Bridging Media: Computers and Human Communication


This paper projects impacts of prospective computer media on human communication based on a theoretical perspective that argues such prediction is possible given knowledge of a medium's characteristics. New media can be visualized by projecting them in a characteristic based "media space" of existing media. Existing computer media occupy a middle ground in this media space between broadcast, publication, correspondence, and telephone media. Projected new computer media, including hypermedia, multi-modal media, voice-into-text interactive media, and virtual reality media, appear to continue this general trend, expanding this middle ground more in the direction of film and art media and face to face communication as well as into an as yet unoccupied high bandwidth, high dynamism media space. The effect of computer media should then be to a range of new media applications and an expanding range of media choices that create a continuum of mass and interpersonal media.


For those who use computers as a routine element of their work, the impact of computer media on communication with others is increasingly obvious. For these individuals, the computer now mediates a large percentage of their daily interaction, in part because computer media facilitate contacts with people they might not otherwise communicate with or even know of. Foulger (3) finds, for instance, that, at least among users of IBM's IBMPC computer conferencing facility, computer mediated communication is becoming a dominant mode of interaction, with total use of electronic mail significantly surpassed only by face to face interaction. These results hardly represent a complete profile of media use. Survey participants are only asked about two computer media and ten media overall. Still, the media surveyed are certainly those one might expect to be the most important in an information intensive business environment like IBM.

Among these media, face to face interaction stands out as the most important, used significantly more than any other. Publications, television, and electronic mail stand together, all used at about the same rate, but used significantly more than any of the remaining media. Group meetings, computer conferencing, and the radio stand out as a second such cluster. The remaining media, including the telephone, correspondence, and lectures, respectively, each record significantly less use. Electronic mail in IBM is barely more than 15 years old. Computer Conferencing in IBM is just coming up on its tenth anniversary. Both make a strong showing relative to fairly well entrenched media, however.

MediumUsage Level by PercentageNMeanSt. Dev.
35 hours or more 17 to 35 hours 6 to 16 hours 1 to 5 hours Less than 1 hour
Lecture 0.0 0.0 1.7 22.0 76.3 177 1.25 0.47
Correspondence 0.0 0.0 2.8 31.1 66.1 177 1.37 0.54
Telephone 0.6 2.8 13.6 55.9 27.1 177 1.94 0.75
Group Meeting 0.0 0.6 24.9 59.9 14.7 177 2.11 0.64
Conferencing 0.0 2.3 24.9 57.6 15.3 177 2.14 0.69
Radio Hours 1.7 5.1 24.9 50.3 18.1 177 2.22 0.86
Electronic Mail 1.1 12.4 27.1 50.8 8.5 177 2.47 0.86
Television 1.7 7.3 43.5 33.3 14.1 177 2.49 0.89
Publication 0.6 6.2 53.1 37.9 2.3 177 2.65 0.66
Interaction 13.0 32.8 42.9 10.7 0.6 177 3.47 0.87

Use of ten media as reported by IBMPC participants: 1988 IBMPC user survey reports of the use of ten media. Media shown here ordered from least used (lectures) to most used (face to face interaction).

These results would not surprise Hiltz and Turoff (4), who have projected computer conferencing as a revolutionary medium of communication with the potential to make McLuhan's (6) "global village" an interpersonal reality (p. xxix). Indeed, they surpass Turoff's (7) expectation that computer conferencing will eventually "become as commonly used by the general public as the telephone is today." At least within IBM, both computer conferencing and electronic mail use significantly surpass telephone use. Hence it may well be, as Keisler (5) has suggested, that computer conferencing "has much in common with past technical innovations, like the telephone and the typewriter, that have had great social impact".

Indeed, other results detailed in Foulger's (3) study of one particularly successful computer conferencing facility in IBM indicate that the impact of these media inside IBM has been very substantial; arguably even revolutionary. Evidence for this impact can be found in the rapid growth of the facility in terms of conferences, participants, and contributions. It can be found in the evolution of its software to solve problems and meet the needs of its users. It includes a large, rapidly growing, and enthusiastic community of participants, a range of distinctive applications and a diverse set of associated genres of interaction. It includes a variety of formal and informal rules, a range of mechanisms for enforcing those rules, and means for negotiating such rules and their boundary conditions on an ongoing basis. These results do not appear, moreover, to be unique to computer mediated communications in IBM. Similar developments have been observed in other companies and on "publicly" accessible facilities like Prodigy, CompuServe, BIX, Usenet, Internet, and Minitel.

Keyboards are just the beginning

Still, today's keyboard-oriented computer conferencing and electronic mail are just two of a variety of existing and prospective computer media that may enhance and possibly change the ways in which people communicate with each other. As yet underdeveloped computer media, including hypermedia, multi-modal documents incorporating combinations of text, graphics, image, voice, video, and other presentation formats, voice-into-text concurrent interaction, and virtual reality, hold the promise of more radical changes yet to come. It may be that all of these possibilities will be successful. It is more likely, however, that one or a few will be hugely successful in developing a large audience and a range of applications, some will be moderately successful in niche applications, and most will fail.

This appears to be the case in one popular projection of our media future. Indeed, most of these alternatives appear to be largely irrelevant in "Star Trek: The Next Generation". Painting remains the pinnacle of art. Shakespeare remains the pinnacle of theatre. Face to face communication retains its dominance, aided by technology that allows people to "beam" from one place to another at the speed of some subatomic transport vehicle and technology that allows city sized starships to travel across the galaxy at warp speeds. When, for whatever reason, we cannot warp or beam across the space that divides us, we view each other on wall sized videoconference screens. The effect of computer communication in this 23rd century vision of human communication is seen in two venues. First, the computer acts as a repository for human knowledge; a huge multi-modal "memex" (Bush, 1947) that can find source material to answer almost any already-answered question almost instantaneously. Second, the computer creates the ultimate virtual reality in "holodeck" recreations of other realities.

This vision of human communication in the distant future is probably as reasonable as any that can be imagined right now. Shakespeare has weathered 500 years and seems likely to find continuing relevance in a distant future that will continue to value live performance. Although technologies for beaming people across space and eclipsing light speed are unlikely given our current understanding of the way the universe works, face to face communication will probably persist in its dominance of all other media. Something like a computer probably will replace libraries as the repository of accumulated human knowledge. Something like a holodeck probably will bring users into remote and alien frames, although perhaps not with the full sensory reality associated with Star Trek's holodeck.

Still, Star Trek's answer seems less than satisfying knowing the wide range of media that already exist and the early success of computer conferencing and electronic mail. The question, then, remains, what kind of impact will computer media like those suggested here have on human communication in the foreseeable future.

Making tracks through media space

Foulger (3) suggests a theoretical perspective and typology of media that may be useful in throwing insight on these questions. The perspective suggests a general process by which people create and evolve media to meet their needs; a process that is particularly obvious in computer conferencing but which is clearly evident in the evolution of many media. The success of a medium, according to this theoretical perspective, is a function of its relative costs and benefits to its prospective users, with key benefits measurable in the distinctive applications to which a given medium is applied. The perspective suggests that media with similar characteristics compete for the same applications, with one medium's success usually obtained at the expense of others.

Media, in this perspective, occupy positions in a "media space" that is directly analogous to the "animal space" proposed by Dawkins (2). There are an infinite variety of possible media (proto-media), but only a few succeed. Each is distinguished, in its success, by its ability to ply one or more "trades" or applications better than its competitors. Computer media, like all other proto-media, must succeed in the face of a variety of existing media, each of which is entrenched in its use for specific applications. Success, in the face of such competition, must come through either displacing other media from applications or creating new trades that do not, as yet, have competition.

Evidence for the validity of this perspective can be found in Foulger's (3) typology of 52 media, a representation of which is shown in the figure below. This representation shows six clusters of media, each of which represents a generalized set of trades. These clusters are depicted in a three dimensional space in which audience size occupies the vertical, with mass media found nearer the top and interpersonal media found closer to the bottom; bandwidth occupies the dimension extending to the right from this vertical, with higher bandwidth media found near the left wall and lower bandwidth media found in the right foreground; and dynamism (a combination of interactiveness and message persistence) extending to the left from the vertical, with more dynamic media found near the right rear wall and less dynamic media found in the left foreground.

The six clusters of media occupy distinct vectors through this multi-dimensional space. Face to face media are found deep in the media space, extending up from the bottom rear and intersecting, about midway across the bandwidth wall, with an almost perpendicular vector associated with telephone media. Correspondence media occupy the bottom left foreground. Television and radio media occupy the upper right foreground. Publishing media occupy the middle upper foreground. Film and art media describe a vector across the dynamism wall, attracting larger audiences as dynamism increases.

Foulger's (3) typology of media

What is interesting, in this depiction, is the relationship of these media to each other and, in particular, the relationship of computer media to these media. Specifically:

The first of these observations serve to support the perspective that media with similar characteristics do indeed compete for the same applications. The second supports the notion that the surest route to success for a medium is finding a way to avoid competition.

Computer media and the middle ground

With the observation that computer media occupy the middle ground, however, we begin to reach a better understanding of the role the computer plays in the media it participates in. Conventional assessments of the characteristics of computer media dwell on such characteristics as asynchrony, written form, transcripts, immediacy, shared memory, and the integration of the communication process with other activities (Turoff and Hiltz, 8). Perhaps the most important characteristic the computer brings to its media, however, is flexibility. Computer media can create a middle ground between other clusters of media because the mediators associated with computer media can be manipulated to highly specific ends.

Consider, for instance, computer conferencing. Open computer conferencing in the style Foulger (3) observed in IBM occupies the middle ground of the formal typology. One can, however, by simply restricting access to a select group of people, create a variant of computer conferencing that is very correspondence-like in its characteristics. By contrast, one can also create a very publishing-like variant by simply adding a person (an editor) that screen contributions before they are posted and schedule their distribution. One notes that these three variants of computer media form a nearly perfect vector across the media space. The changes entailed in this distribution are small. Indeed, all three variants can be implemented by enforcing different rules using the same software.

Many other variations of computer conferencing, electronic publishing, and electronic mail should also be possible with only small changes in the rules that govern existing computer conferencing software, the interfaces that are available for authoring and viewing computer mediated information, and the filters used in computer media to select, order, control, and systematically modify that information. The flexibility offered in these variations should be such that no computer media need never exactly match the characteristics or trades of an existing medium. It can be expected, then, that computer media will continue to find success in flanking actions that work to merge the characteristics of existing media in new ways.

Multi-modal documents

One such possibility is found in multi-modal documents (sometimes called multi-media or compound documents) that combine two or more presentation formats, possibly including text, graphics, image, sound, voice, and video in integrated documents. Such combinations are, of course, hardly new. Newspapers have integrated text, image, and graphics for over 100 years. Magazines, books, movies, television, and multi-media presentations have all combined such formats to differing degrees. Computer media remain relatively young, however, and while the computer has proven to be a valuable tool for producing newspapers, books, and other hard copy documents that combine text, voice, and image, it has yet to achieve any general solution to the problem of sharing multi-modal electronic documents.

The biggest problem for such documents is standards. There are only two generally used formats for computer text representation, the international standard ASCII format and IBM's widely used EBCDIC. These two representations of the English language character set readily translate to each other, moreover, making it possible to automatically translate documents from one format to the other with minimal losses or, more commonly, perfect fidelity. This relatively high level of standardization has made is possible for text based computer media like computer conferencing and electronic mail to become highly successful, at least within IBM.

The same cannot be said of graphics, image, sound, voice, or video storage formats, where a growing variety of often incompatible "standards" contend for dominance, or even for the ways in which these formats are joined to create multi-modal documents. Assuming such standards are agreed upon, and there are at least six major format groups in which agreement needs to be reached, there remain other problems to be solved. Text generation remains a fairly simple task on a computer, and recording voice and facsimile images is probably similarly simple. Generation of graphics, video, and sound, by contrast, remains a more complex enterprise. Each of these elements of multi-modal documents require both substantial storage and transmission capacities (which are just beginning to be realized) and major user interface solutions if they are to become widely used.

Computer graphics tools continue to improve, but are hampered by limited interface devices. Signing your name with a mouse or trackball is an almost impossible exercise that demonstrates this need quite graphicly. Video entails even greater complications, and unless one is willing to stay very still, taking a picture of oneself talking requires a second person operating the camera. Here again, interfaces can be improved with intelligent camera control computer software that can recognize the image it should track and then do so. If these problems aren't solved, multi-modal documents that incorporate graphics and, in particular, video will remain largely a mass production medium, more similar in its characteristics to books, newspapers, and Filmstrips than to interpersonal correspondence or telephone media.

Applied to the formal typology, the effect of multi-modal documents should be to increase the bandwidth associated with publication and correspondence media. This appears to be the case for the two multi-modal document media presented in the figure. Voice over data (text with voice annotation) is presented above, to the right, and somewhat behind correspondence media, indicating a movement toward both publishing and telephone media. These general directions indicate larger audiences, increased dynamism, and increased bandwidth relative to other correspondence media.

This appears to be what is happening in what is perhaps the most successful multi-modal media venture thus far, Videotex systems like PRODIGY (sm). Although neither correspondence nor daily news delivered by prodigy are multi-modal, many of its features and virtually all of its advertizing are. PRODIGY already has a readership that exceeds that of all but the largest newspapers and features highly dynamic multi-modal features and advertising.

Voice over mixed text and graphics appears to the right and somewhat behind publishing media stretching over in the direction of television. Mixed video, text, and graphics should move even further in that direction, clustering more closely with television and radio media than it does with publishing media. Indeed, combined with hypermedia structures, such a medium might make the closest approach to television of any medium yet devised, but with non-linear cross reference capabilities that should create a wealth of distinctive new applications.


These hypermedia structures describe a potentially non-linear mode of document organization that differs strongly from the usual format of books, newspapers, movies, television, and radio formats. Traditional documents, whether letters, books, newspapers, or movies, have a linear, sequential organization. One idea is typically presented at a time, with one following another in some structured manner. If two or more facets of the same idea must be presented, they are sequenced, with one shown before the others. Ideas may be presented together at the beginning of end of a sequence, but will be serialized in a specific author-determined order in the body of the document.

Hypermedia breaks this serialization by allowing a text to be organized in a non-sequential manner. A frame that introduces two ideas together will most likely point to two or more other frames that explain the ideas with additional detail. These frames may themselves point to additional frames of related or detail information. Except for the paths by which frames are joined, the order in which these frames are read, if at all, is left to the reader.

It is helpful, in visualizing hypermedia documents, to think of them as a sort of cave system in which passages diverge and intersect in several dimensions. Getting from the entrance to the cave to a river a quarter mile underground may require use of a series of passages and chambers, with each passage accessible by only a limited number of paths. If you are exploring, the cave system will allow you to stray from the path into interesting side passages. If you have a goal to achieve and know where you are going, you can travel quickly through the minimum number of passages.

It is clearly possible, in such a system, to build highly linear documents in which the side passages are mere footnotes. It is possible to build parallel serializations that occasionally intersect. But it is also possible to build highly convoluted knots of documents in which complex ideas express themselves in complex relationships. The flexibility associated with computers doesn't constrain hypermedia access to cave exploring. Indexes, tables of contents, and search engines can all be added to the viewer's interface possibilities, allowing them, in some sense, to "beam" straight into a given passage.

These structures change the way we view documents in important ways. They changes the whole vocabulary of, and quite possibly the authoring style associated with, reading and writing. Letters and books no longer have meaningful page numbers. They simply have parallel frames that cross-reference each other in much the same way different books and articles reference each other. Each frame is, in some sense, a complete mini-document, and the resulting macro-document can be traversed flexibly according to the reader's wishes. The effect of these changes should be to make documents appear more dynamic and to increase their overall content. The effect should be correspondence and publishing related media that stretch across the dynamism dimension of the formal typology in the direction of telephone and broadcast media, and across the bandwidth dimension in the direction of film and art.

This stretch should assure hypermedia-based computer media some measure of distinctiveness and may prove very challenging, in particular, to newspapers. This trend will only be accentuated if, as will almost certainly be the case, computer hypermedia incorporate graphic, image, and video multi-media document technology. One notes, however, that it remains to be seen if hypermedia composition can be easy enough for hypermedia to be used in any but publication contexts. The difficulties currently associated with compound document composition currently appear to give hypermedia correspondence applications a low probability of widespread adoption.

Voice-into-text concurrent interaction

Continuing progress in the area of computer voice recognition opens up possibilities for additional computer media including one that might be called "voice-into-text concurrent interaction" in which people talk to each other verbally but read a real time transcript of the text rather than listening. The major advantage of voice over text, given highly accurate transcription of voice into a textual transcript, is the relative speeds with which people can talk and write, read and listen. Other advantages can be found in reduced requirements for synchrony and turn taking and increasing opportunities for review relative to purely verbal (face to face and telephone) media.

The speed advantages are fairly obvious. Most people can talk at 140 to 150 words per minute with little effort but have difficulty creating new typewritten text at a rate better than 20 words per minute. By the same token, most people can read at a rate of at least 300 words per minute (and sometimes much more quickly), but are generally unable to listen at speeds much greater than 150 words per minute. In other words, where speed is the measure of a medium, voice is the preferred mode of text creation and reading is the preferred mode of text review.

This speed advantage is accentuated by reduced synchrony requirements, which allow two speakers to literally talk at each other continuously. When interacting via voice-into-text, there will frequently be no need for speaker A to wait until speaker B has finished before replying. Interaction, under these circumstances, becomes parallel and simultaneous in the manner that is usually associated with non-verbal interaction in face to face settings. The ability to review text easily may also enhance interaction speed, as the need to ask someone to repeat themselves (all too common in face to face and telephone exchange) will be reduced.

These advantages are clearly not obtained without cost. Much of the non-verbal dimension of face to face and, to a lesser extent, telephone verbal exchange will be lost in such exchanges, and computer substitutions for such non-verbal elements as pauses (...) and emphasis (boldface text) will be poor substitutes for this element of verbal interaction. Translation accuracy is also a potential issue, and voice-into-text may need to be augmented by an on-demand voice over text option that allows readers to occasionally review an associated voice transcript of the translated text.

Still, the advantages are potentially substantial, especially in contexts where the emphasis is on the sharing of information rather than affect. An individual might participate in several voice-into-text interactions simultaneously, work in a voice-into-text small group interaction environment, converse voice-into-text with one individual while interacting directly with another, etc. At extremes of translation, interactants might converse at some reasonable level without being able to speak a common language.

Viewed relative to the formal typology, one would expect media based on voice-into-text to fall somewhere in the gap separating telephone and correspondence media, but with an increased level of dynamism and information bandwidth that pulls it into unexplored media space in the general direction of face to face media. voice-into-text concurrent interaction can be expected, then, to provide the basis for one or more primarily interpersonal media that break new ground in media space. The result should be a truly unique mode of interaction that may well provide the basis for a whole new cluster of what might be called translation media.

Virtual reality

Perhaps the most ambitious goals for computer mediation of human interaction are found in recent efforts to create "virtual realities" via computer. Virtual reality is, in some sense, a covering term for a wide range of experiments in computer interfaces, including video goggles, motion sensing data gloves, and other technologies that attempt to bring an observer into a dynamic and high bandwidth frame. Star Trek's "holodeck" is the ultimate virtual reality fantasy in which one can create and recreate alternate realities at will, interacting with real and imaginary others in a world of sight, sound, touch, taste, and smell. We're a long way from this kind of complete virtual reality, but the first steps on this path are being taken.

These first steps are more oriented to presentation than interaction. One direction of such development attempts to bring the viewer more closely into the frame of a video game. A virtual reality golf game would substitute a real golf club for the keyboard or joystick of today's video games. The behavior of the ball would be based on the swing of the club, contact with the ball would create tactile feedback, and a large screen TV or video goggle would create the impression of actually being on a golf course. This kind of recreation is already found in arcade simulations of automobiles and motorcycles, and can be expected to be extended, over time, to encompass a range of games and game activities.

It remains too early to say very much about the future of virtual reality, which will probably emerge as a half or dozen or more distinct mass media, including an individual and small group gaming medium, one or more presentation media, and one or more publishing media. If adequate interfaces and transmission bandwidth are established, moreover, there may be a variety of interpersonal virtual reality media, including a remote small group discussion medium and a virtual reality face to face telephone that might allow two people (mom and daughter; grandparents and grandchildren, boyfriend and girlfriend, etc.) to interact under the illusion that they were in the same room, perhaps even with some level of tactile interaction extend the possibilities for remote telephone-like interaction into the face to face realm.

What is clear is that virtual reality will break new ground in media space. by occupying the high bandwidth, high dynamism space above face to face media. Traditionally, new ground in media space represents new opportunities, as-yet-unimagined applications, and breakthrough media that revolutionize the world in unimaginable ways. Virtual reality, more so than any of the other media described here, appears to have this potential for producing revolutionary media that change our world in fundamental ways.


This paper attempts to project the kinds of impact computer media will have on human communication in very general terms, based on a theoretical perspective that argues such prediction is possible and typology of media that creates a media space in which it is possible to project other media. Space limitations necessarily preclude anything more than a cursory analysis, restricted to general directions rather than specific applications and effects. It is hoped, however, that these directions have been presented in a way that will allow the reader to imagine the kinds of things these new media may make possible.

It is argued here that the early success of electronic mail and computer conferencing in the face of a wide range of existing media is just a beginning. When projected in a multi-dimensional media space today's relatively low-bandwidth text-based computer media cluster with a variety of existing media, including telephone, correspondence, and publishing media. Interestingly, however, they form what can be thought of as a "seventh" cluster of media, a middle ground between existing clusters of media that unifies these clusters and creates new opportunities for mass interpersonal communication and more interactive correspondence. This middle ground gives computer media a distinctiveness, relative to other media, that probably helps to account for the increasingly apparent success of electronic mail, computer conferencing, and electronic publishing.

New computer media, including multi-modal media, hypermedia, voice-into-text concurrent interaction, and virtual reality, appear to have characteristics that may make them even more distinctive. Multi-modal computer media should increase the bandwidth and, to a somewhat lesser extent, the dynamism associated with correspondence and publishing media, thus extending the possibilities associated with these media in the direction of both film and art media and television. Hypermedia should have a similar effect in increasing the dynamism and, to a lesser extent, the bandwidth, of these same media, thus pulling these clusters in the direction of television and telephone media. Voice-into-text concurrent interaction should increase the dynamism and, to a lesser extent, the bandwidth and audience sizes associated with telephone and correspondence media, thus stretching these clusters in the direction of face to face interactive media. Finally, the high bandwidth and dynamism associated with virtual reality promise to stretch the limits of face to face interaction and create a new cluster of interactive mass media in the empty space above face to face interaction.

The effect of these media, broadly stated, will first be to expand and fill in the computer mediated middle ground between existing clusters of media, and then to explore new ground. The effect of hypermedia, multi-modal media, and voice-into-text will be to expand this middle ground to encompass broadcast media (including computer mediated television and radio variants), film and art media, and, to a limited extent, face to face interactive media. Complexities of multi-modal and hypermedia production argue that such media will, at least initially, more likely be oriented to mass audiences. The general structure of voice-into-text argues that it will produce interpersonal media. Virtual reality, by contrast, promises to expand into the unexplored regions of highly dynamic, high bandwidth media space above interpersonal media.

It can be argued, based on this analysis, that the role of the computer in human communication is one of an integrator. Where there were once clearly interpersonal media that were obviously different from what were clearly mass media there is now a growing continuum that allows any individual to interact, almost as if on a one to one basis, with a large audience. The effect of this growing range of media will be an increasingly complex media environment in which individuals will have many choices depending on the kind of interaction they want to have or are constrained to having, the kind of message they want to deliver or receive, and the kind of audience they want to reach or be a part of.


  1. Bush, Vannevar. As we may think. Atlantic Monthly, July 1945, 101-108.
  2. Dawkins, Richard. The Blind Watchmaker: Why the evidence of evolution reveals a universe without design. New York; W.W. Norton and company, 1986.
  3. Foulger, Davis. Medium as Process: The structure, use, and practice of computer conferencing on IBM's IBMPC computer conferencing facility. (Doctoral Dissertation, Temple University, 1990). U.M.I. Dissertation Services, Order Number 9107898, 1991.
  4. Hiltz, Starr Roxanne, and Turoff, Murray. The Network Nation: Human Communication via Computer. Reading, MA: Addison-Wesley, 1978.
  5. Kiesler, S. The hidden messages in computer networks. Harvard Business Review, January-February, 1986, pp. 46-60.
  6. McLuhan, Marshall. Understanding Media: The Extensions of Man. New York, McGraw Hill, 1964.
  7. Turoff, M. The Anatomy of a Computer Application Innovation: Computer Mediated Communication (CMC). Technological Forecasting and Social Change, 1989, 107-122.
  8. Turoff, M. and Hiltz, S. R. Computer Mediated Communications and Developing Countries. Telematics and Informatics, 1987.

October 21, 1992

Davis A. Foulger