THE PSYCHOLOGY OF TEXT-BASED RELATIONSHIPS

THE PSYCHOLOGY OF TEXT-BASED RELATIONSHIPS

Computer mediated communication (CMC) is a fi eld of research that dates back to the fi rst use of computers as communication devices, well before the current proliferation of home personal computers (PCs). Early work in this fi eld centered on the study of small work groups of people who communicated by computer, in the process of working together to design and / or produce a commercial product (Kiesler et al., 1984). In the past decade the price of a no-frills desktop PC that can connect to the Internet has fallen from over $2,000 to around $500 today. The increase in Internet use has been fast and broad, and the language used to describe it is still evolving. New words, such as “cyberspace” and “virtual communi- ties” have been added to the common lexicon, and it is common to refer to the use of the Internet as being “online.” Online is used here to describe the entire range of CMC activity, from asynchronous e-mail transmissions, browsing the World Wide Web (WWW) to synchronous chat room participation. Current research in the CMC fi eld includes studies designed to make explicit the psychological and social factors involved in text-only relationships.

Internet research fi ndings need to be interpreted with the understanding that there is no such thing as “the Internet.” The term “using the Internet” includes a wide range of possible activities, within a wide assortment of text-based and graph- ical Web-based environments. What someone reports as their experience online depends on what their intentions are, and the particular Internet text-based envi- ronment in which they participate. A chat room for “single 20s” is a far different social and psychological environment than a moderated e-mail list devoted to recovery from cancer. Real-time chat rooms that function as Internet self-help groups also exist, especially on provider networks such as America Online (AOL). Due to the constraint of having to schedule one’s time online to participate in such groups, though, they are far less populated than the asynchronous e-mail lists. Note, however, that the psychology of real-time and asynchronous text relationships are not the same. Due to the scarcity of research results from synchronous text-based self-help groups, this chapter focuses mostly on the pros and cons of asynchronous text-only Internet self-help groups.

When discussions of the pros and cons of Internet self-help arise, it is impor- tant to acknowledge that any aspect of the Internet can function as either a positive or a negative infl uence, depending on the motives of the individual user and the overall context of their participation. Some factors, such as anonymity, are both a major pro and a big con. Typing to text-only public forums from one’s private space frequently enhances the online experience, and also creates some challenges.

Text that is conversational in nature is a relatively new phenomenon. During

a lifetime of exposure to text in books and print media, people develop the fi rm impression that these written messages always represent the well-thought-out and

224 Storm A. King and Danielle Moreggi

carefully edited views of the writer. Text-only social interactions through CMC, however, are frequently the product of someone “typing off the top of his or her head.” Text-based communications often appear colder and more impersonal than the author intended, and humor and sarcasm are particularly diffi cult to convey in text-only environments. Without tone of voice or body language to give it the

correct context, typed sarcasm may come across as anger and aggression. People access virtual communities to fi nd others with similar interests (Madara et al., 1988). People who are traditionally socially ostracized due to having a stig- matized disorder or condition are becoming empowered as they fi nd and connect with each other online, in a manner not possible face-to-face (f2f). Internet virtual communities are fulfi lling people’s need for affi liation, information, and support, and allowing diverse groups a greater political voice (McKenna & Bargh, 1998).

There is a diminished sense of social presence in text-based communications (Sproull & Kiesler, 1984). The lack of tactile sensory feedback and the privacy of being in one’s own home contribute to a different sense of being connected socially. It is easier to communicate to strangers, at least partly because there is less interpersonal risk and little logistic or social cost involved (Sproull & Faraj, 1995; Wellman, 1996). The kind of social situational feedback that one normally uses to regulate one’s behavior is missing (Kiesler et al., 1984). People experience an enhanced opportunity to feel at ease with others, and, conversely, to be challenging toward others. Conditions of age difference, ethnic group membership, and gender have much less infl uence over online interactions than they do f2f.

In text-based environments, the differences between people that would otherwise inhibit the formation of a personal relationship are hidden. This promotes

a sense of group membership that is dependent solely on the limited perceptions of the individual available through CMC. When individual differences are less conspicuous, group membership becomes more prominent (Postmes et al., 2002). Virtual community participation is “an imaginative rather than a sensory experience” (Reid, 1994). Control over impression formation is enhanced in text- only mediums, because people have more command over the timing and content of their self-disclosures (Walther, 1996). People judge one another online based on perceived group similarity or difference. They engage in an over-attribution process (Lea & Spears, 1992) and assume things about others based on their own uncon- scious projections. In their mind’s eye, they fi ll in a picture of others online with whatever cues they have, never fully aware that a large part of that picture is based on their own assumptions and misattributions.

“If all computer-mediated communication systems can be said to have one single unifying effect upon human behavior, it is that usage tends to cause the user to become less inhibited.” (Reid, 1994). This tendency toward uninhibited behav- ior has been noted since the earliest studies of text-only communications (Sproull & Kiesler, 1984). This disinhibition increases self-disclosure and allows people to meet in a more profound manner than they would f2f (Donn & Sherman, 2002).

9 Internet Self-Help and Support Groups

The emotional content of the interaction is more controllable than a f2f encounter (Noonan, 1998). People tend to self-disclose more, and sooner in the relationship, than they would if they had met f2f ( Joinson, 1998).

Becoming less inhibited can have negative consequences. The improbability of any local, real life repercussions for online social behavior has produced a new psycho- logical phenomenon; people can feel free to express themselves in a manner that is much more unrestrained compared to the way they would act f2f. Communication occurs without the normal pressures that are imposed by the need to maintain social order (Huang & Alessi, 1996). Interactions in text-only environments occur without the f2f group norms normally used to regulate behavior (Finn & Banach, 2000). Identifi ed factors of text-based environments that contribute to the poten- tial for disinhibition are anonymity, lack of visual cues, asynchronous communica- tions, unwarranted projections about others, opportunities to present alternative identities, and minimization of authority (Suler, 2004).

It is easy and common for misunderstandings to occur in text-only communication. The only context one has available to interpret the communica- tion is any prior communications from that source. The only cues to guide under- standing conversational text are any “emoticons” added to the message. Emoticons are extra text added to indicate the state of mind of the writer, such as humor :-) or sadness :-(. Judgments of others made without the normal sensory clues can consist of distorted, emotionally laden projections (King, 1995). “The tendency to project stereotypical attributes on others occurs precisely because of the lack of individuating information communicated by the medium, and is promoted by the deindividuating conditions of CMC, such as the physical isolation” (Walther, 1997, p. 364). The cognitive process of stereotyping takes place more easily in text-only environments. Emotional content is the hardest to correctly place in context. One group of researchers told of a misunderstood Internet self-help group member as follows: “The original message sender complained that group members were ‘overreacting’ to a previous message’s level of despair, taking the message out of its context, or misinterpreting a previous comment as a personal attack, rather than simply an observation” (Waldron et al., 2000).

The lack of sensual clues and relative anonymity in text-based relationships creates a level playing fi eld for online social interactions. The normal situational and visual clues that tell someone about the status and position of another are absent (Kiesler et al., 1984). In contrast to f2f communications, text-based interactions do not automatically include information about social status, such as age or how expensive one’s clothing is. Indications about race, body language, and facial expres- sions that might normally inhibit inappropriate responses are not available online (Finn & Banach, 2000).

There is a mood congruent effect to text-based relationships (King, 1995). The frame of mind one is in at the time they read a particular message is a much larger part of the context in which the message is interpreted than it would be

226 Storm A. King and Danielle Moreggi

if one had f2f sensual clues available. To demonstrate this effect, suppose a group member is responded to with a message such as “I disagree with what you said.” There is range of interpretations to that statement, depending on the overall context. When tone of voice is absent, all messages seem somewhat ambiguous; people unconsciously decide what contextual information to use to guide their understanding of interpersonal communications, using cues from what is salient to them at the time. If someone has had a good day, and is feeling confi dent, calm, and composed, they will understand “disagree with you” as an opportunity to elaborate their point. If the same person had a bad day and is primed to become angry, they will interpret the same message as a personal attack on their integrity. From that state of mind, they may reply with a derogatory comment about the sender, deviating from the topic of the group. This is how a “fl ame war” gets started, that is, an online discussion that has degenerated into a series of personal attacks (Goode & Johnson, 1991). During a fl ame war, a moderately active virtual community will suddenly become very active, producing many times the normal volume of daily messages. Flame wars do not occur at the same rate in all regions of online text-based communication. Just as you would expect more arguments to occur in a bar than a church, more fl ame wars occur in open political and religious discussions groups than in Internet self-help groups. Of all the different types of text-based groups, fl ame wars (as measured by the rate of hostile commu- nications posted to the group) actually occur least often in Internet groups that have the highest frequency of emotionally supportive messages, such as Internet self-help groups (Preece & Ghozati, 2001).