers from 1 to 10, rating it. I think we are compelled to take a more active role in the development of the technology itself, if for no other reason than to survive the current fierce competition for resources in the academic environment. The trick will be to think of concepts that bridge both the world of the engineer and that of the communication researcher. Two ways we might span those boundaries have to do with the technology's interface and its architecture.

The Net as Architecture and Interface

Multimedia and hypertext bear on the interface. I think of the user interface, that is, the place messages are physically embodied, when I hear about multimedia. The newspaper uses ink smeared on dead trees as its interface. The interface for television and computers is similar from the viewpoint of the cognitive psychologist in many ways. The idea of user-friendliness translates into the study of problem-solving strategies and issues of mental effort.

Similarly, I see the discussion of what will happen to the printed word on the Net as an interface issue. One interesting difference between television and computers as communications media is the prominence of text in the latter. We both agree that text is not inherently linear. It becomes linear when it is constrained on a printed page.

Your remaining two dimensions bear directly on communication-system architecture issues, that is, stuff going on behind the interface. The idea of packet switching goes to the heart of the Net's parallel- distributed architecture, its most distinguishing and compelling feature. Not much attention has been paid to this topic, perhaps because it is the least visible of all the dimensions.

For me, packet switching is simply the engineering innovation that made true interactivity possible in technologically mediated communication (analog technologies such as the telephone and ham radio notwithstanding). You should be given credit for recognizing the importance of interactivity to communication technology as early as you did. Interactivity is critical to the communication process and, ironically, the aspect which communication technologies have had the least success in implementing. Beyond the hype, the real promise of the Net as a communication technology may be in its ability to capture and even amplify this dimension.

Finally, I see one striking epistemological similarity among all the categories on your list. Currently, this new communication technology is being invented in the context of a discussion among electrical engineers, computer scientists, and cognitive psychologists, with some philosophers and linguists at the periphery. Com munication researchers are conspicuously absent from this discussion. You mention what a marvelous opportunity the Net provides for empirical research; however, the field currently stands in the shadow of more or less 50 years of mass media research that has yielded disappointingly weak effects. It is curious that as divided as the critical approach and empirical science seem to be concerning the study of mass media, neither group has ever seriously considered the possibility that the true impact of mass media is small.

On one hand, empiricists attributed the problem to statistical noise, attempting to capture more variance in their regression equations by refining their theories and honing their methods. On the other hand, the critical approach suggests the effects simply defy empirical scrutiny because of their complexity. Notice, however, that this perspective still supports the notion of strong effects; its supporters are simply throwing up their hands in abandon at the prospect of using empirical methods, especially cross-sectional surveys, as instruments to detect them.

Suppose for a minute, though, that effects-empiricists did their job just fine, that they found what was to be found. Suppose that the strength of the effects was muted, or even distorted, by mass media systems' inability to support truly interactive communication. Now imagine that vastly interconnected computers can support and even enhance interactivity in human communication. Such a prospect holds out new hope for the empiricist! However, that may be true primarily when the unit of analysis for such studies, at least initially, is at the individual level, where the experimental methods of the psychological are especially appropriate.

I am not staking out a radically reductionist position, quite the opposite; over time, questions at cultural and societal levels of analysis may well be the unique contribution of our field to the understanding of this new technology. However, to truly understand this new technology, those embracing the critical approach have to be comfortable with the issues across multiple levels of analysis, whereas the empiricists will have to show a greater, more eclectic tolerance for experimental science.

Rafaeli

Are effects the only focus of study? I would argue that some of the more important contributions of communication research are in a better understanding of what goes on, even without these “goings on” necessarily getting anyone anywhere. Intended effects or salient dangers play an important part, but there is much more to studying communication than just documenting what it actually does to people. I join in your assessment that the Net is, and will be, particularly fruitful in the study of indirect effects.

In mass communication research we have a tradition of studying uses and gratifications, that is, why people are engaged in this particular mediated communication or another, and what they get from it. Similar questions exist in the literature on interpersonal communication. In both cases, the spotlight is on the motivations (biological, psychological, sociological) that drive people to take part in receiving or exchanging messages. What are the uses and gratifications of Net use (e.g., Rafaeli, 1986)? What do we get from such use? What are the relative weights of prurience, curiosity, profit seeking, sociability?

We are likely to be reusing familiar paradigms and tools in answering these questions; however, more so than ever before, we will be asking about the uses and gratifications of providing information and of participating in an exchange. Why do people expend so much effort presenting themselves on the Web, creating and maintaining and updating home pages? Interaction, of course, is likely to play a major role in studying this. The Net and its use are likely to be the venue for a rejuvenation of the uses-and-gratifications type of study.

Second, we need to remember that the evolution of mediated communication rarely leads to extinction. We have had conversation, lecture, letter writing, storytelling, playing, acting, exhorting, defaming, creating-and we still have them. The Net will no doubt become one more place where these occur. Because media change faster than do our needs, the issue is not whether one medium will replace another, but of media and their contexts evolving in their roles.

Perhaps, in addition to worrying about how we study the effects of media on people, we can worry about the effects of our study on the media. There is nothing like the potentially interactive Net as an opportunity to do so. Some technologies, such as the telegraph, have withered. On the whole, however, niches in our communication ecology have developed, with media adapting to the niches more often than niches responding to media abilities. Will the Net create a new niche, or adapt to fill out previously existing needs? Herein lies a major use of the classification I suggested earlier. The five proposed concepts will, I predict, play a major role in figuring out the proper, effective, or likely niche for the Net in our communication diet.

Returning to the dilemma you raised of the vanishing direct effects: Is it possible that some of the postulated-but-not-documented effects are not there because the traditional media are still not very good at realizing their potential? We used to say that television is a medium because it is neither rare nor well done. Will the Net eventually be well done? That could be up to those of us in communication research. We have a role, as you pointed out, in holding up referees' score cards. We are also coaches in this particular field and track. For example, we will take part in the massive relegislation processes that stem from CMC, then be in the enviable position of studying what we have created :-).

Thinking about academia's role vis-à-vis the Net, we are reminded that what we call the Net today has roots in the Internet, Bitnet, and Arpanet, all partly academic institutions. Just at the point in history when critical voices speak of the decreasing relevance of research and universities, along come the Net and its attendant large-scale commercial, industrial, organizational, and social relevancies. In large measures, the Net can be considered an academic accomplishment. As you indicate, this alone behooves our involvement. Much of the morphology and culture of the Net, the practice of information exchange, and the very emphasis on information and the symbolic are all traditional academic messages. Furthermore, one set of strong direct effects (returning to your theme) is the massive macrolevel changes happening all around us as a result of the Net and its use. These changes are in the economic markets, the value of information and knowledge, the nature of work, the rewrites of entire legislative systems, to redefinition of international relations and sense of belonging somewhere, the definitions of nationhood, the sense of place, and much more. Perhaps more of communication research should focus on these macrolevel effects, and perhaps this is where a good cooperation among qualitative, critical and quantitative, empirical scholars can occur.

Newhagen

I think my point is that we, as a field, may find it useful to reconceptualize just what an “effect” is in the study of the Internet. Mass media research frequently conceptualizes media messages as independent variables and thinks in terms of how audiences are somehow changed by exposure to them. Interpersonal communication may think more in terms of media artifacts, that is, computers, telephones, or other communication appliances, and conceptualize them as variables intervening in the communication process. Media, however, are almost never conceptualized as dependent variables, whereby their use could be the effect caused by user-centric phenomena.

You mention uses and gratifications. That perspective gained a lot of attention about 20 years ago primarily because it did just that, it thought about how different motivations might lead to selective media exposure. Dusting off that perspective seems to hold some prospect for understanding the Internet because it addresses the problem of its mutability. That is, the Internet offers the user a broad range of communication opportunities: It can be used to visit an art museum (McLaughlin, this issue), to make friends (Parks, this issue), or for a host of other activities.

In any case, because of the Internet's chameleon-like character, I do agree that uses-and-gratifications offers a vehicle to lay out a taxonomy of just what goes on in cyberspace.

However, I disagree with your recommendation that the field should jump right in to a macrolevel agenda in its study of the Net. Doing so may result only in the same theoretical ambiguity that plagued uses-and- gratifications research in particular and mass media effects in general. To its credit, uses and gratifications did focus the field's attention on the audience, but it tended to do so in a peculiar way. Uses-and- gratifications work was pitched at an aggregate level of analysis, that is, audiences were conceptualized to have behaviors and needs that were resolved by selectively attending to media. That raises epistemological questions about knowing how or when an audience makes such decisions. For instance, it is frequently not clear when exposure was the result of conscious and deliberate selection and when it was not.

What I am suggesting is that we learn from the mistakes of the past, especially in regard to television research-and particularly, assuming a media technology has strong psychological effects, based largely on intuition, and then moving straight to societal-level policy discussions. Television has been blamed for making viewers fat, lazy, stupid, and violent, based on very little empirical evidence about key psychological concepts such as memory and attention.

It may already be too late: The popular press is already depicting the Net as having the power to snatch our babies right from the cradle and poison their minds. Based on virtually no empirical evidence, the power to mesmerize and seduce our youth is being attributed to the Net, just as it was to television and film before. Perhaps we should stand off from this discussion until we have a solid empirical understanding of the relationship between the individual user and the technology.

Reconceptualizing Access

I think this point can be illustrated by thinking about the concept of access. McChesney (in this issue) reviews the history of bandwidth allocation and regulation in terms of interactions among institutions, that is, corporations, government, and the public. This aggregate level of analysis seems appropriate to the discussion because access to television is not particularly interesting as a psychological problem. Receivers are inexpensive and easy to use, virtually everyone has one, and, until recently, viewers had littlecontrol over content.

Discussion of access to the Net, however, is problematic because of the inherent interactivity of its architecture. Whereas it may be possible for the corporate barons to turn the information superhighway into a toll road, it will be very difficult for th em to control the kind of traffic that streams across it. To the degree that this is true, the key issue concerning access may not have to do with the allocation of bandwidth, as it did with television and radio.

Will it make a difference who my Internet provider is in terms of how I use this technology? Examples of provider's attempts to control content, such as Prodigy's restriction on discussions of gay issues, are exceptions that prove the rule. As long as there is sufficient diversity of physical access, in the sense of being able to find a place to log on to the Net, Prodigy's position is quixotic. Some governments, such as those of Singapore and China, have attempted to control content and limit access to the Net. They may be successful, but in so doing, they will have to distort the technology to the point that it will no longer qualify as the kind of communication network we are talking about here.

For me, issues relating to the architecture of the Internet also are interesting. Even when the interface problem is solved, there will continue to be the issue of access related to the nature of an interactive communication system. The Net engages users in cognitively effortful tasks and challenges them to be active. My point is that we ought to take the time to understand what those challenges and barriers to access are for the individual user before we address some of the profound societal implications they may suggest. What cognitive skills will be required for full access? What effect will variable cognitive skills, such as the ability to solve problems or form searches, have on the ability to fully exploit the Net's potential? For political communication, this may shake out in terms of the study of empowerment, where media use is a behavior, much like attending rallies.

Finally, I agree with you that the Internet does not have to put newspapers and TV out of business for it to be truly revolutionary. I do not think there has ever been much empirical support for displacement theories. Whereas older systems may not go into instant extinction because of the Internet, they will be radically transformed by it.

Moments of transition allow students of media the opportunity to reconsider their most basic assumptions, gaining fresh insight into the old technology and setting the stage for understanding the new one. To some small degree, and for at least a few, we hope that this symposium will accomplish those ends.

Ancillary