of The American Society for Information Science

Vol. 26, No. 3

February/March 2000

Go to
Bulletin Index

bookstore2Go to the ASIS Bookstore


Social Interaction in Computer-Mediated Communication

by Jong-Young Kim

As a doctoral student at the University of Illinois, I have been studying the sociology of science and technology. One of main themes in the field is how society and technoscience co-produce each other, generating a new assemblage among heterogeneous factors and forces. In the field of sociology, cyberspace is increasingly viewed as a locus in which social life and interaction achieve new meanings and patterns. For the past 20 years, computer-mediated communication (CMC) researchers have tried to understand which media characteristics of CMC influence users' interaction patterns. This line of research, however, has generated contradictory results, producing little consensus on how social interaction is shaped in CMC.

Our seminar in social informatics has given me a platform from which to explore the question of why CMC research has produced such disparate results. It seems to me that several prominent theoretical positions on social interaction in CMC are limited by their assumption of technological determinism, a view that is reinforced by the general reliance on experimental methods. Thus, the degree to which social and technical systems are mutually constructed a primary tenet of social informatics has been lost. In this article, I examine several prominent theories related to social interaction in CMC and conclude with a brief discussion of why the "adaptive structuration" theory appears worthy of support and further research.

Cues-Filtered-Out Theory

One of the most dominant perspectives on social interaction in CMC is cues-filtered-out theory, which forms the basis of the important body of work produced by Sara Kiesler and her colleagues. Its basic argument is that text-based computer-mediated communication lacks physical and social cues, which fosters anti-normative and uninhibited behaviors. Bodiless cyberspace has an essential shortcoming in which users cannot use gesture, voice tone and facial expressions. In a similar vein, CMC lacks shared social norms and standards. It can lead users to be more aggressive and impulsive. 

One positive outcome of the inability to present socioemotive cues, however, is that CMC democratizes relationships. According to cues-filtered-out theory, CMC has a significant social impact on organizational hierarchy. A relative lack of social cues due to the anonymity and depersonalization of CMC liberates individuals from rigid and hierarchical systems and can change group dynamics. In face-to-face interaction, there is a strong correlation between social hierarchy and the amount of participation in an organizational meeting with, for example, managers speaking much more than their subordinates. The relative lack of social status cues renders electronic communication more democratic, providing a "voice for the voiceless" as people forget their social position, appearance, age, race and even gender. In addition, the absence of social barriers makes people express themselves more openly. (See Lee Sproull and Sara Kiesler's classic Connections: New Ways of Working in the Networked Organization.)

Generally, cues-filtered-out theory assumes that, because there are few physical cues in text-based CMC, there will be fewer socioemotional bonds forged online. That view helps explain why this theory concentrates on those human activities that occur in organizational settings, rather than examining a broader cross-section of CMC use environments and human activity spheres. Further, the cues-filtered-out theory is supported by observations obtained from time-limited experimental studies, rather than longitudinal research in natural settings. It may be that the CMC interactions observed in this genre of research tend to be more impersonal and task-oriented, precisely because time-limited experiments are not conducive to the formation of strong social bonds. To understand social interaction in CMC fully, analysis must also include long-term studies in natural settings. 

The Social Information Processing Perspective

Joseph Walther (1996) insists that the view of CMC as inherently impersonal must be revised and he suggests the "social information processing" perspective as an alternative to the cues-filtered-out theory. Instead of relying on the experimental method, he proposes a relatively long-term examination period in order to reveal true human interaction patterns in CMC. Social information processing theory asserts that in CMC message senders portray themselves in a socially favorable manner in order to draw the attention of message receivers and foster anticipation of future interaction. Message receivers, in turn, tend to idealize the image of the sender, overvaluing minimal, text-based cues. In addition, the asynchronous character of CMC gives the sender and the receiver enough time to edit their communication, making interactions in CMC more controllable and malleable and reducing the stress of the immediate feedback loop inherent in face-to-face (FTF) interactions. Lastly, the idealized perception and self-presentation intensify the feedback loop. This idealized relationship makes cyberspace hyperpersonal, actually exceeding FTF interactions in intensity.

Social Identity Model of Deindividuation Effects (SIDE model)

The SIDE (Social Identity Model of Deindividuation Effect) counters the prevalent view suggested by cues-filtered-out theory that CMC will liberate individuals from social constraints and norms and break social boundaries. Instead, proponents of SIDE assert that CMC reinforces existing social boundaries (Postmes, Spears, & Lea, 1998).

This model accepts the contention that anonymity in CMC deprives individuals of individual identity awareness. However, a social or a group identity replaces individual identity in CMC. Through experimentation, SIDE theorists found a salient shift from an individual identity to a group identity with individuals adhering to group norms, which become the conservative force in society. Anonymous individuals in CMC were inclined to accept "in-group" norms and identity and reject "out-group" norms and identity. Ingroup favoritism increases, while stereotyping and bias between groups prevails. The SIDE model can be applied to broader social categories such as gender, race and nationality. When anonymity maintains and gender is revealed in CMC, individuals tends to behave in terms of gender norms. The prediction is that these social categories remain firm or become fossilized in anonymous CMC. SIDE asserts that our actions stem in part from invisible social norms and identity, despite the absence of physicality, while they are nonetheless tempered by individual awareness and identity.

Adaptive Structuration

The three theories outlined above presume that characteristics of CMC determine the nature of social interaction, creating a homogeneous pattern of actions for all people, given a particular technological setting. This technological determinism does not consider the social, cultural, historical and political context of use. Further, it assumes that users of CMC are passive receivers. It is notable that each of these three theories is derived primarily from observations obtained in behavioristic experiments conducted over short time periods, which clearly truncates the amount of social embedding of technological use that can have taken place. To fully understand the nature of social interactions in CMC, we need a more naturalistic mode of observation and longer-term research.

Adaptive structuration borrows Anthony Giddens' structuration theory, which argues that social systems (including entities like social groups) are produced and reproduced by social actors' interactive use of structure, i.e., rules and resources. Structures are both the medium and the outcome of social action. Adaptive structuration theory contends that CMC is comprised of a technological dimension, a contextual dimension and structuration interaction (Poole & DeSanctis, 1990). First, the technological dimension is implemented with structural features and a "general spirit," i.e., the original purpose and orientation of CMC technology fit to local context. Unlike other media such as television, radio and film, CMC has a variety of forms for local purposes: e-mail, social MUDs, game MUDs, World Wide Web, Group Decision Support System (GDSS), Internet Relay Chat (IRC), etc. For example, e-mail in an organization might be implemented to increase the speed of exchange of official reports. While game MUDs are implemented for the purpose of entertainment, social MUDs introduce people who are interested in developing friendships or romantic relationships.

In this view, the technological dimension cannot be separated from contextual dimension. Technological structure features include the communication channel (one-to-one, one-to-many, many-to-many), the temporal structure (asynchronous or synchronous), the spatial structure, the degree of user control, the interface design (visual or textual), etc.

The variety of forms and many combinations of technological components make various and desirable social interactions in a local context possible. Although technological and contextual dimensions constrain social action, users appropriate CMC technology "faithfully" or "ironically." For example, users in text-based cyberspace create new grammar, cues, emoticon and action words to break technological constraints. Typically, people use other media, like the telephone, to supplement the shortcomings of CMC. Moreover, people use CMC in ironic ways that do not mesh with the intentions of system designers and implementers. E-mail systems in organizations, for example, soon became a tool for socioemotional support.

In 1967 Garfinkel coined the term cultural dope in his Studies in Ethnomethodology to describe the sociologist's research invention of a social actor who follows preprogrammed and stable social rules. In our everyday life, social interaction in CMC is open-ended precisely because people are not "technological dopes," but "bricoleurs" of the technological system, building with whatever comes to hand. They are imbued with indigenous reflexivity and accountability and capable of continuously transforming their actions and social world.

References for Further Reading

Cues-Filtered-Out Theory

Sproull, L., & Kiesler, S. (1991). Connections: New Ways of Working in the Networked Organization . Cambridge, MA: MIT Press.

Social Information Processing Perspective

Walther. J. (1996). Computer-Mediated Communication: Impersonal, Interpersonal, and Hyperpersonal Interaction. Communication Research, 23, 3-43

Social Identity Model of Deindividuation Effects (SIDE model)

Postmes, T., Spears, R., & Lea, M. (1998). Breaching or Building Social Boundaries? SIDE-Effects of Computer-Mediated Communication. Communication Research, 25, 689-715.

Adaptive Structuration

Giddens, A. (1984). The Constitution of Society: Outline of the Theory of Structuration. Berkeley: University of California Press. 

Poole, M.S. & DeSanctis, G. (1990). Understanding the Use of Group Decision Support Systems: The Theory of Adaptive Structuration. In J. Fulk & C. Steinfield (Eds.) Organizations and Communication Technology (pp. 173-193). Newbury Park, CA: Sage Publications.

Jong-Young Kim, a doctoral student in the Graduate School of Library and Information Science at the University of Illinois at Urbana-Champaign, can be reached by e-mail at kim24@uiuc.edu

How to Order

@ 2000, American Society for Information Science