2.3.1 Overview of CMC

Computer-mediated communication refers to human communication via computer. The emphasis is on interaction between humans using computers to connect with one another. The computers may be central repositories for human messages, or they may comprise a network of links and nodes facilitating the transfer of human messages. Various modes and media can be combined to facilitate the communication process.

Presently, most applications and much of the research into CMC is focussed on text-based communication. With the increasing sophistication of computer and communications technology, however, additional modes and media are being utilized. Ultimately, in support of human communication via computer, "the human intellect and sensorium are entitled to the fullest array of media we can muster" (Higgins cited in Levinson, 1990, p. 10). Further, computer technologies @CiteMark(LevinsonP90a) that support human intellectual activity as well as mediate communications should be given appropriate attention in the realm of CMC as their power and usefulness increases.

There are numerous incarnations of text-based CMC. Foremost among the asynchronous types is computer conferencing (CC). CC usually refers to those systems that reside on large central computers to which numerous users can connect directly via inhouse terminals or indirectly via home computers, modems, and telephone lines. Messages are composed and saved to files that are designed to permit multi-user access. As such, groups of people can establish "conferences", or "special interest group" (SIG) areas to which they send the messages they wish to share with others. In this manner, the "conference" resembles an ongoing discussion or meeting that does not require the physical presence of the participants in one place. Nor does it require that contributions to the discussion be synchronized with the attendance of other members of the group. All members can link into the conference discussion when they log in, at their own convenience.

Electronic mail (EMAIL) is another form of text-based CMC that has seen tremendous growth in business organizations. With EMAIL, there is no group-accessible central repository of messages. Rather, messages are sent to individual addesses just as conventional mail is sent to individual addresses. Of course, bulk mailings by individuals or by automated message handling software are becoming as common as their counterparts in the conventional mail systems. One variation on this theme, which is popular in higher education networks, is the LISTSERVER. Listservers function primarily to receive messages from individuals and then to distribute them according to mailing lists. Members of these mailing lists, then, constitute a group of communicants who can carry on asynchronous discussions much the same way as users of computer conferencing systems do.

Electronic bulletin board systems (EBBSs, or BBSs) are an emergent form of CMC activity. The original concept was that of the conventional community bulletin board as an area where public messages could be posted and read. In the early 1980s, BBSs were primarily established by computer user groups for the purposes of online distribution of software and related information. The messaging feature became increasingly popular as members found that the posting of bulletins soon resulted in interactive discussions. That is, bulletin messages could either be initial postings, or response postings. The group discussion potential of BBSs led to a broadening interest in the medium. With the advent of post-hobbyist home computers and the availability of inexpensive communications software, BBSs by and for the general public emerged. Many of these new BBSs were no longer oriented to computer enthusiasts. Instead, they came to be used as local fora for educational activities, special interest groups, writer collectives, and academic @CiteMark(HarasimLM88a) discussions.⁴

One popular feature of BBSs and of the larger commercial conferencing systems is the "CHAT" mode of interaction. This feature is conceptually based on another more conventional public communication system: Citizen's Band Radio. On a small BBS, users have the option of requesting a real-time, text-based chat with the SysOp. On larger systems, numerous simultaneous users can interact in real time and all participants can observe the contributions of the others as they occur. Just as the radio waves provide for multiple simultaneous users, the CHAT mode of large conferencing systems allows for similar interaction in text.

CHAT mode is an example of synchronous text-based CMC. Most computers used for communications have at least a rudimentary form of this feature installed. The various incarnations of synchronous CMC differ primarily in the way the exchange is presented on the screen, and in the number of simultaneous users who can participate in any one group activity. The exchange can be character by character, line by line, or in multiple-line blocks. In character by character mode, the recipients see each character as it is typed by the sender. In line by line mode, the message is not sent until the sender presses the key (or ), and in the multi-line blocks mode, several lines are composed before the message is sent. In all cases, however, the transaction occurs immediately and is not held in a mail system or conferencing space for future retrieval. This is not to say that transcripts of the interaction cannot be collected (logged) for future review of the discussion; it is just that the exchange is intended to be in real time.

The user interface for synchronous text-based interaction often involves some partitioning of the terminal screen into window areas. A common design for two simultaneous users involves splitting the screen horizontally. If the top half is used for receiving messages, the bottom half may be used for composing and sending messages. Systems with more advanced GUI (graphical user interface) features can allow for multiple windows of variable size and location to appear on the screen with any or all of them being used to receive or send messages. The CSILE Phone utility developed for the synchronous dyads in the present study (see Figure @Ref(figscreen1)) used one window for both receiving and sending messages character by character. Other window space was reserved for future uses such as review of previous transcripts or access to information files (neither of which was implemented at the time of the study).

2.3.2 Studies in Asynchronous CMC

In The Third Wave (1980), Alvin Toffler documented many of @CiteMark(TofflerA80a) the current trends in society and projected future scenarios. His discussion of "flextime" included the following:

Indeed it is the computer which has made flextime possible ..., but it also alters our communications patterns in time, permitting us to access data and exchange it both 'synchronously' (ie: simultaneously) and 'asynchronously'.

What that means is illustrated by the growing number of computer users who are today engaged in 'computer conferencing'. This permits a group to communicate with one another through terminals in their homes or offices. Some 660 scientists, futurists, planners, and educators today in several countries conduct lengthy discussions of energy, economics, decentralization, or space satellites with one another through what is known as the Electronic Information Exchange System [EIES]. Teleprinters and video screens in their homes and offices provide a choice of either instant or delayed communication. Many time zones apart, each user can choose to send or retrieve data whenever it is most convenient. A person can work at 3:00 A.M. if he or she feels like it. Alternatively, several can go on line at the same time if they so choose. (p. 252)

EIES was designed by Murray Turoff and had its test period in 1977-1978@Cite(HiltzSR78a ", p. 10"). Turoff also designed and implemented the very first computer conferencing system in 1971. It served as an automated Delphi conference for the Office of Emergency Preparedness in the Executive Office of the President of the USA (p. 46). Starr Roxanne Hiltz, a sociologist, began reporting on the effects and impacts of computer conferencing in 1975. Together they wrote the now classic The Network Nation (1978) which remains important because of the breadth of analysis and insight it delivers. Hiltz and Turoff defined, clarified, or predicted many of the issues and problem areas in CMC that remain as the focus of questions and investigation more than ten years later.

The wave of the future in human communication, however, is more than Toffler's characterization of "flextime" as a simple matter of convenience. In fact, asynchronous CMC (computer conferencing) was propelled to the leading edge of a new communications revolution primarily because of the way it alters human communication patterns. As early as 1975, with relatively unsophisticated conferencing software (FORUM), Jacques Vallee, Robert Johansen, and Kathleen Spangler of The Institute of the Future were able to report on computer conferencing as an altered state of communication with significant social implications@Cite(ValleeJ75a). Their report presented many of the same factors and impacts that continue to be discussed in the more recent CC literature@Cite(HarasimLM90a, RapaportM91a, SteinfieldCW86a, TuroffM89b). Three main themes or issues predominate: alterations in time and place, text-based interaction, and socioemotional content.

Turoff (1989b) is particularly concerned with the use of computer conferencing for group support. He contends that the asynchronous mode of interaction is most important. This is not necessarily because of time and place convenience, but because "the potential for real improvement in the group process lies in the fact that individuals can deal with that part of the problem they can contribute to at a given time, regardless of where other individuals are in the process"(p. 11). The result, he claims, "is to free the individual to deal with the problem in ways consistent with his or her cognitive style"(p. 11).

It appears then, that Turoff is discussing a particular type of problem or decision task, or perhaps a certain phase in the problem-solving or decision making process. This phase would be one that involved exploration of possibilities and ideas for general consideration. It would have to be a phase where the consensus of the group would not yet be necessary, since once consensus is needed it becomes more important for individuals to synchronize their contributions with those of other individuals in group activities. Turoff suggests the software should be designed to synchronize the asynchronous interaction of the group@Cite(TuroffM89b ", p. 11").

With regard to text-based interaction, Rapaport (1991) states that it "addresses a potentially wider set of domains than any other teleconferencing or office automation system. Its purpose is the association of ideas and experience of people, communicated in written words" (p. 29). Feenberg, in his paper "The Written World" (1989) adds, "A group which exists through an exchange of texts has the peculiar ability to recall and inspect its entire past" (p. 25). @CiteMark(FeenbergA89a) In these two statements we have the makings of a natural hypertext: the association and recall of ideas in text. With the aid of computer based utilities to facilitate the linking and association of ideas, we can presumably make a quantum leap beyond any bandwidth constraints on text-based communication. Besides, whereas a picture is said to be worth a thousand words, one might also proclaim that certain combinations of words can elicit a wealth of images, sounds, feelings, and other sensations.

Provenzo (1986) refers to the proliferation of electronic text and information networking opportunities as the "electronic scriptorium" (pp. 50-51) and he predicts that:

As we go beyond the Gutenberg Galaxy [@Cite(McLuhanM62a)] and enter a post-typographic culture, we will see the traditional methods and results of scholarship redefined. Through the establishment of information networks and knowledge pathways made possible by the microcomputer and telematic systems, we will have the potential to extend and refine intellectual discourse in a manner that has never been possible before. @Cite(ProvenzoEF86a ", p. 50").

Considering CMC in the context of the evolution of media, Levinson (1990) describes the development of communications technologies in three historical stages. First, there is the stage of natural communication that takes place within the confines of biological boundaries. The next stage involves the development of technologies which help transcend the physical limitations of human biology. Writing is a technology of this second stage. It evolved in order to facilitate communication across greater distance and through time. According to Levinson, however, new communication technologies in the second stage often result in the sacrifice of certain features and qualities of communications in our natural environment of the first stage. The written word, and ultimately the advent of the printing press, resulted in the loss of interactivity in our communications. In the third stage, technologies evolve in a manner that recaptures features of natural human function; a process that Levinson refers to as "anthropotropic" (p. 5). @CiteMark(LevinsonP90a) Therefore, computer conferencing is proclaimed as "the first progress in the evolution of media since the mass application of the printing press in fifteenth-century Western Europe" (p. 9) because it recaptures the lost interactivity in text.

Without encompassing the full range of human sensory and expressive capabilities, text-based interaction is often thought to be an impersonal medium devoid of social context cues and nonverbal communication. Experience and research, however, are demonstrating that socioemotional content can be communicated in text. Steinfield (1986) states that: "Evidence continues to mount showing that CMC will be used for emotional interaction. People seem to work around the nonverbal cue limitations and actively provide their own text-based translations of nonverbal cues" (p. 176). @CiteMark(SteinfieldCW86a) Tracz (1980) bears out this perception in a comment on his experience: "I was pleasantly surprised, nevertheless, that most users of EIES attempt to incorporate many little expressions to compensate for the lack of face-to-face contact, and on the whole, gentleness prevails" (p. 17). @CiteMark(TraczG80a)

In terms of cognitive effects, very few studies address the issue directly. Most report on social relations, interaction and outcomes of problem-solving tasks. Kerr and Hiltz (1982), however, surveyed the most prominent CMC projects of that time by soliciting opinions and observations from key representatives of the various groups and organizations utilizing this technology. In line with much of the research preceding their study, and much that would come after, a fairly standard set of factors and effects were analysed. These included system software design factors, and factors affecting user acceptance of new systems. (p. 174-175)

Kerr and Hiltz also gathered information about the cognitive impacts of CMC on individuals and groups. They reviewed the existing literature, and used reports from evaluators to confirm or disconfirm their expectations . They found that, "the more socially significant cognitive impacts, such as those including conceptual skills and learning [were confirmed], whereas those which may be more trivial, such as spelling and typing skills, and those which are clearly negative in impact, such as information overload [were not confirmed in the literature and reports]" (1982, p. 102). @CiteMark(KerrEB82a)

2.3.3 Studies in Synchronous CMC

As noted previously, most computer supported communication systems have some capability for synchronous text exchange. Strom (1982) notes, interestingly enough, that the @CiteMark(StromBI82a) PARTY-LINE system developed by Murray Turoff (prior to EMISARI) was the first to implement a synchronous exchange facility. (p. 291). Rapaport (1991) confirms this and goes on to state: "Initially, synchronous support was expected to provide the system's primary value. Dr. Turoff quickly discovered, however, that most of the real work was done asynchronously" (p. 4). Hiltz and Turoff (1978) also state: "Of all the communication forms and conditions permitted by computerized conferencing, the synchronous discussion seems to cause the most difficulty and feelings of confusion" (p. 90). The evidence provided by Hiltz and Turoff is rather slim. They relate the experiences of people in one real-time "party" where insufficient computer core memory allocation resulted in line by line messages taking up to 30 seconds to transmit. There was no structuring of the communication protocols. In fact, there was overt refusal by Roxanne Hiltz, who initiated the "party", to take a leadership or moderating role. The following excerpt characterizes the situation: (see Hiltz and Turoff, 1978, p. 93)

14570 M ELAINE KERR (ELAINE, 114) 4/1/77 6:05 PM
   Private message: Roxanne, this is awful, please be
dictatorial, seems nothing's being accomplished.

14577 M ROXANNE HILTZ (ROXANNE, 120) 4/1/77 6:18 PM
   I am simply not going to try to tell you what to do
or say, so please stop asking me, people!

Fortunately, there are those who have had different experiences with synchronous CMC and consider it worthy of further investigation. Dobos and Grieve (1985) studied the decisional productivity of synchronous online conferencing by comparing the ratio of decisional output messages to messages representing social presence, task related input, and group procedural input. Their results suggest that the use of turn-taking protocols and the use of rotating moderators (with each participant "handing-off" to the next when their contribution was finished) increases decisional productivity. Further investigation demonstrated increased participant satisfaction when turn-taking protocols where implemented in online synchronous groups. (pp. 376-377)

In their report, Dobos and Grieve comment on the lack of attention given to synchronous text-based interaction in the literature. According to their reference list, however, they were unaware of the work of Dubrovsky, Kiesler & Siegel (1983). @CiteMark(DubrovskyV83a)

Kiesler and co-workers at Carnegie-Mellon University investigated the social psychological aspects of computer-mediated communication@Cite(DubrovskyV83a, KieslerS84a, KieslerS85a, SiegelJ86a). Their 1984 paper was published in American Psychologist, became very popular, and is often cited. It was also selected for inclusion in an important collection of papers on computer-supported cooperative work four years after its original publication@Cite(GreifI88a). In fact, it is a very well presented report of research that carefully delineates the factors and effects studied.

The problem with this account, however, is that they failed to clarify one very important point. Whereas they used the terminology of computer-mediated communication, they did not emphasize the fact that most of their data was drawn from synchronous text-based interaction, not asynchronous computer conferencing. Further, although they did present a portion of their work comparing users of the synchronous mode with users of electronic mail, they continued to confuse the issue by referring to their synchronous mode as "computer conferencing"@Cite(KieslerS84a ", p. 1130").

The effect of this combination of a popular scholarly report and the confusion in terminology relevant to the field, is that some of its main findings have been misapplied to the more widely implemented and researched asynchronous mode of computer conferencing. Most significantly, the issue of uninhibited verbal behavior, reflected in the fiery term "flaming", came to be associated with CMC in general. The truth of the matter is that the data presented in Kiesler et al. show higher levels of uninhibited verbal behavior in synchronous CMC, but the levels for EMAIL (an asynchronous mode) are more comparable to the lower levels found in face-to-face interaction. Figure @Ref(fkies1) shows their graph comparing uninhibited behaviour in face-to-face (bars a & f), simultaneous computer conferencing (bars b, d, & g), simultaneous computer conferencing with turn-taking (bar e), simultaneous computer conferencing with anonymity (bar c), and computer mail (bar h)@Cite(KieslerS84a ", p. 1129, Fig. 1"). Note how the term "computer conferencing" is used in the legend without clarification as to synchronicity. This deficiency is corrected in a more recent reproduction of the graph (see Sproull & Kiesler, 1991, p. 119) were "real time" is added @Cite(SproullL91a) to the "computer conferencing" legend item.

Figure 2-4: Uninhibited verbal behaviour (from Kiesler et al.)

This figure was adapted from Kiesler et al. (1984). They use
"Computer Conferencing" in the legend without noting that
it is synchronous CMC.

An important new perspective is offered by Ferrara et al. (1991) who refer to synchronous CMC as interactive written discourse (IWD) and suggest that it represents an emergent linguistic register. The concept of "register" deals with language variations in the context of language use, as opposed to "dialect" which refers to language variations according to user. Previously studied registers include "Baby Talk", "Foreigner Talk", "Sports Announcer Talk", and "Note-taking" (pp. 10-12).

Ferrara et al. studied synchronous text-based CMC in the form of message exchanges between a travel planning agent and 23 computer professionals (or their spouses) who were making travel arrangements. Their findings support three claims about IWD:

... first, that it is a naturally occurring register, perhaps a reduced register; second, that it is a hybrid language variety, displaying characteristics of both oral and written language; and third, that norms of its use are in the process of becoming conventionalized. (p. 10)

"Reduced register" refers to registers that frequently omit components of the language such as copulas, articles, and pronouns. Note-taking is a reduced register that is generated under real time conditions with pressure on the writer to save time and minimize effort. Ferarra et al. point out that IWD is a register comparable to Note-taking in the type of register reduction it exhibits; except for one crucial difference. IWD is interactive and there is need for somewhat improved clarity in the communications because the audience is other than the self (1991, p. 13).

IWD as a hybrid of oral and written language varieties is demonstrated by several characteristics. For example, IWD compares to oral language in its "use of informal discourse particles such as okay, sure, sorry, and now, and colloquial usage such as Let's and How about ..." (p. 23). Some IWD compares to written language in the use of cataphora, or forward referencing: "Cataphoric reference was quite common in the corpus. Phrases such as `the following criteria,' `the following two airlines,' or `the below mentioned connections' ... were used ... "@Cite(FerraraK91a ", p. 24").

These findings represent the work of individuals who are only remotely familiar with the CMC literature (as revealed in their citations and bibliography). They have introduced techniques and concepts of a discipline not previously applied to the study of CMC, and as such, they present an important outsider's view: a perspective from which many more studies should arise. Their projections about the conventionalization of IWD, however, may not be well founded. They state, for example:

There is ample reason to believe that this type of written communication will proliferate in the future, perhaps even influencing or outdating the stylistic conventions of traditional writing styles@Cite(FerraraK91a ", p. 30").

Well, the online activity of computer professionals is certainly not evidence enough to make such broad generalizations. Another view might be that IWD is simply an interim artifact doomed to be reviewed in history as one of those strange things that humans did as human and social factors research strove to keep up with a stampede of technology. Perhaps the Kiesler et al. view is closer:

Based on current trends, text-only electronic communication systems will become more popular. In that case, we should study both their transient effects (like those likely to disappear when the technologies are mature) and their more permanent and secondary effects (1984, p. 1130).

2.3.4 Computer-Supported Cooperative Work (CSCW)

Most human work involves both individual effort and group participation. The "personal" computer and common applications software such as database managers, wordprocessors, and spreadsheets have provided significant support to people working independently in various organizations. These people, however, must step out of their computer support environment whenever group activities such as meetings are scheduled. In response to this problem, and in conjunction with developments in media and communications technologies, a new area of research and practice has evolved. "Over the last half-dozen years, Computer-Supported Cooperative Work has emerged as an identifiable research field focussed on the role of the computer in group work"@Cite(GreifI88a ", p. 5").

CSCW has drawn together research interests from a variety of fields in order to investigate groups of humans using groups of computers in their work. Sociology, anthropology, and organizational behaviour offer methods and knowledge about the study of humans in groups. Computer science offers methods and knowledge concerning computer-human interaction (CHI), interface design, and communications networking. Together, the integration of these perspectives have resulted in the development of applications that provide research information about group work in shared multimedia spaces. The general direction presently suggests the use of advanced multimedia technology to facilitate group interaction concerning the shared objects of work@Cite(CrowleyT91a, ManteiMM91a). The strength and importance of this trend is revealed in the fact that the latest operating system and user interface for the Macintosh computer (System 7, 1991) incorporates shared document and workspace capabilities as a standard feature of its networking environment. Tocco (1991) gives @CiteMark(ToccoG91a) an example of the use of System 7 features for group work. With the combined effort of several people on a shared document, he states that the document seems to come alive as it evolves with the completion of each participant's contribution (p. 48).

It appears that the time for individual, "personal computing" software applications is fading away (at least in organizational settings). Document sharing and computer-mediated discussions in support of group work are on the rise. The term "groupware"⁵ is being applied to various commercial @CiteMark(KerrEB82a) applications packages that integrate facilities for document sharing and messaging with the more common text-processing, database, and spreadsheet functions. Greif declares,

All software will be groupware. But CSCW as a research field will still be addressing the larger questions of how to design and refine good groupware - software that will allow people to work together with the best help they can get from the computer@Cite(GreifI88a ", p. 12").

A related development in the field of management science is the work involving group decision support systems (GDSSs). This represents a variation on the CSCW theme@Cite(KraemerKL88a). After the groupware model, GDSSs attempt to integrate group sharing and communication structures with management software. In particular, GDSSs are best understood as the groupware version of more conventional decision support systems (DSSs). DSSs use operational data from an organization to provide simulations and models to help inform management decisions.

Baecker (1991) provides a taxonomy of CSCW technology (see Figure @Ref(fcscwtax)) that classifies applications according to

Figure 2-5: Baecker's Taxonomy of CSCW Technology

This figure was adapted from Baecker (1991, p. 6).

The concept of a single-site asynchronous utility may represent an area of knowledge and interest dissonance between proponents of CSCW and proponents of CC. In fact, the blank cell in Baecker's taxonomy may indicate a lack of appreciation for the asynchronous text-based interaction promoted by Turoff and others. Turoff on the other hand, is quick to downplay the importance of GDSS and CSCW@Cite(TuroffM89a ", pp. 111-112") by suggesting that computer conferencing already offers GDSS via a flexible, tailorable system and that CSCW, being mostly synchronous in nature, cannot match the group support power of asynchronous CC. It should be clarified as well, that Turoff has good grounds for his view that work in CC has already addressed a lot of the issues now arising in the GDSS and CSCW work. In The Network Nation (1978), Hiltz and Turoff discuss DSS applications specifically (p. 273) and CSCW indirectly as "multimedia technology":

The ideal system to some technologists appears to be the ability to talk to someone while watching him or her on a TV screen and transferring written material digitally.

...

We see little real benefit in such multimedia setups, since they defeat the time dispersion benefit of computerized conferencing and the psychological atmosphere encouraging reflective thought CCS's [computer conferencing systems] generate. Multimedia communication systems foster formality, impose status relationships on the participants, or accentuate such relationships that already exist (p. 409).

Specific critiques of CC appear to be directed more at the issue of it being mostly text-based than at the issue of synchronicity. Greif notes:

... when computer conferencing is applied to some tasks, the model breaks down. The unstructured body of messages is suitable for the free flowing text of natural language, but does not let us set the computer to work on our problems. Designers who draw pictures, software developers who jointly write code, financial analysts who collaborate on a budget - they all need coordination capabilities as an integral part of their work tools (1988, p. 7).

Kraemer and King (1988) report on a more general feeling:

... most participants report that the computer conference hardly feels like a meeting at all, and many are unwilling to participate in them more than a few times (p. 125).

This argument about how it feels would carry little weight with Turoff, however. One of the most important points made throughout The Network Nation is that conferencing technologies would ultimately change the way we think, understand, and feel about work just as much as what we do as work will change.

---