This chapter seeks to construct a more formal typology of media using the techniques of numerical taxonomy (see methodology). For this study, 52 different media were compared, using cluster analysis and factor analysis, on 12 different characteristics of media. The result of this analysis is a typology of media which contrasts computer conferencing and other computer media with a broad array of interpersonal and mass media, including newspapers, magazines, movies, radio, television, letters, face-to-face communication, and the telephone.
Six broad classes of human communication media are revealed in the cluster analyses. Three of the clusters represent interpersonal communication media. Three represent mass communication media. The characteristics that define each, and differentiate them from each other, are intuitively obvious. Indeed, the six classes are what one might expect, given the divisions which permeate the field of communication today. Still, the results are not entirely clear-cut. The two cluster analysis methods have produced two results that, as might be expected, are highly similar. There are points, however, at which the two analyses have produced strong contrasts. Indeed, each of the analyses can be argued to clearly reveal only five clusters, with a sixth cluster partially subsumed within another cluster. Each analysis hides a different cluster, however, and in one case the pairing is moderately surprising. There are other interesting differences between the results of the two analyses, as we shall see.
A cluster representing face-to-face communication media is clearly apparent in both analyses. It includes the three "broad bandwidth" interpersonal communication media that were identified, in our description of different communication media, as traditional communication media. This cluster is centered by face-to-face interpersonal communication, but also includes Small Group Communication, and Public Speaking.
One might expect this cluster to end here, but it doesn't. It also includes one or two technologically mediated human communication systems. Both cluster analyses place one of these systems, video conferencing, in this cluster. One analysis, the Ward method, also identifies the video telephone as a member of this cluster. These should not be surprising additions. Of all the technologically mediated and computer mediated communications systems that can be used for interpersonal communication, the Video Conference and Video Telephone offer the highest personal bandwidths, with both visual and auditory cues available to the target.
The telephone has, over the course of the last 100 years, revolutionized the way we communicate with one another, making it possible for people to interact with others quickly and easily, even when they are widely separated. It is basically a single channel variant on face-to-face communication, marked by high speed, low persistence, verbal messages. Several other forms of communication, including the intercom, Citizens Band Radio, and Ham Radio, are very similar to telephone communications in these essentials, differing from the telephone only in terms of their distance constraints, the way messages are switched, and their transmission medium. A fifth medium, teleconferencing, differs from typical telephone conversations only in the number of people involved, and the way the floor is controlled during conversation.
It is not surprising that these media cluster together to form the cluster we call "Telephony". They are highly similar media that present a strong contrast to the characteristics of other media. This basic cluster is joined, in both analyses, by a sixth communication medium, the electronic quip. Although written, this medium matches the general profile of other media in this cluster (high speed, low persistence, verbal messages). It can, moreover, be used conversationally.
Telephony is a small, tight, and fairly obvious cluster, and the electronic quip makes an interesting addition to this core group, but the most interesting aspect of the telephony cluster is its edges. In one analysis (Average Linkage Between Groups), this core set is joined by the Video Telephone, which groups with face-to-face communication in the Ward clustering. In the other analysis (below), it is weakly paired with Broadcast media.
Both edges are interesting, because the media that form the telephony cluster are all interpersonal variants of broadcast media. Indeed, the telephone was marketed, in it early years, as a broadcast medium that would allow current events to be brought directly to people everywhere (Pool, 1977). Hardly anyone imagined the telephone as a home appliance at that time. The aim was to place a telephone in every town, and the Western Union's participation in the early stages of the telephone industry reflects the perception of the telephone, at that point in time, as a broadcast vehicle.
Other media in the telephony cluster accentuate that relationship. A C.B. Radio, for instance, is really a low power radio station, and a Ham Radio isn't even low power. Both involve two way radio equipment, however, and this allows both to be the interpersonal communication media that, at root, they are. Electronic quips can take on a broadcast quality as well, and they are occasionally used to quickly transmit messages to large numbers of computer users.
The telephone evolved into an interpersonal communication media, however, and all of the media in the telephony cluster are basically interpersonal media. Here, the Video Telephone provides an interesting second edge. The connection formed by the VideoPhone is an important one, as it establishes a continuum between the telephony cluster, which encompasses a number of single channel interpersonal communication media and the face-to-face cluster, which encompasses several multi-channel interpersonal communication media. The number and nature of these channels provide the biggest differences between face-to-face communication and telephony.
Radio, Talk Radio, and Television, all technological media that have emerged during this century, form the first of several mass media clusters. This cluster seems fairly obvious, as these media have a great deal in common. All involve the broadcasting of messages, on the part of a few people, to many people. All deliver low persistence messages to the target quickly. In the Ward analysis, broadcasting clusters weakly with Telephony, but it is fully independent in the Average Linkage Between Groups analysis.
A third cluster of basically interpersonal human communication media might be called "correspondence". This cluster is clearly apparent in both analyses, and, although an additional four items cluster with it in the Ward analysis, a core set of fourteen media are shared by both analyses. The traditional media in this cluster include letters, notes and memos, which were discussed as a group when we were describing media. All are written interpersonal communication media.
Several technologically mediated communication media also belong to this cluster, including Telegrams, Telex, Tape Letters, and facsimile. None of these are surprising additions. Telegrams and Telex messages are really nothing more than high speed letters, and this characteristic hasn't changed since their inceptions. Indeed, one variant on the Telegram is delivered by the post office. Tape letters are really nothing more than "voice" letters. It turns out that, at least in terms of the characteristics used here, they do not differ significantly from written letters. Finally, facsimile is an image oriented variant on Telex, in which an image is copied between two distant points at high speed.
The real core of this cluster, however, are a set of related computer media, including the following:
It seems obvious that Electronic Mail and Voice Mail would belong to this cluster. Both assert themselves, as forms of "mail", as correspondence media, and their characteristics are much like those of letters, telex, telegrams and other media in this cluster. Voice over Data is, to a large extent, a combination of these forms, and is a highly sensible addition to the cluster.
Asynchronous and Computer Synchronized Conversation declare themselves as conversational media, however, and, before performing the cluster analysis, one might have expected them to cluster with face-to-face communication or, at the very least, with a more instantaneous form of interpersonal communication like the telephone. The same can be said, to a lesser extent, of closed computer conferencing.
Clearly, however, these media have a great deal in common with the other forms of correspondence clustered here. Like most of the correspondence media, computer conversation and conferencing are written media. Unlike telephony and face-to-face communication forms, moreover, these media have very high persistence. And it should be noted that any of the correspondence forms that have been listed here can be used to conduct a conversation at some speed. Hence the clustering of computer conversation and closed computer conferencing with correspondence forms is a sensible one.
It should be noted, moreover, that the forms of computer conversation, along with cooperative composition, form, in both analyses, a distinct, if moderately weak, cluster that could be split off from the other forms of correspondence. It is this clustering of cooperative composition that is, perhaps, the most interesting result of the cluster analyses thus far. Of all the computer media that have been introduced in this chapter, cooperative composition is probably the most innovative, as it provides for a form of communication that cannot even be imitated well except without the use of a computer. As a highly interactive medium that allows people to write and edit documents conversationally, its clustering with computer conversation makes sense. As a written medium, its clustering with correspondence makes sense, but it would have been difficult to predict its positioning among media before performing the cluster analyses.
The only other real surprise so far is the failure of the various forms of computer conferencing to cluster together. This is actually only half true, however, as one of the cluster analyses, the "Average Linkage Between Groups Method", adds another four computer media to the correspondence cluster.
These four media form a fairly strong cluster on their own. Using the Ward method, these media cluster with other "published" media. Using the Average method, they appear as correspondence. The proper location is probably somewhere in between, and these media can be argued to straddle the line between correspondence and print media.
Taken together, the media clustered together as correspondence have a great deal in common. Most are written media, and all can be symbolic media. All have fairly long persistence. All can be interactive, and most can be conversational. The common thread among these media is the letter, whose roots can be traced into all of these media in one way or another.
If one were to construct a list of conventional print media, books, newspapers, newsletters, journals, and magazines would lead the list. This is exactly what occurs in the highly predictable Publishing cluster, which also includes audio recordings, video recordings, and several computer media. Two computer media, both of which were described together under the name "compound documents", stay with this cluster in both analyses. Four others, including Open Computer Conferencing, Electronic Bulletin Boards, Electronic Publishing, and On-line Information, are included with the publishing cluster in one analysis (Ward Method), and are counted with correspondence media in the other analysis.
There are no surprises here. Audio Recordings and Video Recordings are published and marketed in much the same way books are. The preparation time for computer-mediated documents that involve the use of both text and graphics make them improbable (given current interfaces) interpersonal media. Among the borderline members of this cluster, open computer conferencing is probably the least publishing-oriented, but because messages in a computer conference reach tremendous numbers of people, a a contribution to a computer conference can be easily seen to have both publishing and correspondence qualities.
The only other interesting thing to note about this cluster is its relationship to the sixth cluster in the analysis, Film and Art. In one analysis (Ward method) these clusters are fully distinct. They are harder to separate in the Average Linkage Between Groups analysis, however. Publishing mediates the two distinct subclusters that make up the Film and Art cluster in that analysis, and the distinctions between publishing and Film and Art will only be muddied when we apply the results of the cluster analyses to the factor analysis later in this chapter. On the whole, however, this ambiguity is not unreasonable. It may stretch the definition of publishing a bit, but an artist or art studio can be viewed as "publishing" a painting, and a film studio can be seen as "publishing" a film.
The film and art cluster is built from two distinct subclusters. The first of these subclusters ("Art") includes the following media:
The second subcluster ("Multi-media" or "film") is composed of:
Film and art is perhaps the least obvious of the clusters revealed in the two cluster analyses, a fact that is underlined in the cluster's ambiguous linkage with publishing in the Average Linkage analysis. The media in this cluster are varied. Some engage multiple channels. Others use only one. Some reach audiences quickly. Some more slowly. Some involve verbal and written messages. Others do not.
There is, nonetheless, some strong common ground among these media. All involve visual nonverbal elements. All have high persistence. All require the target to travel to the message. None of these elements, taken alone, will do much to differentiate these media from other media. Taken together, however, they built a distinctive cluster that makes sense. The subclusters make sense as well. The "Art" subcluster describes a series of single channel, visual, non-verbal communication media. The "Multi-Media" subcluster describes a series of multi-channel extensions of the media in the "Art" subcluster.
These six clusters form an intuitively valid description of the field of communication. Indeed, four of the clusters, face-to-face Communication, Publishing, Broadcasting, and Film describe traditional divisions in the field of communication. Journalism (publishing), Broadcasting, and Film are most frequently studied from within departments of Mass Communication. Mass Communication and Interpersonal Communication Media do differentiate into several clusters, and those clusters can be clearly classified as interpersonal media clusters and mass media clusters. Correspondence and telephony form sensible clusters, but do not represent traditional areas of study for anyone but the telephone company. When they are treated at all, they are generally treated as special cases, and they are most frequently ignored altogether. It is in these clusters that computer mediated communications systems are making their presence felt most strongly.
The obviousness of the clusters, and the extent to which they map traditional divisions in the field of communication, is important, as it it lends considerable credence to the usefulness of the cluster analysis in mapping computer media to preexisting media. The edges between the clusters may be equally important, however, as they provide us with a perspective on how media fit together. Several edges revealed themselves in the cluster analyses. The less interesting ones showed that Telephony was strongly related to face-to- face communication; that Film and Art can be regarded, to a certain extent, as publishing media. The more interesting ones showed overlaps between interpersonal and mass media that are often ignored within the varied divisions of the study of communication.
One of these edges ties Broadcasting and Telephony, two classes of communication that have common roots and contrasting destinies. The early history of the telephone as a broadcast medium is frequently forgotten in today's highly interpersonal telephone environment. The edge is a tenuous one, weak by any measure, but a useful one to remember, especially as we look at a new generation of media whose most important uses probably remain at the edges of, if not beyond the range of, our imaginations.
The more interesting edge, from the perspective of this study, is the link that computer conferencing creates between the worlds of correspondence and publishing. A large open computer conferencing facility like IBMPC straddles the fence between correspondence and publishing exquisitely. Anything said on IBMPC is effectively published for thousands to see, yet open to immediate dialogue. The ability of computer media to straddle the edges between traditional classes of media is an important one.
This ability of computer media to span classes of media is not, however, the most important thing revealed in the cluster analysis, however. What is important is that unlike the telephone, the radio, or the printing press, each of which effectively created a new class of rather homogenous media, the computer has not provided the basis for a new class of computer media. Instead, the computer media that have been discussed here have distributed themselves around several existing classes of human communication media, and as a group, they cannot be labeled as either mass communication media or interpersonal communication media. Individual computer media fall on both sides of the line, and in some cases, they straddle the line.
The most surprising result found in the factor analyses is also the least surprising result: with all the new media that have emerged in the last 150 years, the underlying dimensions of communication haven't. In each of three factor analyses, one of traditional media (see the defacto classifications in the listing of media), one of traditional and technological media, and one of traditional, technological, and computer media, three factors are revealed. When those factors are matched up, side by side, they are fundamentally the same.
This result is surprising because of the large numbers of new media that have emerged in the last 150 years. Three of four media in this analysis are technologically mediated or computer mediated, and some of those media combine characteristics in ways that simply weren't possible for traditional media. There was, for instance, nothing like broadcast radio and television before technology made "high speed" mass media possible. Distributing information to large numbers of people always took time. Similarly, there was no such thing as "fast correspondence" before electronic mail or, perhaps, telex.
These new combinations of characteristics, when combined with large numbers of new media, could have radically altered the underlying dimensions of communication media. This has not happened. There have been some shifts over time in the relationships between some of the characteristics studied here. One can argue on the basis of the results of the factor analyses, moreover, that there has been a shift in the relative importance of the underlying dimensions. It seems clear, however, that these shifting relationships have not changed the three underlying dimensions of communication.
The first of these factors cleanly differentiates mass media and interpersonal media. Its most powerful members are the measures of audience size and potential audience size. These characteristics, along with Nonverbal Bandwidth, define this dimension in all three analyses. The table below shows the Interpersonal and Mass Communication dimension of human communication as it appears in each of the three factor analyses.
| Characteristics | Traditional (14 Cases) Factor 3 | + Technological (38 Cases) Factor 3 | + Computer (52 Cases) Factor 1 |
Potential Audience Size | .79945** | .93398** | .98099** |
| Audience Size | .86841** | .97392** | .97739** |
| Interactivity | .23369 | .63910** | .71079** |
| Speed | .28662 | .54028* | .69562** |
| Theoretical Speed | .28254 | .36614 | .55541** |
| Nonverbal Bandwidth | .81355** | .42115* | .48379* |
| Persistence | .18900 | .21472 | .42386* |
| Personal Bandwidth | .26046 | .08063 | .18654 |
| Theoretical Persistence | -.14532 | .01014 | .15378 |
| Verbal Bandwidth | -.12100 | .08304 | .07358 |
| Distance | .19784 | .08655 | .05095 |
| Travel | .09861 | -.01180 | .00484 |
Several things should be noted about this dimension:
The emergence of higher speed correspondence-like media, starting with the Telegram, and accelerating with Telex, Electronic Mail, Computer Conferencing and other Computer Media, has changed the landscape, however. Although correspondence media are still slower than face-to-face media when both groups are taken as a whole, it is no longer reasonable to draw a line between face-to-face media and correspondence media and say that the members of one group are always slower or less interactive than the members of the other. Mass media remain consistently slower and less interactive than interpersonal media, however. Hence the shift in the relationship of speed and interactivity to the dimension.
Before the emergence of technologically mediated communication systems, human communication was the subject of discussion in the study of both literature and rhetoric. Indeed, the study of mass media and interpersonal communication as distinct areas of scholarly research is, like the emergence of the Interpersonal and Mass Communication dimension as a strong dimension, a comparatively recent development. This coincidence is notable because it may provide unusual evidence concerning the origins of the shift from studying communication as a rhetorical event to studying communication as an interpersonal or mass media event. This is particularly evident when one considers the next dimension of communication, which has slipped from a former position of dominance.
Another dimension revealed in the factor analyses might be called "Dynamism". It is represented by several characteristics which, on the surface, seem rather diverse, but which share a deeper common thread. The most powerful element of this dimension in the analyses is persistence, but personal bandwidth and, to a lesser extent, speed, also load on this dimension.
The common thread that runs through all of the elements is a contrast between static media and dynamic media. Dynamic media take advantage of more communications channels (Personal Bandwidth) with messages that move more quickly, but which are not always easy to preserve. Static media produce messages that are generally preserved in the very act of creation, but that only utilize one or two communications channels.
As can be seen in the table below, the relationship of some characteristics to Dynamism has changed considerably with the introduction of new communication media. Speed, for instance, has been reduced from a primary element of the dimension to a secondary role, and Interactivity has been reduced from a primary role to no role at all. A rationale for these changes has already been forwarded in the description of the Interpersonal and Mass Communication dimension. Among traditional media, correspondence media are slower and less interactive than face-to-face media. With the emergence of technological media and computer media, however, this is no longer the case.
| Characteristics | Traditional (14 Cases) Factor 3 | + Technological (38 Cases) Factor 3 | + Computer (52 Cases) Factor 1 |
| Theoretical Persistence | .95306** | .94251** | .84790** |
| Persistence | .91268** | .85374** | .71787** |
| Personal Bandwidth | -.95716** | -.47697* | -.65512** |
| Theoretical Speed | .83862** | .74602** | .49797* |
| Speed | .85492** | .61818** | .47239* |
| Verbal Bandwidth | -.60800* | -.59247** | -.45577* |
| Interactivity | .91119** | .46969* | .33536 |
| Nonverbal Bandwidth | -.52871* | .00924 | -.26442 |
| Audience Size | .29885 | -.02970 | -.08626 |
| Distance | .26631 | -.16623 | .08544 |
| Travel | -.30264 | .16037 | -.08365 |
| Potential Audience Size | .39212 | .03702 | -.05034 |
It is also worth noting, given the related discussion of the Interpersonal and Mass Communication dimension, that the Dynamism dimension provides a fairly good description of the forces that divide the study of rhetoric from the study of literature. Rhetoric has traditionally focused its attention on public speaking and other dynamic interpersonal communication media. Literature, on the other hand, has traditionally looked to books and correspondence, both of which would be considered static on this dimension, as objects of study.
The last dimension revealed in each of the analyses can be regarded as representing the bandwidth of various media. All of the measures of Bandwidth used in this analysis emerge on this factor, and the two primary characteristics of the dimension, travel and distance, both reflect the effects of Bandwidth.
| Characteristics | Traditional (14 Cases) Factor 3 | + Technological (38 Cases) Factor 3 | + Computer (52 Cases) Factor 1 |
| Travel | .82712** | .90254** | .92523** |
| Distance | -.83606** | -.82638** | -.89190** |
| Verbal Bandwidth | -.88443** | -.33940 | -.64331** |
| Nonverbal Bandwidth | .27217 | .67295** | .58467** |
| Personal Bandwidth | .01803 | .73223** | .45135* |
| Theoretical Speed | -.12322 | .02719 | .24376 |
| Interactivity | .13822 | -.01300 | .15758 |
| Potential Audience Size | -.01695 | .06505 | -.07995 |
| Audience Size | -.13746 | .00609 | -.07274 |
| Speed | -.09102 | -.05740 | .07091 |
| Persistence | -.07520 | .00006 | .03991 |
| Theoretical Persistence | -.12275 | -.07814 | .01328 |
Communications bandwidth is expensive and difficult to obtain. As a result, it is generally difficult to transmit messages long distances without sacrificing channels, channel capacity, and/or speed. Indeed, certain elements of the Personal Bandwidth measure used here defy transmission over distances much greater than a few feet. There is no known means of communication for providing a sense of touch or smell over long distances, and we are only beginning to master the technologies necessary to transmit high resolution images into the home. As a result, long distance communication generally means low bandwidth, and very high bandwidth generally requires that the target travel to the message.
The factor analyses have provided us with a picture of some of the forces that differentiate human communication media. Some media work best for delivering simultaneous or near simultaneous messages to large numbers of people. Others work best for allowing people to interact in small numbers. Some media provide very high bandwidth, but extract distribution costs, while others provide very low bandwidth, but can travel large distances. Finally, some media are very dynamic, while others are somewhat static.
This simple three dimensional structure does not capture all the rich possibilities that differentiate human communications media. A good case can be made, however, that these dimensions capture some important underlying themes in the differentiation of media. The same dimensions appear when the data set is reduced to traditional communication media. The same dimensions appear when technological media are added to traditional media. Incidentally, the same dimensions also appear if one forces a 3 factor solution after removing the characteristics that, by definition, create the highest preanalysis correlations (Potential Audience Size, Theoretical Persistence, and Theoretical Speed).
Finally, the dimensions are sensible. Bandwidth, dynamism, and the split between mass communication and interpersonal communication media are constantly recurring themes in the study of communication. Clearly, the three dimensions seem to reveal something about the basic forces underlying communication media.
For the current purpose, however, they need not make any sense at all, for our primary objective in performing the factor analysis has been to create a multidimensional space in which we can contrast the varied forms of communication that were identified in the cluster analysis. It is nice that the dimensions are both highly interpretable and sensible, as it makes the job of describing the fit much easier, but it is hardly a necessity. So long as the relationships between clusters can be explained within the context of the dimensions, the factor analyses have served their purpose.
We will explore these relationships by examining the positioning of media within the three-dimensional space revealed in the factor analysis. In an effort to make the positioning and subsequent comparisons more comprehensible, we will not consider elements in all three dimensions simultaneously. Instead, we will draw three pictures, taking the dimensions two at a time.
+---+----+----+----+----+----+----+----+----+----+----+----+----+-+
+ | *Television |
| |
| Public Speaking* *Radio |
| |
| *Face-to-face Ham Radio* *Talk Radio |
+ *VideoConference +
| |
|Small Group* C.B.* *VideoPhone |
| |
| |
+ *Telephone +
| *Intercom *Teleconference |
| |
| |
| |
D + *Electronic Quip+
Y | |
N | Voice over Mixed Text & Graphics* |
A | |
N | *Newspapers |
I + +
S | |
M | *Weekly Newspapers |
| |
| On-line Information & |
0+ Electronic Publishing* +
| Magazines |
| Mixed Text & Graphics* |
| Computer Synchronized Conversation* Open Computer|
| Asynchronous Computer Conversation* *Confe|
+ Film Voice Mail Video Rec+
| Film with Subtitles * Coop Comp** *Elec B B|
| Multi-Media**Talking Autom Recordings* *Boo|
| Note* *B.Films Electronic *Voice/Data |
| Filmstrip* Mail/Telex* *Newsletters |
+ Facsimile* *Closed Computer+
| Journals* Conf|
| Silent Film* *Automatons *Telegram |
| Dioramas*Art *Memos |
- | |
+ Letters & Tape Letters* +
+---+----+----+----+----+----+----+----+----+----+----+----+----+-+
- BANDWIDTH 0 +
In our first comparison, we will contrast Dynamism and Bandwidth. The above figure provides us with a picture of this comparison, and the reader is encouraged to refer to this figure in the discussion that follows.
Opposing Dynamism and Bandwidth does not provide for a great deal of insight concerning either the differences between media or the role of computer media among other media. Hence the above plot is probably the least interesting of the three figures that will be presented here. Three clusters -- face-to-face communication, telephony, and broadcasting -- can be clearly differentiated in this picture. Face-to-face communication occupies the upper left hand corner of the two dimensional space. Broadcast media occupy the upper right hand corner of the space. Telephony occupies the upper middle ground, with its close connections to both Broadcasting and face-to- face Media apparent. All are highly dynamic media which are primarily differentiated by their relative bandwidths.
The other three clusters are hopelessly intermingled in the lower left corner of the picture. Some differentiation is possible, as Film and Art Media are separated from Publishing Media, but correspondence media mix it up with both, and muddy the water considerably. The same is true when one examines computer media within the space, and the only real news is that computer media share the characteristics of other media. This is not, of course, news, as computer media vigorously resisted forming independent clusters in the cluster analysis.
Ultimately, there are only two things that are really interesting in this pairing of dimensions. One is the clear relationship between broadcast media and telephony. A weak relationship between the two was revealed in the cluster analysis, and among the pairings of media dimensions that will be examined here, this is the only pairing which shows a rationale for that relationship. The common ground between these clusters is their dynamism and their relative bandwidth.
The second interesting revelation is the verification of the interpretations of all of the factors that is implicit to this pairing. Although both Bandwidth and Dynamism individually have elements that should differentiate mass communication from interpersonal communication, these elements express themselves, when bandwidth and dynamism are opposed, as an interaction effect. Face-to- face communication occupies the upper left hand corner of the space. Correspondence occupies the lower right hand corner of the space. Broadcasting occupies the upper right hand corner of the space. Film and Art are located near the lower left hand corner of the space. Clearly, neither of these dimensions captures the difference between interpersonal media and mass media.
The reasonableness of individual dimensions can be observed as well. Bandwidth moves from left to right at a slight downward angle. The highest bandwidth media, face-to-face communication, is all alone on the left hand side. The numerous single channel media crowd against the right side. Dynamism moves from top to bottom, with highly dynamic face-to-face and broadcasting media at the top, and rather static written media at the bottom. Neither of the dimensions works perfectly. One can easily pick pairs of media which seem to have the wrong relative positions, but this is more a by-product of the insensitive measures used than a reflection of the quality of the interpretation of the dimensions. Taken as a whole, the interpretations work very well. The exceptions, which would generally occur anyway, simply underline the value of taking a statistical approach to typology.
+----+----+----+----+----+----+----+----+----+----+----+----+----++
+ | Television*|
+ +
M | |
A | |
S | |
S | Books |
+ * Film +
M | Recordings**Video Recordings Voice over Mixed |
E | Talking Automatons *Text & Graphics |
D | *Automatons Radio* |
I | *Film *Magazines *Newspapers |
A + Art* w/subs *Mixed Text & Graphics +
| Dioramas * *Weekly Newspapers |
V | * *Multi-Media Electronic Publishing |
S | Journals & On-line Information *Talk |
| * *Beeping Filmstrip Radio |
I + Silent Film +
N | *Filmstrip |
T | * *Open Computer Conference |
E | Newsletter *Electronic Bulletin Board |
R | |
P0+ +
E | |
R | |
S | |
O | Voice |
N + over* *Closed Computer Conference Ham Radio* +
A | Data Public Speaking* |
L | Electronic Quips* *Video |
|Letter & *Voice Mail C.B.* Confer.|
M |Tape Letter |
E + * Telegram +
D | * * *Facsimile TeleConference* *VideoPhone |
I |Memo **Telex *TelePhone |
A | Electronic Mail |
| *Asynch Computer *Face- |
+ Cooperative Conversation Intercom* * to-face+
| Composition* *Computer Small |
- | Notes* Synchronized Conversation Group |
+----+----+----+----+----+----+----+----+----+----+----+----+----++
- 0 DYNAMISM +
Contrasts the Dynamism dimension with the Interpersonal and Mass Communication dimension, is much more interesting. Indeed, it fairly explodes with useful information. To begin with, it provides a real key to understanding the basis of the relationships between clusters that were revealed in the cluster analyses.
We find mass media in the upper half of the space, with interpersonal communications media in the lower half of the space. This is not a casual division. It is a concise division that brooks no exceptions. If one assumes the clusters revealed in the Ward analysis, every mass medium is in the upper half of the two dimensional space; every interpersonal media is in the lower half of the space. If that were not sufficient emphasis, however, it should not escape notice that the dividing line between the two halves is the value zero. This is the cleanest division between two classes of objects I have ever seen in a factor analysis, and it bears notice.
The division even works in relative terms. The cluster of computer media that switches allegiances in the two cluster analyses, with a commitment to publishing in one analysis and correspondence in the other, is close to the line that divides interpersonal and mass media. Indeed, two of its four elements are closer to that line than are any other medium. Talk Radio, the most interactive of broadcast media, is the closest of those media to the dividing line. Public Speaking and VideoConferencing, both of which have mass media aspects, are close to the middle line as well. Indeed, it appears that one can pick any two points in the space and, in comparing their relative position from top to bottom, justify their relative position on the continuum from pure interpersonal media to pure mass media.
The dynamism dimension is clearly visible from left to right, as well. Dynamism should not be regarded as a straight left to right dimension, however. Recall that the dimensions are the result of oblique rotations. Hence dynamism is seen most clearly as a dimension if it is regarded as having a slight down-slope moving from the left to the right.
Four of the clusters revealed in the cluster analysis can be clearly differentiated in this space, and one can almost differentiate all six. Indeed, the results tightly parallel the results of the Average Linkage Between Groups Cluster Analysis, and the two dimensions that don't quite separate in the two dimensional space, Publishing and Film and Art, are the same clusters that had difficulty separating in that clustering.
Positioning in this two dimensional space is significant, and media can be seen to separate into four highly polarized "super-clusters". The middle ground of the plot is empty. The clusters occupy the four corners of the picture.
Correspondence occupies the lower left hand corner, and pretty much has the space all to itself. The same is true of Broadcasting, which is all alone in the upper right hand corner.
The lower right hand corner of the figure is shared by two clusters, but they clearly differentiate themselves. Face-to-face communication occupies the extreme lower right hand corner, and telephony grows toward the middle from that corner. The close relationship between face-to-face communication and telephony is clearly apparent here, as are the reasons for the cluster analyses' indecisiveness in assigning the VideoPhone to one cluster or the other.
The upper left hand corner is shared by the Publishing cluster and the Film and Art cluster. Film and art generally occupy the extreme left hand positions, with publishing distributed more toward the middle. This differentiation is not clear-cut, however, and we will need to examine one last two-dimensional representation to fully appreciate their relationship.
It isn't quite time to do this, however, as there is another relationship apparent in this two-dimensional space. As might be expected in the relationship left for last, this relationship may be the most interesting:
Consider the following:
If you look closely, you discover that every computer mediated communication system explored here forms a middle ground between its class of media and other classes of media. Indeed, you can draw a roughly elliptical shape in the middle of the interpersonal/mass communication versus dynamism plot and include nothing but computer media.
This ellipse doesn't form a cluster in the current analysis, but it does suggest that computer mediated human communication systems, which remain fairly new to the world, are establishing the base for a new class of media. It can be argued that new technology often starts by providing small variations on the existing functions of media, and only later realize their potential to create new categories of media. It may well be that, sometime in the near future, computer media establish a unique position among other media, involving media whose natures we cannot now even conceive of.
The basis for such movement is already clear, for instance, in the promise of Voice Recognition. Existing English language voice recognition systems can already handle continuous speech for vocabularies in excess of 5000 words, and it appears highly likely that affordable consumer voice recognition systems with much larger vocabularies will be available within the next generation. The impact these systems may have on computer mediated systems is hard to estimate now, but it seems clear that the speedup involved in moving data entry away from the keyboard via voice recognition systems cannot help but move computer-mediated systems closer to face-to-face media and the middle ground of the interpersonal/mass communication versus dynamism plot.
It is also likely to change the dynamics of computer media and the way they are used, as people will be able to respond to the comments of others at nearly the speed with which they can read those comments. This kind of instant verbal feedback can be mediated by computer to prevent "floor fights". Each party to a conversation will be able to have the floor whenever they want it, and conversation can become effectively simultaneous. These kinds of media dynamics may well move computer media into a distinct cluster (or set of clusters) that integrates existing media into new media.
+---+----+----+----+----+----+----+----+----+----+----+----+----+-+
| *Television |
+ + +
| |
M | |
A | |
S | Books|
S + Video Recording * +
| Talking Automaton*19 *Film * *Record|
M | Voice/Mixed Text & Graphics* |
E | Film with Subtitles* *Radio |
D | Automatons* Newspaper* *Magazine|
I + Dioramas*Art Mixed Text & Graphics* +
A | Weekly Newspaper* Electr Publishing*On|
| Multi-Media* Journals* Line|
V | Talk Radio* Info|
S | Silent Film* *Beeping Filmstrip |
+ Open Computer+
I | *Filmstrip Conference|
N | Newsletters* * |
T | *Electron|
E | Bulletin|
R0+ Board +
P | |
E | |
R | |
S | Voice over Data |
O + Ham Radio* * *Closed Comp+
N | *Public Speaking Conference |
A | *Videoconf. *Electr. Quips |
L | C.B.* *Voice Mail |
| Tele- Letters & |
M + VideoPhone conf. *Tape Letters +
E | Memos* ** * *Telegram |
D | Facsimile * *Telex & |
I | Telephone Electronic Mail|
A | *Face-to-face *Asynch Computer |
+ *Small Group *Intercom Conversation +
| Computer Syncrhonized Conversation**Cooperative |
- | *Notes Composition |
+---+----+----+----+----+----+----+----+----+----+----+----+----+-+
- BANDWIDTH 0 +
Brief mention should be made of the final set of contrasting dimensions before we bring this chapter to a close, as some fairly interesting relationships are revealed here. First, the Publishing cluster clearly differentiates itself from the Film and Art cluster when bandwidth comes into play. Here it is useful to look at the above plot and then reexamine the interpersonal/mass communication versus dynamism diagram, using the above figure to provide depth. If the left hand side of the above is taken as the top of interpersonal/mass versus dynamism, and the right hand side as the bottom, most of the media, and all of the computer media, fall to the bottom. Face-to-face media, however, are positioned at the top of the lower right hand corner of the interpersonal/mass communication versus bandwidth diagram. In the upper right hand corner, moreover, publishing media are at the bottom of our imaginary three dimensional space, and Film and Art media are floating above them in the middle of the space.
What is revealed in this typological exercise is a communication landscape that encompasses four severely polarized clusters of media, including what we might call distribution media, broadcast media, correspondence media, and real-time interactive media. Two of these clusters cleanly subdivide. Distribution media include both publishing media and film/art media. Real-time interactive media include both telephony and face-to-face media. While these differentiations provide six differentiable forms of media, the four corners of the interpersonal/mass communication versus bandwidth diagram clearly differentiate communication into four broad clusters of media.
Computer mediated communication enters this highly polarized world with a message of compromise. Every computer media examined is located in a typological middle ground. Each is associated with a particular cluster of media, but each reaches out from the edge of its cluster to other clusters. The statement made here is a strong one. Computer mediated communications systems, and computer conferencing in particular, must be regarded as a cross between interpersonal and mass communication. The result can only be communication media that allow people to communicate in new ways, and perhaps more effectively than has been possible in the past.
+--------------------------------------------------------------------+ | | | Distribution Broadcast | | | | | | | | | | | | | | | | | | Computer | | Media | | | | | | | | | | | | | | | | Real-Time | | Correspondence Interactive | | | +--------------------------------------------------------------------+
Note that this typological message, summarized in the above table, is very different than that of the informal typology. Although the placements of real-time interactive and correspondence media in that typology are similar to the one developed in this chapter, the two characteristics it uses, potential recipients and potential for feedback, are insufficient to the task of differentiating broadcast media from distribution media. Placements of computer media are completely different, moreover. Electronic mail is a real-time interactive medium in the informal typology, but a correspondence medium in the present typology. Computer conferencing is a "third extreme" in the informal typology, but is somewhere in the middle ground between distribution media and correspondence in the present typology.
This difference is ultimately rooted in nothing stronger than the number of characteristics considered. The twelve characteristics used in the formal typology probably give a fuller picture of the relationships between media than do the two characteristics of the informal typology. These relationships are complex, and the informal typology may be easier to understand than the formal typology, but the formal typology is probably to be preferred as a better summary explanation of the relationships between media.
Such a preference has implications. A computer conferencing that represents a third extreme of communication practice may be very different than one that combines elements of broadcast, correspondence, distribution, and real-time interactive media. Hence there may be much to gain in reexamining the formal typology from a somewhat different perspective.