+ Page 31 + ---------------------------------------------------------------------------- ####### ######## ######## ########### ### ### ## ### ## # ### # Interpersonal Computing and ### ### ## ### ## ### Technology: ### ### ## ### ### An Electronic Journal for ### ######## ### ### the 21st Century ### ### ### ### ### ### ### ## ### ISSN: 1064-4326 ### ### ### ## ### April, 1994 ####### ### ######## ### Volume 2, Number 2, pp. 31-49 --------------------------------------------------------------------------- Published by the Center for Teaching and Technology, Academic Computer Center, Georgetown University, Washington, DC 20057 Additional support provided by the Center for Academic Computing, The Pennsylvania State University, University Park, PA 16802 This article is archived as METZ IPCTV2N2 on LISTSERV@GUVM ---------------------------------------------------------------- COMPUTER-MEDIATED COMMUNICATION : LITERATURE REVIEW OF A NEW CONTEXT J. Michel Metz, University of Georgia ABSTRACT Ever since computer-mediated communication came into existence in the 1970s, researchers have used it as a tool to examine its effectiveness within organizational, interpersonal, and mass communication contexts. This paper analyzes the existing literature regarding CMC, and finds no continuity. It therefore argues for the existence of a CMC context of communication, which would guide future research along a cohesive vein, encompassing all the different sub-contexts of CMC (such as electronic mail, computer conferencing, Relay, and Multiple User Dungeons). STATEMENT OF THE PROBLEM From a time when computer use meant entering punch cards and receiving greenbar printouts without the use of monitors, came the discovery of the ability to send one-line messages between college campuses. Thus, to alleviate the boredom, programmers then decided to try to create "chat programs" which would also enable somewhat more reliability in sending and receiving these messages (Kell, 1988). + Page 32 + It is a common cliche that necessity is the mother of invention. If this is true, then boredom must be one productive mother. From these chat programs came more intricate Relay programs -- programs which linked individual chats together -- involving several higher-level commands, allowing several people to talk on individual channels at once, or "move" to private channels for private discussions. As with any new resource (i.e., telephones, radio, television, fax machines), there is an awkward period when even the people who develop them have no idea to what extent their inventions will be used and, in certain cases, abused. The developers of the Internet, for example, had no idea that the most influential aspect was to be email (Metcalfe, 1992). It is one thing to produce a program and eradicate bugs from it (problems which hamper the effectiveness of the program), but it is another to predict how it will be used. Once the element of humanity is involved, the margin for error with control becomes much greater. Computer-mediated communication (CMC), for purposes here, can be defined as any communication patterns mediated through the computer. These include, but are not limited to, computer conferencing, electronic mail (e-mail), relay chat lines, and Multiple User Dungeons (MUDs). CMC has been in existence since 1969, when the creation of the ARPANET produced unexpected benefits regarding email. Even those who helped develop the technology had no idea that communication would be its most important asset (Metcalfe, 1992). The impact came as such a surprise, in fact, that research concerning CMC has scrambled along behind the technology and the culture with no theoretical base to guide it. Much of what has been studied, therefore, looks very disorganized. CMC researchers have failed in their duty to organize and define their field of study. Many researchers have struggled to get their studies and opinions out into the forefront with (in certain cases) disregard to what has been studied before, or what needed to be studied. As a result, certain topics (i.e., computer conferencing) have been analyzed to an almost ridiculous extent, while others (such as computerized bulletin boards, online relays, and the newest member of CMC: Multiple User Dungeons, or MUDs) have been alarmingly neglected. It is the thesis of this paper that there is no cohesive groundwork for studying CMC, and, as a result, there is no fundamental theory which guides CMC research. There are no metatheoretical sources available on CMC, leaving research about the research nonexistent. + Page 33 + Therefore, this paper is broken down into four major sections. The first section postulates the work which has been done considers CMC as a subset of previous communication contexts, primarily organizational communication. The second section gives evidence to support the hypothesis that much of the work has been done in isolation to other studies concerning CMC, and provide little or no continuity from one body of research to another. Third, this paper will examine the few attempts to combine CMC findings, and examine how such attempts have missed the mark. Finally, this paper argues that CMC is a field of theoretical study in its own right, not merely a channel to study within other theoretical contexts (e.g., interpersonal, organizational communication). CMC in Organizational Settings For the most part, research involving CMC has been designed to determine worker productivity and effectiveness within organizational settings. Rice and Case (1983) examined the uses and effects of a computer-based message system within an organizational setting. The study focused on a pilot program for the introduction of the Terminals for Managers (TFM) program at a major west coast university. The program was an attempt to "facilitate communication within the university's administration and eventually to provide other management aids" (Rice & Case, 1983, p 132). They particularly discussed the results related to managerial communication because "the payoff... lies in the use by managers; (and) TFM was specifically designed for managerial use" (Rice & Case, 1983, p. 133). In essence, the report was broken down into three separate studies. The first study examined the expected benefits of the program in terms of the frequency (number of times per day), duration (number of minutes per day), and experience (number of weeks on the system) of the users on the system. The researchers surveyed 89 managers and administrators, and 110 computer science (CS) staff in two waves of questionnaires. They found that managers may use TFM less over time, but became more efficient by logging on to the system fewer times per day while staying on nearly the same total number of minutes. Overall, the researchers found no significant relationship between the amount of time a respondent reported using a computer system and their perceptions of how appropriate computers are for communication tasks. However, heavy users found TFM substitutable for face-to-face communication (Rice & Case, 1983). + Page 34 + Rice and Case then investigated whether there was a correlation between the duration of use and the benefits arising from use. It was hypothesized that heavier users of TFM would decrease the amount of telephone use and paper production. Rice and Case also proposed TFM would affect the quantity and quality of work. The second study, using the same sample, identified duration of use as the independent variable, and the amount of telephone use, production of paper, and the quantity and quality of work as the dependent variables. Surprisingly, heavy duration of use did not lead people to believe that benefits outweighed the costs. Apparently there was a diminishing returns effect on the use for low to medium level users. Heavy users acknowledge benefits early on, and continued use over time did not alter these feelings. Also, they found that CMC may serve as an additional form of communication rather than as a substitute for any existing one. This is perhaps the first indication that computer mediated communication might be seen as a separate entity, and possibly one worth studying, but neither Rice and Case nor subsequent researchers followed through with that hypothesis. The third study found that greater duration of use was associated with more positive benefits and impacts, albeit these benefits and impacts were not reported to change over time. There were several possible explanations for these results. For example, equilibrium could have been attained rapidly after the introduction of the new medium and was unrelated to the number of weeks on the system. Or, people had unrealistic images of expected benefits and impacts unrelated to actual usage over time. A third possibility could be that there were more complicated relationships among usage, media use, and reported benefits and impacts (Rice & Case, 1983). According to the researchers, the possible implications of the study were threefold. First, positive impacts of CMC in an organization that people report may not be related to their actual use of the system. Essentially, what this means is that users of CMC might sustain their preconceived notions about communicating via the computer, and therefore perceived benefits (read positive impacts) might have been attitudinal rather than developmental over use and time. Second, computer- based technologies should be used where they are appropriate to organizational tasks and managerial styles, rather than indiscriminately thrust into any communication activity. Implementing a new technology, or any such new device, for the sole purpose of having something new could be counter- productive. Third, personality traits, job tasks, positions, and media styles that affect how people use technology will be a major factor in acceptance and consequences of computer mediated communication within an organization (Rice & Case, 1983). For example, individuals who have already determined that they are "computer illiterate" and completely inept with new technologies might not find any perceived benefits, regardless of the amount of time spent on the system. + Page 35 + Steinfield (1986a) provides a second example of the organizational pigeonhole for CMC research in his study of electronic communication within a business organization. The dependent variables in the study were task-related use, which was the amount of use of CMC directly related to task-functions, and the amount of social use of CMC, which was the socioemotional use not related to task functions. There were four independent variables: a) task and environment variables, such as task complexity, routineness, standardization, unpredictability, time pressures, working with new people, and occurrences of crises; b) channel characteristics, which were defined as "subjects' perceptions of the attributes of electronic mail" (Steinfield, 1986a, pg. 788); c) individual and demographic variables, such as e-mail experience, education, and level within the organization; d) and access to the system. Two hundred twenty randomly chosen CMC users, selected from a list of registered system users, completed self-administered questionnaires. Steinfield, like Rice and Case (1983), found that task-related use depended upon availability of access, meaning that completion of a task was dependent upon the accessibility of a computer terminal. It was also found that the task environment related positively to the measures of complexity of the task, task interdependence, uncertainty in the environment, and need for communication across locations (Steinfield, 1986a). In addition, Steinfield found that the perceived utility of e- mail had the strongest relationship to task-related use. In other words, the more a user thought CMC was useful, the more he would be likely to use the system for a specific task. The individual and demographic variables played no part in predicting task use. Steinfield then examined the variables in terms of social use. It was found that the accessibility of a terminal had no effect on social use, nor was the task environment a factor. Only the perceived attributes of the system, and the individual and demographic variables were significant in predicting social use. For example, newer and younger company members were more likely to use CMC socially. The difficulty in attributing these results to a CMC context is the fact that they presuppose that CMC is a tool, rather than an element which affects communication. In addition, this is a study involving only one organization, and only one communication program, whose features were not described. One is forced to make an assumption, therefore, about the ease of use of the program compared to others available. Also, if these results were to be attributed to CMC as a whole, there would be a risk of singling out specific groups as potential users of CMC (in this case, newer and younger members of a business). + Page 36 + Steinfield (1986a) noted that this research will help assist in the future implications of electronic mail systems and help create necessary applications for existing systems. I disagree. His research was valid and topical for the then-current existing systems. However, the results fail to provide accurate predictive power for the ever-changing world of CMC. By orienting his research to organizational research, the scope of Steinfield's research is outdistanced by the very medium he is studying. The third example in this review is a study which examined the introduction of computerization (including electronic mail) into a community-oriented nonprofit organization (Rubinyi, 1989). The study followed 72 small sub- organizations over a two-year period from their initial adoption of computer technology. Surveys were administered at the baseline, or adoption, point and one month, three months, six months, nine months, one year, and two years after the computer system was received. An "internal success" index was constructed through a factor analysis of variables related to successful internal implementation of the computer. These variables consisted of frequency of use and organizational productivity. The high- and low-scoring groups were analyzed at the baseline, one-year, and two-year points to determine if they differed significantly on various measures, such as integration of the computer organization with productivity (Rubinyi, 1989). The study found that organizations which had successfully implemented the use of computer networking had access to more resources than unsuccessful organizations. Also, "the major obstacles to computer adoption faced by resource- poor groups were the lack of staff time and lack of training" (Rubinyi, 1989, pg. 118). Finally, the study found that the computer did not create new and dramatic forms of interaction among the groups involved in the project, although several groups did benefit from implementing the internal applications such as word processing or financial management. Rubinyi concluded with a plea for investment in helping community- based social service and interest groups implement new information and communications technology (Rubinyi, 1989). Perhaps the lack of finding in this study was due to flaws in the methodology. In describing the study, at no point did the researchers discuss the type of CMC used, what resources that were available due to CMC, or any other communication-related differences due to computer use. + Page 37 + This study perhaps best exemplifies the problem with identifying CMC as an organizational tool taken to its logical extreme. The research has no ties to any theoretical background, but rather is simply a report on organizations which attempts to incorporate CMC into its day to day activities. As a result, the study has little potential for inclusion into a mainstream of communication literature, providing little assistance in furthering understanding of CMC or organizational uses of the medium. CMC Research in Isolation The trend towards a link-less research base is not reserved to organizational studies, though. Sudden breakthroughs in software and communication technology have raised much curiosity concerning the new applications, yet much of the work apparently has been done in isolation, providing no coherent link to any other research. A prime example of such isolation is a theoretical work on the effects of computer mediated communication on human linguistics. Baron (1984) studied whether the use of the computer as a linguistic medium will affect the very shape and functioning of traditional language itself. Analyzing and comparing computer and linguistic research studies and literature, Baron first examined CMC in the context of a broader set of linguistic issues. She analyzed CMC as a formal modality of linguistic communication, and made some general predictions about the kinds of linguistic change we might expect as a result of CMC. For example, she determined that "we might expect to see an improvement in... speech. That is, the degree of logical coherence and grammaticality in our speech might begin to approximate more closely that of our written language" (Baron, 1984, p. 138-9). As evidence she examined the concept of "professorial speech," where academians begin to "talk the way they write" (p. 124) Baron considers the computer as one of the "few moments in human history when technological innovation has the potential for radically altering human society within the space of a generation" (p. 139). Although she proposed that such a change would be beneficial, she gave an inordinate amount of credence to the power of the computer to change the linguistic functions of human social interaction, determining that "computer mediated communication may serve to discourage actual human encounters" (p. 136). Her evidence for this rests shakily upon the fictional work of Isaac Asimov where inhabitants of another world never saw each other, but rather viewed their conversational partners through videophones or computer conferencing. + Page 38 + A second example of a literary isolate would be Gratz and Salem's (1984) study on the potential influence of emerging information technologies on a psychological reality: self identity. In their essay, Gratz and Salem saw that although these technologies offer great opportunities, they must be managed with care to avoid serious communication problems. As evidence, they reviewed both popular literature and literature on how new technologies have altered self concepts and identity to a significant degree (Gratz & Salem, 1984). For example, they cited Kosinski's Being There, where the hero Chauncey becomes so influenced by television, that his responses "in the social environment he was forced to encounter were simply orienting reflexes directed at coping with the incoming social stimuli. He was unable to present himself to others, because he did not know a self capable of presentation" (Gratz & Salem, 1984, pp. 99- 100). They concluded that computers cannot fulfill many social functions, and, as a result, users may avoid negative consequences. The authors suggested finding methods to complement new technologies with opportunities to develop relationship-building skills, but they fail to indicate what those methods might entail. In addition, the authors relied too heavily on these popular literature models, such as Burgess Clockwork Orange, Mitchell's Gone with the Wind, and Kosinski's Being There, as demonstrations of how a medium "cannot assist in constructing the relationships that confirm self" (Gratz & Salem, 1984, p 100). As a result, many of their conclusions are based upon fictitious character development, rather than actual studies done to substantiate their claims. A third example of isolated research involves an attempt at determining whether CMC contributes new information in an organizational setting. Sproull and Kiesler (1986) explored how electronic communication related to self-absorption, status equalization, and uninhibited behavior as well as these effects on the contribution of new information in a Fortune 500 company. They used questionnaire data and actual electronic messages sampled from users to examine CMC in all levels of the organization. They found that decreasing social context cues, such as the lack of nonverbal and facial cues, had significant deregulating effects on communication. In their study, Sproull and Kiesler used specific message attributes, such as "length (of the message), opening, closing, positive affect, negative affect, politeness, energy, and topic" (Sproull & Kiesler, 1986, p. 1499) as the independent variables. The dependent variables were partner attributes, absorption effects (where the user was determined to be either self-absorbed or other-centered), status equalization effects, and uninhibited + Page 39 + behavior. For example, people overestimated the value of their own contributions to e-mail communications, while underestimating the value of their own group messages, and that people appeared to focus on themselves more than on others in message salutations and closings (Sproull & Kiesler, 1986). Their findings determined not only relatively weak social context cues in CMC, but also that people preferred to use CMC to send messages to superiors rather than to subordinates, and for sending bad news. While confirming what had been a long standing belief that a user's behavior is altered in CMC when compared to face-to-face, the authors found that people behaved more "irresponsibly" using CMC (Sproull & Kiesler, 1986). When evaluating that behavior, Sproull & Kiesler neglected to justify or even define the term "irresponsible." As a result, the following questions arise: what do the researchers deem "irresponsible behavior" on an electronic mail service? Is the behavior deemed irresponsible to an organizational setting, or in terms of basic human etiquette? What effect, if any, did irresponsible behavior have on the message as it was being perceived? Was irresponsible behavior a hindrance in overall communication? These questions do not necessarily undermine the results of the study, but one is forced to contemplate the context in which a message is "irresponsible." A fourth example, by Reid (1991), concerned an attempt to examine another context of CMC that had previously been ignored. This context is called Relay or, depending upon the network which facilitates its use, Internet Relay Chat (IRC). Underscoring the lack of contextual and social cues, Reid (1991) examined the behavior and attitudinal changes of "online" communication, using actual messages sent via IRC. She found that such a reduction in social context cues forces the users of CMC to adapt and adjust by providing substitute nonverbal cues and allowing "the shy and socially ill-at-ease" a way of learning social skills in a non-threatening environment. Ultimately, she argues for the existence of an emerging community of IRC users into a subculture of its own. She refused to place a value judgement based on that culture, since IRC is a medium in which behavior that is both outside and in opposition to accepted norms is accepted and encouraged (Reid, 1991). However, while she managed to defy the temptation to morally judge the IRC, it has come under attack in the past (as reported by Kell, 1988). In addition, Reid offered no reasoning or arguments for its continued existence in light of efforts to remove IRC from CMC. As a result, her research is exemplary within the context of new forms of CMC, but does not offer any possibilities for future research, nor does it connect with any previous form of CMC research. It is, effectively, isolated from any other body of literature. + Page 40 A final example of completely isolated research can be found in Furlong's (1989) survey study of senior citizen CMC users at the University of San Francisco. There had been no significant differences between these senior citizens and other age groups who have been taught about computers. Her report examined the possible uses and gratifications of CMC for these people. Her study, which was simply a survey of an unknown quantity of individuals, attempted to determine the population of senior adults which was using the computer network, and what they hoped or expected of computers to improve or change their lives (Furlong, 1989). The rest of her report was a qualitative analysis of senior citizens' messages, nine in all, in order to establish exactly what type of computer communication exists for these subjects. All in all, she found that CMC is beneficial to this group, that CMC provides a "window to the world," and that the information age cannot forget this population (Furlong, 1989). Furlong attempted to apply the atheoretical model of uses and gratifications to CMC. Consequently, no theoretical advancement of CMC is accomplished. This "study" provides the most poignant evidence of how the problems befalling CMC "research." This work provided no review of literature, and had no tie whatsoever to existing research, indicating a "don't forget about me!" flavor. Interpersonal Communication via the Computer(1) From the earliest beginnings of computer message systems, one of the first and most influential discoveries was the complete lack of expressive (nonverbal) behavioral cues (Sproull & Kiesler, 1986). As a result, traditional forms of communication, such as the nuances of a conversation created with nods, smiles, eye contact, distance, tone of voice and other nonverbal behaviors, become "mystery" variables, causing much misunderstanding as a result of perceived meaning derived from context and the tone automatically attributed by a human's imagination. For example, Reid (1991) stated that users must and do compensate for the lack of these contextual cues. CMC users have managed to develop "ways of sending computerized screams, hugs, and kisses" (Reid, 1991). It is implied that these "emoticons," a hybrid name inferred from the words "emotive icons," are necessary for the comprehension of the meaning behind the message. Emoticons have developed into an artistic display of emotion and meaning, without which much of computer communication would be spent verifying the intent of each message. There are four different forms of emoticons, each with its own specific purpose. + Page 41 + First, there is the common practice of simply verbalizing physical cues. For example, a humorous comment is difficult to reply in genuine laughter, so the literal typing of "hehehe" is considered acceptable. Second, physical actions taken by CMC users are typically described within two asterisks (Reid, 1991). This form of expression usually is determined within the context of the communication, however, and is primarily utilized in on-line interactive situations, such as Relay and MUDs. For example, a hypothetical conversation between two individuals might read something like this: *Lady Marion arrives in the room with a flair and arrogance fitting of her status* *Bowing low in respect to Lady Marion's presence* How fair thee, Lady Marion? where the two individuals use pseudonyms to hide their true identity and are engaging in a role-playing game. In academic use, however, asterisks play an important part in terms of stressing what would have normally been a speaker's vocal emphasis. Here, an example would be: ... and I find it ludicrous that some system administrators would have the utter *gall* to censor material to a public forum. Such behavior is unethical, immoral, and I _won't_ stand for it. Here, the terms 'gall' and 'won't' are to be represented as stressed for emphasis to compensate for the lack of vocal ability. Likewise, the third form of visual expressive content includes the ability to stress, but with a subtle, yet important, difference. This is the inclusion of terms addressed in all capital letters. Here, the stress of the word is to be perceived as near-yelling, and quite frequently other computer users will comment as such. An example from a Relay episode: <2hot4U> HEY EVERYBODY, WHAT'S UP?!? 2hot: Why are you yelling? 2hot: Quit yelling! You're giving me a headache! <2hot4U> Sorry everybody, my caps lock stuck. Finally, the fourth form of emoticons is a "shorthand" for the description of physical condition (Reid, 1991). Commonly called "smileys," these characters placed together represent the sentiment within a message, or to lessen the impact of a sarcastic comment, or sometimes just for esthetic effect. Some examples of this are described by Reid (1991): + Page 42 + :-) or : ) a smiling face, as viewed side-on ;-) or ; ) a winking, smiling face :-( or : (an 'unsmiley': an unhappy face :-} or : }someone having a so-so day :-(*) someone about to throw up 8-) someone wearing glasses :-P someone sticking out their tongue >:-O someone screaming in fright, their hair standing on end :-& someone whose lips are sealed @}-`-,-`-- a rose [Author's note: Examine the icons by looking at the page 90 degrees counterclockwise] Overall, there are well over 500 of such icons, although only about five to nine of the ones represented above are in constant use. It would seem relatively obvious to trace back these icons to handwritten notes where smileys were drawn as circles with features. Nonetheless, the sheer quantity of these icons goes beyond such humble beginnings into a depth of imagination and creativity that defies such a simple explanation. It is Reid's (1991) hypothesis that such inventiveness and lateral thinking demands skill. An inference about the communicative competence of these users would be a logical conclusion, but she refrains from attempting to do so, leaving conclusions up to the reader. However, such creativeness implies that computer users have the ability to adapt to limited communication situations, and even substitute graphic symbolic cues to replace conversational shortcomings. Reid (1991) indicated that computer-mediated communication led to the behavior disinhibition due to the lack of regulating cues, yet she was not the first to do so. Sproull & Kiesler (1986) found that much of the information distributed through electronic mail was unique in that it could not be conveyed through another medium. As Adkins (1989) defines, deregulating is when "regulating" cues, such as speaking when spoken to or recognizing when one comprehends the message and moving on to another topic, are absent from a conversation. In email conversations, for example, it is not uncommon to witness the practice of inserting responses in the midst of the original message, addressing each point as it is presented. More drastic, however, are realtime conversations, where individuals can send single-line messages across the network to each other. Empirically, it is feasible to conduct more than one conversation at the same time with the same person, due to the time differential in receiving feedback to statements. One individual might move on to another topic, and then receive a response to a statement sent several minutes before, and respond to the previous topic. Maintaining cognitive awareness of each conversation becomes a balancing act at which some CMC users become quite adept. + Page 43 + Even so, the tendency to deregulate behavior has concerned several researchers (Adkins, 1989; Kiesler, Siegel, & McGuire, 1984; Reid, 1991; Sproull & Kiesler, 1986). Reid (1991) has gone so far as to say that Relay encourages disinhibition among its users. The lack of social context cues are read as an obscuring of the boundaries which delineates the forms of behavior which are acceptable or unacceptable. She determined that "with little regulating feedback to govern behavior, users behave in ways that would not generally be acceptable with people who are essentially total strangers" (Reid, 1991). Possible future research along this vein might examine the motivation behind such behavior. Not all uninhibited behavior or communication is so negative, however. Many researchers (Adkins, 1989; Kiesler, Siegel, & McGuire, 1984; Reid, 1991; Sproull & Kiesler, 1986) indicate that there are advantages to disinhibition. For example, the anonymity associated with conversing with a terminal between the participants is a boon for those people who suffer from shyness (Reid, 1991). Computer-mediated communication can allow users the freedom to "open up" to a complete stranger, disclosing information that one would hesitate to mention tete-a- tete to an individual of the same acquaintance level. In fact, an uncommon, but nonetheless true, phenomenon is the emergence of what may be called "net.romances." Here, self- disclosure among CMC participants results in attachment and, ultimately, involvement. Going beyond even the professional involvements, which in and of themselves are most notable, these relationships, which hold all the subtle nuances and difficulties that close proximity relationships have, are long distance romantic relationships carried out over e-mail or Relay. As Reid (1991) warns, "such expressions of feeling are not in any way thought to be shallow or ephemeral." Relay has been known to create and foster extremely emotional bonds between people, eventually leading up to successful marriages. These romances are indeed computer-mediated interpersonal relationships at their most idyllic. A seedier side of the communication medium also exists, with little attention paid it due to its nature. However, the sexuality which pervades CMC is a force to be reckoned with. Human sexuality seems to bloom outward over the computer, with little appearance of being hindered by system administrators. Only one researcher (Reid, 1991) has acknowledged its existence, attributing the label "net.sleazing" to the attitudes and situations involved. It is also known by "CompuSex." or "link-sex," where descriptions of sexual behavior are graphically depicted, written in the first person and directed towards one another. Whether this is a result of disinhibition or a cause of it is uncertain, as are the effects of this behavior on attitudes and perceptions of each other. Given the frequency of this behavior, sexual communication via the computer would be a rich area of future research. + Page 44 + Conclusions and Future Research Questions The lack of an underlying theory regarding CMC research makes prospective research questions difficult to ascertain, since much of what has been studied has little or no continuity (i.e., Baron's study of computer linguistics). Perhaps there is no theory of CMC because all other theories support different models of communication. CMC's "models" are delineated by the structural definition of the programs which support electronic communication. Therefore, an electronic model replaces a theoretical one, and a theory, up until now, has not been seen as necessary. CMC theory has been usurped by the concreteness of the program "models," thereby placing CMC under contextual formats rather than theoretical ones. For example, the models are used to place an image, or method, of examining a specific context of communication. In CMC, however, such concreteness is provided by the actual software, and not by imagination. As a result, there has never been a perceived need to organize a theory. The research thus far has indicated that this not need be the case. If future studies examined CMC not as part of a broader set of communication (e.g., organizational communication research), but rather as its own context, a major step will be taken in grasping the gestalt of CMC. The first step in achieving this is recognizing that CMC does *not* equal electronic mail, a serious mistake made by nearly all early CMC researchers. Electronic mail itself is but one small subset of CMC, and is not all inclusive. CMC itself is made up of two distinct classes of communication: synchronous and asynchronous. Sproull and Kiesler (1986), among others, made that mistake when defining all CMC as asynchronous. Such is not the case. Asynchronous CMC, such as email, and the findings resulting from studies of such material, cannot explain the occurrences of culture within ReidUs (1991) analysis of Internet Relay Chat, a synchronous form of CMC. While the two are both CMC, as well as legitimate forms of communication, the results of the former cannot begin to address the issues brought up by the latter. In this way, CMC studies should address the medium first, in order to understand the consequences of its two subsets (e.g., asynchronous and synchronous forms of CMC). Only at that point in time can the full scope of CMC interactions been completely revealed. As indicated earlier, one of the main criticisms of CMC research is that the majority of findings cannot be used to explain the empirical experience across CMC platforms. + Page 45 + At certain points in CMC research history there have been points of contact where researchers have managed, or at the very least attempted, to tie up many loose ends in nearly exhaustive analyses of existing research (Chesebro & Bonsall, 1989; Steinfield, 1986). It is from there that we should re-address the questions they raise. For example, examination of electronic bulletin boards has not increased since Steinfield's (1986) essay, although the usage of them have increased exponentially since that time. It is also necessary to examine the implications of multiple-organizational research involving CMC. Much of the emerging research concerns the controversial topic of CMC cultures (Reid, 1991), and a combination of research in these areas could facilitate the continuity which is desperately needed. On a positive note, the myths and superstition regarding computers and computer technology appears to have been phased out of research involving CMC. While this may appear to be inherently logical and self-evident, it indicates that a shift has taken place; computers are no longer regarded in a mystical, glorified manner. As has been indicated by Chesebro and Bonsall (1989), there is now a notion of responsibility which permeates human-computer based communication. Perhaps with this theme, researchers can take responsibility of their studies and guide them toward an underlying theory which up until now has been largely undocumented. Cathcart and Gumpert (1986) provide a starting point with their discussion of mediated interpersonal communication, which could have lead directly into a method of examining CMC. According to the authors, definitions of communication "have minimized the role of media and channel in the communication process. The focus has been on the number of participants, source and receiver relationships, and forms and functions of messages. The media of communication have been accepted, more or less, as fixed or neutral channels for the transmission of messages among participants" (Cathcart & Gumpert, 1986, pg. 27). It is important that research emphasize the role of the individual in the mediated communication process rather than on the role media plays in shaping interpersonal behavior (Cathcart & Gumpert, 1986). Even so, all too often the term + Page 46 + "media" is used synonymously with the term "mass media," when the two are distinctly separate. As a result, the authors attempt to bridge the definitional gaps and reconcile the role of media in human communication by making some generalizations: 1. There are interpersonal situations which require media for the purpose of communication. 2. Second, the media are part of a complex of variables that influence behaviors and attitudes. 3. The content of media is both a reflection and projection of interpersonal behaviors. 4. An individual's self image and its development is media dependent (Cathcart & Gumpert, 1986, pp. 27-8). It is then argued that if the above claims accurately reflect the realities of the media, in this case CMC, then what would be needed is a new topology which would include media technology. First, research needs to address the possibility that heavy users of CMC (including relay and chat lines) have modified the symbolic realm of communication. As indicated by Reid (1991), symbolic emotive messages have been established in order to compensate for the lack of social contextual cues. Also, users have abbreviated traditional sentence structures, almost forming a separate language code. It will be important to determine exactly how these modifications affect the perceptions of the users. Second, and more importantly, it needs to be determined whether CMC users, Relay users in particular, are less proficient at interpersonal communication than other people who do not use the computer as often, or at all. Because research has delved only slightly into the area of behavioral modification due to computer use, the next logical step would be to determine what behaviors might arise, and how those behaviors affect communication both via the computer and interpersonally. At that point we should be able to take the concept of computer communication and its effects out of fiction and regard the importance of this new area with the attention it deserves. + Page 47 + As has been stated previously, CMC research falters primarily when examined as a "tool," rather than a context in itself. For example, the telephone has become a likely metaphor when describing CMC interactions. However, such a paradigm seriously limits the way in which CMC is analyzed. If, for example, CMC is examined within an organizational communication setting, only the task-oriented aspects will get analyzed; social functions of CMC become ignored. If examined under the social aspect of Internet Relay Chat, only the conjectures of social behavior will be examined. Instead, it is imperative that CMC become examined as a context, an umbrella under which electronic mail, computer conferencing, MUDs, and Relay/Chat lines would be examined. The advantages of such a perspective are obvious. First, it would be easier to theorize communication patterns across each individual subcontext of CMC (e.g., from electronic mail to computer conferencing, from MUDs to Relays). Second, there would be predictability across each CMC sub-context. Due to the high amount of change involved when examining CMC, as in the constant changing, modification, and updating of computer communication software, current research has been limited in terms of this predictability power. Once a study has been published, the conclusions and predictions were no longer valid either because a) the program being used is not universally utilized in all businesses or universities, or b) the program has been modified, undermining the assumptions on which the conclusions are based. Perhaps the best way to examine such a new context is under the human action perspective. What rules do users of CMC follow, both written, unwritten, and within the limitations of computer programming? Communicators have been able to begin conversations with other users, and while waiting for a response, begin a second conversation with that same person by asking another question, or initiating a change in conversation while continuing the original conversation. Under a rules perspective, what rules do users create and follow in order to maintain conversations which multiply, even between the same two conversants? In order to demonstrate how CMC is distinct within its own context, we must first acknowledge that it is not merely a channel through which people communicate. CMC is unique in that it is perhaps the only medium that its users change the very nature of communication. A perfect example is the aforementioned + Page 48 + interspersing of conversations within conversations during realtime communication. In no other form of communication do conversants 'flip-flop' topics with such rapidity. In addition, there are the rare cases (yet not improbable) where an individual's programming ability can modify the very program which people use to communicate. Also, it is important to note how users of CMC perceive that medium. In that sense, it would be ridiculous to force-fit a telephone paradigm onto a medium whose users wouldn't look at it as such. Users of CMC change their medium(2), evidenced not just by the advent of emoticons, but also the choice of immediacy and interaction. Both asynchronous email and synchronous Relays are CMC, yet the level of interaction varies between the two. In no other medium does communication become subject to personal modification; in what way does the viewer change the way a message is broadcast over television? How many different ways are there to write a letter? Most importantly, treating CMC as its own context raises different communication questions than if used as a subset of another context. For example, under organizational theory, CMC is used under two criteria: Task-related use and social-use. As shown in this paper, not all computer mediated communication can have both elements. In interpersonal communication, for example, under standard theory criteria, the nonverbal cues provided by emoticons would be ignored, since they aren't "true" nonverbal cues. There is no current system of analysis which can completely cover all aspects of CMC. The reason for this is because computer mediated communication has come into its own as a valid form of communication, and is permeating a global society. It would be foolish to ignore the potential for this medium by underestimating its value. Notes (1) Unfortunately, for the most part, little has been published concerning interpersonal communication via the computer. That is not to say, however, that little has been written. In fact, as previously stated, much of the research involving human behavior, interpersonal communication, and communication subcultures has been presented at conferences. While this implies that awareness and consciousness of the phenomenon is increasing, the availability of such material leaves much to be desired. + Page 49 + (2) Here it is crucial to make a distinction between CMC-as- medium and computer-as-medium. One of the common assumptions which have been in error prevalent in the above- mentioned sources, is that the computer is the medium through which people communicate. This simply is not accurate. Communication via the computer is *dependent* upon the software available and chosen. As a result, the *medium* of CMC is changed by the users based upon form, immediacy, and interaction. References Baron, N. S. (1984). Computer mediated communication as a force in language change. Visible Language, 18, 118-141. Cathcart, Robert, and Gumpert, Gary. (1986) Intermedia: Interpersonal Communication in a Media World. New York: Oxford University Press. Chesebro, James W., & Bonsall, Donald G. (1989). Computer- Mediated Communication. Tuscaloosa, Alabama: University of Alabama Press. Furlong, Mary S. (1989). An Electronic Community for Older Adults: The SeniorNet Network. Journal of Communication: 39, 145-153. Gratz, R. D., and Salem, P. J. (1984). Technology and the crisis of self. Communication Quarterly, 32, 98-103. Kell, Jeff. (1988). Relay: Past, Present, and Future Paper presented at NetCon, New Orleans, LA. Metcalfe, B. (1992, September 21). Internet fogies to reminisce and argue at Interop Conference. InfoWorld. p. 45. Reid, Elizabeth M.(1991). Electropolis: Communication and Community on Internet Relay Chat. Unpublished thesis, University of Melbourne. Rice, R. E. and Case, Donald.(1983). Computer-based messaging in the university: A description of use and utility. Journal of Communication, 33: 131-152. Rubinyi, Robert M. (1989) Computers and Community: The Organizational Impact. Journal of Communication, 39: 110-123. Sproull, L., and Kiesler, S. (1986). Reducing context cues: electronic mail in organizational communication. Management Science, 32, 1492-1512. Steinfield, C. (1986a). Computer-mediated communication in an organizational setting: Explaining task-related and socioemotional uses. In M. McLaughlin (Ed.), Communication Yearbook 9. Beverly Hills: Sage. Steinfield, C. (1986b). Computer-mediated communication systems. Annual Review of Information Science, 21, 167-202. + Page 49 + --------------------------------------------------------------------------- Interpersonal Computing and Technology: An Electronic Journal for the 21st Century Copyright 1994 Georgetown University. Copyright of individual articles in this publication is retained by the individual authors. Copyright of the compilation as a whole is held by Georgetown University. It is asked that any republication of this article state that the article was first published in IPCT-J. Contributions to IPCT-J can be submitted by electronic mail in APA style to: Gerald Phillips, Editor IPCT-J GMP3@PSUVM.PSU.EDU