"Miss Kelly said that when you talk to somebody it's like you're playing ball. First the somebody asks you a question, and that means they throw the ball to you. But you have to do more than just catch a question like you catch a ball. Here's the important part. You have to throw the ball back. When somebody asks how you are, you can't just say, 'Fine.' You say, 'Fine, thank you, and how are you?'"
"What does this have to do with ?"
"Everything," I said. "Miss Kelly said you have to throw the ball back. So I threw it back, and by mistake the ball hit Miss Boland." (Peck, 1974, p. 5f)
Vygotsky (1978) writes that " children solve practical tasks with the help of their speech, as well as their eyes and hands" (p. 26). In Vygotsky's view, speech is an extension of intelligence and thought, a way to interact with one's environment beyond physical limitations:
the most significant moment in the course of intellectual development, which gives birth to the purely human forms of practical and abstract intelligence, occurs when speech and practical activity, two previously completely independent lines of development, converge (p. 24).
This higher level of development enables children to transcend the immediate, to test abstract actions before they are employed. This permits them to consider the consequences of actions before performing them. But most of all, language serves as a means of social interaction between people, allowing "the basis of a new and superior form of activity in children, distinguishing them from animals" (Vygotsky, 1978, p. 28f).
The ability to use language to help solve problems is a tool. Rather than trying to understand the world alone, a child can enlist the help of older children, adults, or other authorities. As a result, Vygotsky believed that a child's potential should be measured not merely in terms of what a child already understands, but should include the child's capacity to profit from what others can help the child to understand (Spencer, 1988; Vygotsky, 1962). This difference between what one can do and one's potential to engage the help of others and profit from it Vygotsky called the zone of proximal development, "the distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance or in collaboration with more capable peers" (Vygotsky, 1978, p. 86). The more children take advantage of an authority's support, the wider is their zone of proximal development and, ultimately, their own capacity.
For example, I have never studied Japanese. If I were tested on the subject today, I would do very poorly. One might infer, based on those results, that my Japanese ability was very poor. However, if I were to enrol in a Japanese course -- enlist the help of others to make me a better Japanese speaker -- another test might indicate that I am rather good at the language. My ability to learn Japanese is the same as it ever was. What is different is the inclusion of my zone of proximal development -- my use of the knowledge of others to change my understanding. On the other hand, even with the help of others, I might still be unable to grasp the language. If this were the case, my zone of proximal development (at least for Japanese) would be small; my ability to use an authority's support to learn Japanese would be near zero.
Vygotsky "viewed intelligence as the capacity to benefit from instruction, with language having a powerful developmental role" (Spencer, 1988, p. 170). In this sense, language is a tool for learning and an aid to understanding. Writes Vygotsky (1978), "human learning presupposes a specific social nature and a process by which children grow into the intellectual life of those around them" (p. 88). As such, language acts as a vehicle for educational development and is important for the apprehension and acquisition of knowledge.
Vygotsky (1978) maintained that the zone of proximal development is an "essential feature of learning" (p. 90), in which one builds on one's knowledge through interaction and co-operation with one's peers. In this sense, the authority or teacher in all learning situations acts as a collaborator and coach, in which he or she "provides scaffolding to lead the student to increased understanding" (Hawisher, 1994, p. 44). In this respect, the basis of education is people interacting with other people (Daniels, 1996; Shale, 1988; Shale & Garrison, 1990).
In the educational context, language is important for comprehension and making use of knowledge. Shale (1988) describes the role of the teacher in the "ideal educational process" (p. 28) in four parts: First, the teacher and the student determine and validate what the student knows. Second, on the basis of what is determined, the teacher may provide additional declarative knowledge. Third, the teacher and the student negotiate the meaning of what is taught. The assumption is that the teacher will clarify points for the student, but in the best exchanges the teacher also gains understanding (Shale & Garrison, 1990). Fourth, through repetitions of steps two and three, both the teacher and the student advance in their knowledge, and the student's knowledge is validated by the teacher.
The zone of proximal development is observed during this third step of the schooling process, in which teachers help "others to gain consciousness and reach higher ground intellectually, transforming the meaning of the lower order concepts" (Spencer, 1988, p. 176), also (Schaffer, 1996; Vygotsky, 1962). In this step there is "room for the negotiation of meaning and the prospect of mutual learning through dialogue and discussion" (Rowntree, 1975, p. 284). In an ideal form of education, the teacher and student engage in what King and Brownell (1966) refer to as "The Great Conversation."
Unfortunately, a great deal of actual schooling time is spent conveying information, rather than ensuring comprehension. This often precludes the formation of an interactive learning environment in which learning is an ongoing process shared between the teacher and students. Hodge (1993) notes that in schooling the difference in knowledge between teachers and students is so great that it simultaneously justifies and impedes the educational process. In the limited amount of time in a class, teachers conveying information do so at the expense of negotiation of meaning. However, when teachers are negotiating meaning, they are not providing declarative knowledge. To use Shale's (1988) description of the educational process, steps two and three are in opposition; one may have more of one, but only by having less of the other.
In the last 30 years, a number of educational researchers have begun to emphasise the role of language in learning, particularly the role of talk in the classroom. The disparity between the amount of talk performed by teachers and students was often seen as an hindrance to learning (Barnes, 1971; 1976; Bellack, Kliebard, Hyman & Smith, 1966; Britton, 1970; 1971; Brown, Anderson & Shillcock, 1984; Bullock, 1975; Cazden, 1988; Flanders, 1970; Hodge, 1993; Sinclair & Coulthard, 1975). As Kingman (1988) noted:
In addition to encouraging the development of speech for communication, teachers need to encourage talk which can be exploratory, tentative, used for thinking through problems, for discussing assigned tasks, and for clarifying thought: talk is not merely social and communicative, it is also a tool for learning (p. 43).
Despite these concerns, however, the nature of teacher-student language in the classroom has remained largely the same (Dillon, 1985; 1994; Hodge, 1993; Jones, 1988).
Within the realm of education, there are three basic types of formal classroom discourse. Over time, it is possible for any given course to include all three of these forms of interaction. It is possible for a course to be taught by means of any of these modes of interaction, or by any combination of them. Each mode has its own distinctive traits, as well as advantages and disadvantages.
The first and least-interactive mode of teaching is the lecture. In a lecture, material is conveyed in what is intended to be one-way uninterrupted discourse, as though delivering a speech (Hills, 1979). Although it is seen in face-to-face (FTF) situations, this is also used to describe the delivery of content which cannot be questioned or altered, such as books, radio, television, audio tape, videocassettes, and some forms of multimedia. In a classroom, lecturing by a teacher would be performed by reading a lesson while soliciting no interaction from the audience.
In this form of teaching, there is no overt negotiation of meaning with the teacher. Rather, students participate in an "internal didactic conversation" (Holmberg, 1986) during which they interact with course materials and "talk to themselves" about this new information and ideas. Lewis (1975) explains that when people ponder what they have learned in solitude, they are actually having a conversation with themselves. In the realm of distance education, this interaction with the course content is described as learner-content interaction. In his analysis of different kinds of interaction which must be acknowledged in distance education, Moore (1989) places it as the "defining characteristic" of education: "Without it there cannot be education, since it is the process of intellectually interacting with content that results in changes in the learner's understanding, the learner's perspective, or the cognitive structures of the learner's mind" (p. 2). Although researchers -- myself included -- have acknowledged the importance of this intrapersonal communication between a student and the instructional content (Dillon & Gunawardena, 1995; Hillman, Willis & Gunawardena, 1994; Holmberg, 1988; Moore, 1989; Wagner, 1994), it does not and cannot replace the teacher for validation and negotiation of learning.
Although the lecture format is effective for disseminating information, it does so at the expense of validating this knowledge and making it meaningful to the student (Shale, 1988; Shale & Garrison, 1990). Although interaction appears to be occurring between the student and the content, it is actually a counterfeit form of interaction (Button & Sharrock, 1995) aping a collaborative environment. In a Vygotskian sense, the zone of proximal development may be entered by the use of lecture-style means, such as books or television programs, but the limitations of the medium restrict the amount of guidance and collaboration that can occur. Holmberg (1988) observes that although pre-packaged materials for distance education can represent a kind of "simulated communication," it is the interaction between humans that "represents real communication" (italics in original, p. 116).
The most typical form of classroom interaction is recitation, which has two predominant characteristics. First, the teacher is the predominant speaker. Although students are now permitted to interact with the teacher, the teacher will guide and control the class by means of asking questions, giving instructions, and giving information (Edwards & Furlong, 1978; Hodge, 1993; Sinclair & Brazil, 1982; Sinclair & Coulthard, 1975).
In most schools, on average the teacher speaks for 59% to 69% of the time (Dillon, 1985; 1994). Kramarae and Treichler (1990) report that it is typical for teachers in college classrooms in the United States to speak for 75% of the time. Bellack, Kliebard, Hyman & Smith (1966) found in their experimental social studies classes, taught to seventeen-year-olds, that teacher speech varied from 60% to 93% of all classroom discourse, with the median at 73%. These numbers are similar in the United States (Dillon, 1985; 1994) and the United Kingdom (Barnes, 1976).
This classroom domination is evidenced in the artificial interactions that take place in the classroom. Edwards and Furlong (1980) describe the educational process as a performance, in which the students' collective attention is focused on the teacher. Adams and Biddle (1970) explain that " despite the presence of thirty or more potential communicators, what has been called a 'central communication system' is frequent and often prolonged." This invariance inspired Flanders (1970) to devise the "two-thirds" rule: two-thirds of every class is made up of talk, and two-thirds of the talk comes from the teacher. This centralised communication is reinforced and maintained by means of rhetorical techniques such as responding to questions by asking another question, traditionally a technique used by teachers (Gere & Stevens, 1985).
The second characteristic of recitation is that the interaction between the teacher and students will follow a regular pattern. The teacher will initiate some form of action, usually a question, the student will respond, and the teacher will acknowledge the student's response (Atkinson, 1981; Dillon, 1985; 1994; Hodge, 1993; Mehan, 1978; Sinclair & Brazil, 1982; Sinclair & Coulthard, 1975; Stubbs, 1983b). This mode of interaction is described by Sinclair and Coulthard (1975) as Initiation-Response-Feedback. They propose that it is the quintessential teaching exchange: (teacher's) initiation, (student's) response and (teacher's) feedback (Stubbs, 1983b). The last stage is also known as evaluation (Mehan, 1978).
These two characteristics of recitation are interrelated. Since the teacher is controlling the class by means of initiation and feedback -- two-thirds of the turns -- he or she will necessarily do most of the talking (Atkinson, 1981). When the student asks a question, however, the structure is reduced to initiation-response since students do not "overtly evaluate teachers' answers" (Stubbs, 1983b).
From a Vygotskian viewpoint, recitation is better than lecture because it includes "the provision of two-way communication so that the student may benefit from or even initiate dialogue" (Keegan, 1986, p. 49). This permits the student to enlist the help of the teacher to negotiate meaning of the course content (Cennamo, Abell, Chung, Campbell & Hugg, 1995; Gunawardena, 1991), and thus more fully engage his or her zone of proximal development.
The third form of classroom interaction is discussion. Like recitation, discussion has several observable characteristics (Dillon, 1994). The first attribute of discussion is that rather than being dominated by the teacher, the students do most of the talking. Unlike a lecture or recitation where the teacher will do all or two-thirds of the talking respectively, the students in a discussion will generate half or more of the talk (Dillon, 1985; 1994).
The second characteristic of discussion is that although the students are participating in the discussion, it will not follow the initiation-response-feedback model of recitation. Rather, there will be a mix of statements and questions by a mix of teachers and students (Dillon, 1994). For Shale (1988), this form of interaction is the ideal educational process for it permits the student to validate his or her " emerging knowledge through collaborative and sustained interaction with a teacher and other students" (Shale & Garrison, 1990, p. 29f).
Discussion is rarely used in classrooms, however. Although teachers believe that they are doing so, as indicated by self-report, external observation indicates that this is not the case (Alvermann, O'Brien & Dillon, 1990; Connor & Chalmers-Neubauer, 1989). This is unfortunate, because from a Vygotskian perspective, discussion is the best method of teaching. It fosters co-operative learning between all of the participants (Fowler & Wheeler, 1995), and reinforces the idea that the teacher's role is that of an active, communicative partner in learning, not merely a provider of a certain learning environment or one who enforces correct behaviours (Jones & Mercer, 1993).
The relevance of this form of communication supports the research of Amidon and Giammatteo (1967) who found that superior elementary-school teachers  were interrupted more by questions from students, were more accepting of student-initiated ideas, tended to encourage these ideas more, and also made more of an effort to build on these ideas than did the average group of teachers. Overall, there was about 12% more student participation in the classes of the superior teachers.
This increase in student participation also reflects the findings of Gabriel & Davey (1995), who interviewed student nurses who were taking their classes by means of distance education. They found that the students could learn elementary facts with no trouble with the lecture materials (self-study packs). For abstract or complex ideas, or issues other than facts, however, the students found FTF interaction with other students important and "particularly necessary when the work involves challenges to existing values and attitudes" (p. 500).
Although research has examined the role of peer collaboration in FTF environments (Gallimore & Tharp, 1990; Newman, Griffin & Cole, 1988; Nunes, Schielmann & Carraher, 1993; Tharp, 1993), only recently has appropriate technology become available to enough people to examine the role of peer collaboration via computer-mediated communication (CMC) in educational contexts.
Historically, distance education was based on a one-to-one (teacher-student) model of correspondence study. This was primarily a response to the limitations of the instructional delivery systems (Barker, Frisbie & Patrick, 1989; Garrison & Shale, 1987; Shale, 1988; Shale & Garrison, 1990; Sherow & Wedemeyer, 1990; Verduin & Clark, 1991). Newer forms of technology have helped to change this, however, permitting the addition of student-to-student interactions (Moore, 1989).
Although research has been performed to examine the effects of peer collaboration via CMC (Hartman et al., 1995; Hiltz, 1990; Hiltz & Turoff, 1978; Hunt, 1995; Kiesler & Sproull, 1992), what has been missing is a tool to examine and compare classroom interaction across modes of delivery, as well as within, while taking into account the characteristics of each mode of interaction. With such a tool, one would be better able to study these modes of delivery's abilities, and to determine how they can best be used to promote interaction and encourage collaborative learning.
New technologies and software have allowed me to construct such a tool. By using commonly available Macintosh software and the AppleScript programming language, one can now handle the large quantities of data necessary to study the differences between the two major forms of interaction, FTF and CMC. First, however, one must know what one is looking at.
It is not the purpose of this work to describe inequalities of students in terms of learning per se. This section examines the differences between students in terms of things which in an ideal world would not matter in an educational environment: innate talents or natural abilities in interpersonal interaction. For the purpose of examination the following attributes have been separated, though they are often interrelated. Although they may occur in both FTF and CMC contexts, they have been placed in the sections in which they are most likely to affect interaction.
In classroom interaction, teachers are always in a high-status position of power. Although this may not be desired it always exists due to their role as teacher (Ellsworth, 1989; Harrington, 1992; Wilshire, 1990). This mirrors other forms of FTF interaction. Social interactions are dominated by participants with high social status, such as managers in business settings. Bales, Strodtbeck, Mills and Roseborough (1951) report that in many groups, participation is unequal and the proportion of the participation can be predicted by group members' social position and personal competencies. In corporate settings, for example, supervisors speak more often than subordinates, males speak more than females, and people in the front of the room speak more than those in the back (Berkowitz & Bennis, 1961; Kiesler, 1986; Kiesler & Sproull, 1992).
People transmit information about their social status by means of physical communication and social artefacts, such as what they wear, how they stand and how loudly they speak. This information subsequently affects how much influence they will have with others (Dubrovsky, Kiesler & Sethna, 1991; Edinger & Patterson, 1982; Humphreys & Berger, 1981; Ichiyama, 1993; Kiesler, 1986; Kiesler, Zubrow, Moses & Geller, 1985; Patterson, 1982).
Reporting observations of brainstorming exercises done by executives in both FTF and CMC environments, Nunamaker, Applegate and Konsynski (1987) noted that the group members who participated in the CMC brainstorming session spent most of their available time entering comments into the computer conferencing system. During FTF interactions, however, discussions tended to be dominated by a few individuals. This was reported similarly by Dubin & Spray (1964) who found that in FTF interactions the amount a person talks has a high correlation with his or her prestige and social status.
An aspect of social interaction which is related to social status is that of sex  issues. Unlike the relative anonymity of CMC interactions that shall be discussed later, an individual's personal characteristics, such as physical appearance, race, and sex are apparent and potentially relevant in FTF interactions. As Bellman (1992) wrote, these are "metacommunicative properties of the meaning-contexts used for interpreting the value and validity of another's talk" (p. 60).
In the classroom, sex-based inequality manifests itself and is reinforced through two means. The first is through unequal interactions with the teacher. LaFrance (1991) explains that sex inequality comes about and continues by means of verbal and non-verbal messages in the interactions between teachers and students. Boys, for example, having called out answers without raising their hands were acknowledged whereas girls exhibiting similar behaviours were reprimanded. Teachers also interact more with boys than girls (Berk & Lewis, 1977; Etaugh & Hughes, 1975; French & French, 1984; Leinhardt, Seewald & Engel, 1979; Morse & Handley, 1985; Swann, 1992).
According to Sadker & Sadker (1990), women are less likely to be called on in college courses and when they do participate, they are more likely to be interrupted and less likely to be accepted or rewarded. This sort of unequal interaction reinforces differences between the sexes, "nudging" female students into "passivity, dependency, and silence rather than activity, autonomy, and talk" (LaFrance, 1991, p. 10).
The second manifestation of inequality is apparent through interactions with other students in the classroom. Sternglanz and Lyberger-Ficek (1977) explain that:
it has come to be taken for granted by many faculty and students alike that men will usually dominate in college classrooms and many researchers have confirmed that women students are less likely to be verbally assertive in co-educational settings (p. 5).
This domination by male students prevents female students from having an equal opportunity to present ideas for discussion. It also means that the women have a less than equal chance to critique ideas warranting such treatment.
These findings are similar to those of Kramarae and Treichler (1990), who interviewed students at a graduate-level humanities course at the University of Illinois. They found that female students complained that the teachers "dominated" the course, and were "judging" one's point of view. They conclude that female students are less apprehensive in environments in which learning is a communal activity shared fairly by the teacher and students.
One form of domination which appears to occur between males and females in FTF interaction is interruption. Besides merely displaying rudeness and a lack of respect for the speaker, interruptions permit interruptors to control the topic or flow of the conversation and thereby to control or dominate others (Greif, 1980; Zimmerman & West, 1975).
The research suggests that when conversing with females, males do more than their share of the interrupting (McMillan, Clifton, McGrath & Gale, 1977; Natale, Entin & Jaffe, 1979; Octigan & Niederman, 1979; West & Zimmerman, 1983; Willis & Williams, 1976). When conversing in same-sex groups, however, males interrupt males as often as females interrupt females (Beattie, 1981; LaFrance, 1981; Marche & Peterson, 1993; Roger & Schumacher, 1983; Rogers & Jones, 1975).
Anxiety is a cognitive and affective response characterised by apprehension about an impending, potentially negative outcome that one thinks one is unable to avert (Schlenker & Leary, 1985, p. 172). In the case of social anxiety, the potentially negative outcome is an undesired evaluation. Grint (1989) notes that a "critical block to participation" in the FTF classroom appears to be this fear of public ridicule. This supports the literature which concludes that people who are highly anxious in social environments are less likely to initiate conversations with other people, speak less often and for a lower percentage of the time. They are also less likely to break silences in the conversation or to disagree with others, while they are more likely to avoid topics and factual matters that might reveal their ignorance, and to reveal less information about themselves. (Arkowitz, Lichtenstein, McGovern & Hines, 1975; Borkovec, Fleischmann & Caputo, 1973; Cheek & Buss, 1981; Glasgow & Arkowitz, 1975; Held, 1987; Leary, 1983; Leary, 1988; Mansbridge, 1983; Murray, 1971; Pilkonis, 1977; Slivken & Buss, 1984; Sniderman, 1974).
High levels of social anxiety are associated with social withdrawal, the avoidance of social situations that "portend possible self-presentational difficulties" and with "prematurely leaving such situations when they are encountered" (Schlenker & Leary, 1985, p. 179). Unfortunately, in a classroom environment, this behaviour may manifest itself as a reluctance to engage in classroom dialogue, or even removing oneself from the class altogether (Brown, 1970; Brown & Garland, 1971; Cheek & Buss, 1981; Grint, 1989; Phillips & Santoro, 1989; Pilkonis, 1977; Twentyman & McFall, 1975; Zimbardo, 1977).
In an experiment of social interaction (Alden, 1986), subjects were asked to talk FTF with a confederate under low-self-awareness conditions (unaware of being observed) and high self-awareness conditions (told that they were going to be observed as well as having a video camera visible). Subjects were instructed to indicate when they had completed their tasks by pressing a button. Subjects that reported themselves as manifesting low-efficacy traits withdrew from interactions more quickly than did high-efficacy subjects when self-awareness was heightened.
CMC, however, appears to be a way to circumvent these difficulties (Grint, 1989; Mabrito, 1991; Sproull & Kiesler, 1986; Sproull & Kiesler, 1991; Zuboff, 1988). Phillips and Santoro (1989), comparing FTF with CMC interactions in a classroom environment, observed a significant difference between the types of interaction in that students could ask questions via CMC without "publicly embarrassing themselves" (p. 159f).
CMC differs from FTF interactions by means of the characteristics and capabilities of the medium itself. As a result, the nature of the interpersonal interactions that develop as a function of those characteristics is often changed.
In interpersonal interactions, not all people are created equal, and subsequently neither are the nature of their communications abilities. Grint (1989) explains that participation in discussion and decision-making groups "tends to be dominated by those embodying institutionalized power, or by the more articulate and less inhibited participants" (p. 189).
These more articulate participants have been described by Sproull and Kiesler (1991), who quote a laboratory director who categorises his scientists into two groups: "leapers" and "plodders." The leapers tend to dominate FTF meetings "because they think quickly on their feet, are witty, and love the punch and counterpunch of intellectual debate." Plodders are less likely to contribute ideas in the FTF interactions, as they prefer to "go back to their offices and think through the implications of an idea" (p. 16). If this were this a classroom situation, it is probable that these scientists' voices would not be heard; they surely wouldn't be viewed as active participants. With asynchronous CMC , however, the plodders are able to share their analyses with everyone in the lab using a laboratory-wide e-mail distribution list, and in this way are "just as influential" (p. 16) as the leapers .
One of the advantages of asynchronous interactions is that they can help to narrow the differences between the leapers and plodders. Even for those individuals who would be inclined to present their ideas in a FTF meeting, as Harasim (1987) notes, " an individual can finish thoughts without fear of interruption by some keen, more outgoing colleague." Bruce and Shade (1994), describing a course taught via compressed video , noted that "chiming in" with a question or comment by the teachers or students "bordered on a competition." This is similar to the findings of Phillips and Santoro (1989), who describe students participating in a course via asynchronous CMC as able to contribute "without having to fight their way into the discussion" (p. 160). Admittedly, this is limited by such confounding variables as a given participant's general speed of thought, access to technology to upload comments, typing speed, and available time, but in comparison to FTF interactions it makes the field a bit more level.
Asynchronous benefits can cut both ways. The technology of CMC permits the quick turnaround of responses and some may be tempted to do so. One may respond immediately, and one's message is transmitted almost instantly. This encourages quick responses that may be more similar to patterns of speech than those of writing (Owen, 1992). On the other hand, since the material is written and thus available for re-reading, plodders have an opportunity to point out the flaws in a poorly-prepared argument that they would not have in a FTF environment. This is not to say that one must take more time to compose a response in a CMC environment. The point, however, is that speed is not a requirement.
Another advantage of asynchronous communications is that participants are free to contribute when it's convenient for them and are not restricted to a certain schedule (Ehrmann, 1989; Harasim, 1989; Hiltz, 1986; Lewis, Whitaker & Julian, 1995). In an educational situation, plodders can take as much time as they need or want to prepare a response. This increases equality between students, which in turn improves the quality of educational interactions. Learning is an interactive process, and any impediment to classroom interaction is necessarily a barrier to effective instruction. Ellsworth (1989) explains that in optimum learning conditions, "all members have equal opportunity to speak, all members respect other members' rights to speak and feel safe to speak, and all ideas are tolerated and subjected to rational critical assessment against fundamental judgments and moral principles" (p. 314). This mirrors Harrington (1992) who writes that "Communicative competence can only be achieved when dialogue is not dominated" (p. 72).
Furthermore, active participation enhances one's commitment to and satisfaction with group activities. Forsyth (1983) and McGrath (1984) describe a positive correlation between the amount a subject talks and satisfaction with and commitment to his or her group. The more one participates, the greater involvement with the organisation one perceives. Hiltz (1986), in her description of CMC courses offered by the Electronic Information Exchange System at the New Jersey Institute of Technology, reported that the only strong correlations between measures of perceived greater interaction with other students were feeling more involved and the perception of having learned more.
Social presence refers to how one is perceived by others. This is often described in fairly vague terms, such as charisma. A strikingly handsome man can often command attention simply by virtue of his physical presence; a very wealthy woman can achieve the same thing by displaying her financial power, and the presence of a teacher can affect students' motivation (Bruce & Shade, 1994; Christophel, 1990; Gorham & Zakahi, 1990; Hackman & Walker, 1990; Sigel, 1991). All of these are forms of social presence. Bales (1950) describes this as the "Actor's range of symbolic manipulation and process of overt action" (p. 44). A decrease in a participant's social presence changes how he or she will be interpreted; an unwashed street urchin has reduced social presence when viewed across the street compared to when he or she is plucking at one's coat. In a similar manner, orators with a mastery of content-free rhetoric may have their limitations pointed out in a computer conference when the social presence -- charisma -- of their oratory is unable to be communicated via CMC.
Short, Williams and Christie (1976) note that lacking the dynamic personal information of FTF or telephone communication, people focus their attention on the words in the message rather than on the messenger. "Group decisions made via CMC are unpredictable, unconventional, democratic, and less constrained by high-status members" (Sproull & Kiesler, 1991, p. 66). Moreland & Levine (1982) note that it's easier to join and be socialised to electronic groups than to FTF ones because physical invisibility decreases the potential stress associated with the newcomer identity.
In addition to increasing the equality of interactions between the participants themselves, CMC thus forces one to focus on the content of a participant's contribution, rather than the person. As is often explained on the USENET newsgroups, "Here on the 'net, you are only what you write." Describing a CMC decision-making environment, Nunamaker et al. (1987) noted that "The status, authority, and roles of the group members were divorced from the comments so that each comment was evaluated on its own merits rather than being evaluated in light of the person who made the comment" (p. 11).
This lack of social presence works like alcohol in a social situation -- it relaxes participants and yet it also may lead to impolite behaviour. Felson (1980) proposes that shared cultural norms which encourage politeness in social transactions hinder the direct expression of feedback to others, particularly if it's negative. People do not typically communicate their appraisals of others directly. With the reduction of social presence, however, "there are few reminders of status differences, the fear of evaluation or criticism declines" (Sproull & Kiesler, 1991, p. 88). Siegel et al. (1986) report that in a decision-making task, three-person groups exhibited swearing, insulting, and name-calling behaviour 34 times during CMC sessions, but never in FTF interactions.
Schrum (1992) reports that in day-to-day FTF interactions an estimated 90% of interactional information comes from non-verbal cues. Due to the text-based interaction of CMC, however, social cues, the means by which one ascertains another person's status and state, are reduced. This reduction of social cues changes the nature of the conventional distribution of power. As a result, high-status people do not dominate discussion in CMC groups as they do in FTF situations (Harasim, 1990; Hartman et al., 1995; Siegel et al., 1986). Note Kiesler, Siegel and McGuire (1984), "Charismatic and high-status people may have less influence, and group members may participate more equally in computer communication."
In a decision-making experiment, McGuire et al. (1987) found that when groups of executives met FTF, the males in the group were five times as likely as the females to make the first decision proposal. When the same groups met via CMC, the females made the first proposal as often as the males.
Huff & King (1988) performed an experiment in which they examined the interactions of pairs of students, each comprised of an undergraduate and a graduate. They found that when the pairs met FTF to decide the topic of a joint project, the pairs were more likely to choose the topic preferred by the graduate student. When similar pairs of students met via CMC, however, they were equally likely to choose the topic initially preferred by the undergraduate.
Zuboff (1988) reports that participants involved in CMC generate their own status based on both their helpfulness and the quality of their contributions. One participant, comparing the difference between computer conferencing and FTF interaction, explained that "lots of people have power that is not knowledge-based: it is forceful and based on their personality or position." In the CMC environment, however, "the power lies in the ability to communicate and pass on knowledge. I have extended my power base through my knowledge rather than through intimidation or style. It is strictly now the quality of your ideas, the way you put things in words, or your sensitivity to what others say that now determines your influence" (p. 371). One might retort, however, that in CMC one's writing style becomes an important social cue in its own right.
Melody (1981) and Poster (1990) state that communications media alter the forms of social organisation, create new patterns of association, develop new forms of knowledge and often shift the centres of power. In addition, Kiesler (1986) explains that "Because computers break down hierarchies and cut across norms and organization boundaries, people behave differently when using them." Lending support to this hypothesis, McGuire, Kiesler and Siegel (1987) report that in decision-making tasks, less negotiation took place before an initial solution was proposed via CMC, and yet group members were equally confident of choices made in FTF and CMC.
In addition, CMC makes it possible for participants to break free of traditional interactions:
Those who regard themselves as physically unattractive reported feeling more lively and confident when they expressed themselves in a computer conference. Others with soft voices or small stature felt they no longer had to struggle to be taken seriously in a meeting (Zuboff, 1988, p. 371).
The reduction of social cues caused by the limitations of CMC "can be expected to encourage open input of creative ideas, discovery of optimal solutions, and selection of an alternative based on its merits rather than on compromise" (DeSanctis & Gallupe, 1987). Indeed, due to the nature of CMC, ideas must be examined with little reference to their creator for as Boshier (1990) notes, "Nobody cares about or even knows if the originator of messages is wearing a pinstripe, clown or birthday suit" (p. 52).
Although this may be viewed as a double-edged sword, participants communicating via CMC are generally less influenced by social conventions. Sproull and Kiesler (1991) explain that in comparison to FTF interactions, " electronic communication will be relatively franker and will demonstrate relatively less audience awareness."
Phillips and Santoro (1989) describe experiences in which shy students who had previously been intimidated in FTF interactions entered CMC discussions as equal participants. Bellman (1992) notes that "The anonymity in online communication makes physical appearance, accent and other speech characteristics, ethnicity and gender irrelevant." The advantage of this decreased attention to social convention in an educational environment is that it changes previously-established structures of power, encouraging students to "think for themselves" and stand by their thoughts, a prerequisite for an egalitarian dialogue (Harrington, 1992).
This is not to assert that classroom dominance is impossible to achieve via CMC. Bray (1995) notes that modes of intimidation can still be used against CMC participants: "Electronically mediated emotional abuse falls among the same continuum of violence as do physical and sexual assault. Emotional abuse can easily be as harmful as physical or sexual violence" (p. 493). In general, this sort of comparative pain is not useful, however. To say that rude e-mail can damage someone as seriously as forcible rape trivialises the importance and severity of the latter. Nevertheless, although she exaggerates by characterising offensive words or language as "harassment and violence," it could reasonably be interpreted as a means of intimidation. This could in turn lead to classroom dominance.
A more realistic concern of distance education is that it denies students personal interaction, an aspect of the educational process female students, in particular, value. A joint survey administered at Germany's FernUniversität and the United Kingdom's Open University in 1986 by Kirkup and von Prümmer (1990) found that female students value local study centres for FTF interaction more than males. The opportunity to engage intellectually and socially with other students is an important part of learning for women distance-education students.
A later survey of distance-learning students at the FernUniversität by von Prümmer (1995) found that both males and females "lean towards personal interaction" and prefer FTF interaction rather than technology-mediated communications. Her results indicate that males and females express equal preferences in their use of FTF meetings or the telephone. Women, however, are more likely to prefer communicating with written correspondence and group discussion. E-mail was the only mode of communication for which male students had a greater preference.
May (1992) interviewed nine women who had studied at a distance through home study or teleconferencing. All were satisfied with the experience, but were bothered by the solitary nature of the courses and recommended more collaboration and interaction. May recommends an increase in collaborative learning techniques, more interaction between teachers and students as well as among students, and more interactive modes of teaching, such as teleconferencing. These results are similar to those of Gabriel and Davey (1995) who, in interviews of females enrolled in a distance-education nursing program, found that the students preferred "a way of learning that is personal, interactive and acknowledges their life experiences" (p. 501).
It is important to remember, however, that although the social cues are changed, simplified, and reduced, they are not eliminated. Using a post from USENET for example, it is still possible to determine social cues from the available data. From the sender information one may often infer the sex of the poster from which one may make assumptions of what is "appropriate" behaviour (Matheson, 1992). The status of the posting site can also become an issue. In a discussion about supercomputers, for example, email@example.com is likely to have more clout than firstname.lastname@example.org, yet perhaps less than email@example.com.
On USENET in particular, the status differences between posting domains are often argued. One posting from a given domain, be it academic or corporate, is associated with that institution. For example, participants from commercial domains often assume that posters from academic domains are young undergraduates with no understanding of the real world. Even among commercial Internet service providers a status hierarchy exists between those which offer a command-line interface and require a knowledge of UNIX and those which boast of their point-and-click ease for absolute beginners .
Even when all CMC participants post from within a single domain, one may still detect social cues. As Hampton (1994) explains in a USENET newsgroup: "We are not egalitarian, we are elitist. We care not for status or rank, nor for income or title -- only skill, intellect, wit, acumen, and ability are of any import." It is therefore surprising to find the following in a book about computer conferencing:
" it was an editorial decision not to re-print online messages in this book with all their mistakes, misspellings and lack of punctuation. The reader of a conventional text has different expectations of the written word, and uncorrected messages, acceptable in online interactions, simply revert to looking incorrect and being hard to read" (Mason, 1992, p. 14).
The editorial decision to correct mistakes is not unusual. The assumption that these errors are "acceptable in online interactions," however, is rather surprising. In a CMC setting, one's writing style is one's voice and speech; a speaker who makes grammatical errors loses status, and this is true online as well as in FTF situations. People cut each other slack, in both cases, but the errors are noticed. An error remains an error regardless of its medium of delivery. Granted, mistakes, misspellings and lack of punctuation may not immediately cause one to plummet to the bottom of the CMC social order, but they add nothing to one's status. They may, indeed, particularly in the case of electronic submission of a journal article or curriculum vitae, subtract a great deal.
These errors are exclusive of a "typically informal and 'conversational' register of conferencing" (Robinson, 1992, p. 115); the correct usage of "your" and "you're," for example, does not preclude casual written discourse. If letters typed on letterhead convey social context information (Siegel et al., 1986), what do poor spelling, incorrectly formatted software and dubious grammar convey online? Computer conferencing may permit immediate responses, but that does not free the writer from the responsibility of proof-reading and correction. One has only to see the popularity of spelling and grammar flames on USENET for verification of this. An ungrammatical sentence or an incorrectly configured newsreader provide plenty of social cues: ignorance and inability, or, at the very least, inattention to detail.
Finally, information is passed along within the language with which the text is written. For example, Lakoff (1990) describes disparities in the vocabularies of men and women. For a male to exclaim, "What a lovely steel mill!" would be considered "syntactically deviant" (p. 225). Readers would be likely to presume the author of such words was either a female or an effeminate male, and would judge the worth of the contribution with this mental image of the sender in mind. Although Poster (1990) suggests that communication via CMC can be de-gendered, it would still have to be through the use of a common or neutral language.
When interacting via text-based CMC, one is isolated from social cues and feels safe from observation and criticism. CMC decreases these reminders of a possibly critical audience, providing a sort of mask to overcome factors that inhibit participation. This often encourages people to be more open and less inhibited. It also makes it easier to dispute or confront others' opinions. (Grint, 1989; Sproull & Kiesler, 1986; Sproull & Kiesler, 1991). Harrington (1992) reports a student's remark about a computer conference: "It was easy to say what I really wanted to say because I was uninhibited and no one knew who I was" (p. 77).
In a study of a writing-evaluation class, Mabrito (1991) described the students as high- and low-apprehensive writers . He found that although the high-apprehensive writers generated only 36% of the classroom dialogue FTF when discussing each other's writing, they contributed 45% of the material when the interactions took place via CMC. Perhaps most interesting was the nature of the change of the interactions. Directive comments -- in which the participants advised others how to revise their work -- accounted for only 11% of the high-apprehensive writers' interactions FTF, but increased to 22% with CMC interactions, making the nature of their contributions more similar to those of the low-apprehensive writers (26% FTF and 25% CMC). In a similar manner, Siegel et al. (1986) report that in a decision-making task, CMC three-person groups showed twice as much equality of participation as FTF, and that participants tended to speak for their appropriate one-third of the time in CMC. In FTF, there was a much greater disparity in the amount of talk between the three participants.
CMC also creates conditions which are similar to that of deindividuation, a state of "unself-consciousness and impulsivity" (Kiesler et al., 1985) that describes people caught up in the action of gangs, crowds, or mobs (Diener, 1979; Festinger, Pepitone & Newcomb, 1952; Festinger, Schachter & Back, 1950; Forsyth, 1983; Kiesler et al., 1985; Prentice-Dunn & Rogers, 1980). Festinger et al. (1952) propose that people "obtain release" in groups and as such are "able to indulge" in behaviours in which one would not engage if one were alone:
There occurs sometimes in groups a state of affairs in which the individuals act as if they were "submerged in the group." Such a state of affairs may be described as one of de-individuation; that is, individuals are not seen or paid attention to as individuals. The members do not feel that they stand out as individuals. Others are not singling a person out for attention nor is the person singling out others (p. 382).
This merging of the inward and the outward perceptions of oneself may result in altered behaviours, such as inappropriate language or flaming.
Spears, Lea and Lee (1990) suggest that if the group is considered an aspect of the self, the inward/outward aspect tends to fall away. Communicating via CMC, people feel as though they are interacting with their computers as an extension of themselves, rather than with another person. Mason (1992) explains that even for CMC enthusiasts, this aspect may be unsettling, prompting one to question " am I talking or writing, am I reflecting or interacting, am I isolated from or connected with others?" (p. 5).
This feeling of disassociation manifests itself in two ways. In one sense, shy people or those with low social status are more likely to participate in situations in which they would normally withdraw or be passive (Johnson & Downing, 1979; Kiesler et al., 1984; Kiesler et al., 1985; Short et al., 1976; Sproull & Kiesler, 1991). On the other hand, this sense of anonymity and increased assertiveness can go too far, leading to antisocial or unrestrained behaviours (Carver & Scheier, 1981; Kiesler et al., 1985; Lea, O'Shea, Fung & Spears, 1992).