It is widely held that computer-mediated communication (CMC) filters out many of the social and affective cues associated with human interaction with consequent effects on communication outcomes and the medium’s suitability for interpersonal tasks. The relationship between paralanguage and social perception in CMC in different social contexts in investigated in two experiments.
In Study 1 it was hypothesized that there would be significant differences in subjects’ perceptions of anonymous communicators as a function of the paralinguistic content of the electronic mail messages they received. Subjects read three sets of messages containing different types of paralinguistic cues and a fourth set of control messages. They also completed a set of person perception rating scales in respect of each message-sender. The hypothesis was supported for both novice electronic mail users and for experienced users drawn from a large telecommunications organization.
In Study 2 subjects participated in group discussions over a CMCS under four conditions, manipulated in a 2×2 between subjects design. The salience of the task group was either high or low, and subjects were either deindividuated (physically isolated and visually anonymous) or individuated (physically co-present and visually identifiable). From social identity theory it was hypothesized that de-individuated subjects for whom group identity had been made salient would evaluate users of paralanguage more positively than when group salience was low, in accordance with a social attraction response associated with perceptions of group identity. The hypothesis was supported.
The results suggest that paralanguage is one means by which social information is communicated in CMC and that the meaning of paralinguistic marks is dependent on the group or individual context that is pre-established for the communication. The studies therefore question earlier assumptions that the social context is dramatically reduced or eliminated in this medium. The implications of contextual effects for the use of CMC by work groups in organizations are discussed.
One observation that has been repeatedly made about computer-mediated communication (CMC) is that it filters out the interpersonal and normative aspects of communication. On this basis, wide-ranging predictions have been advanced about the computer-mediated interactions of individuals, social groups, and organizations (e.g., Kiesler, 1986; Kiesler, Siegel, & McGuire, 1984; Siegel, Dubrovsky, Kiesler, & McGuire, 1986; Sproull & Kiesler, 1991). A related view point, which has emerged from comparative media studies, is that CMC may be less appropriate for many interpersonal tasks compared to other, “richer” media (such as the telephone) or face-to-face communication (Daft & Lengel, 1984; Rice, 1987; Rice & Williams, 1984; Sproull & Kiesler, 1986; Sumner, 1988; Trev ino, Daft, & Lengel, 1990).
These predictions and conclusions about the properties of CMC have been questioned by work that has examined contextual effects (particularly social influence effects) on media choice and on the processes and outcomes of CMC (e.g., Fulk, Schmitz, & Steinfield, 1990; Hiltz, Turoff, & Johnson, 1989; Lea, 1991; Lea & Spears, 1991, 1992; Matheson, 1992; Matheson & Zanna, 1988, 1989; Myers, 1987; Spears & Lea, 1992; Spears, Lea, & Lee, 1990; Steinfield, 1986). The experiments reported in the following contribute to this debate on filtering in CMC by investigating the effects of paralinguistic cues contained in messages on subjects’ social perceptions of message senders. It will be argued that paralanguage is one source of information contained in CMC that people use to form impressions of each other when communicating. Furthermore, the interpretation of paralinguistic cues, it will be argued, is influenced by the specific social context in which the CMC is situated.
Cues filtered out
According to the “filtered-cues” hypothesis, the absence of visual and auditory cues, which are normally present in face-to-face interaction, is of fundamental importance in predicting the process and outcome of CMC. Kiesler et al. (1984) argued that the filtering of social information in CMC produces feelings of isolation and social anonymity in communicators, which results in reduced impacts of group norms and standards upon behavior, thereby encouraging the expression of deregulated and uninhibited behavior. “Flaming” and extreme decision making are considered to be two examples of antinormative behavior so-produced. According to this account, groups using CMC as a medium for discussion should show reduced susceptibility to social influence effects on their attitudes and decision making (Siegel et al., 1986). Information about the communicators themselves-their status, authority, personality, mood, motivations, and intentions-are also thought to be largely absent in CMC (Kiesler, 1986). Instead, it is contended that users immerse themselves in the content of the messages, and are able to evaluate ideas and take decisions that are un affected by social influences. In short, the filtered-cues account of CMC is that the medium frees communicators from many of the constraints that are usually exerted by interpersonal and normative factors, and thereby promotes extreme and antinormative behavior (Kiesler, 1986).
If these predictions are found to be the general case in empirical studies, they would clearly have important practical implications for the use of CMC for organizational communication. They might suggest, for example, the sorts of tasks for which CMC should be used and those for which it would be better to employ an alternative form of communication, such as the telephone or face-to-face interaction (Trevino et al., 1990). However, studies of users’ media comparisons offer little consistent support for informational accounts of CMC use and frequently implicate other important factors, such as social influence, to explain media choice (Fulk et al., 1990; Lea, 1991).
Recent studies of social psychological processes in group CMC have reported social influence effects under certain conditions that are inexplicable by the informational account. To summarize these results briefly, when CMC takes place under conditions where group norms have been previously made salient, individuals tend to behave in accordance with those norms. For example, the attitudes of group members polarize in the direction of the norm, and group decisions favor the norm. Moreover, the “deindividuated” conditions under which CMC normally takes place (i.e., between physically isolated and visually anonymous individuals) can serve to reinforce group identity and the prevailing norms (Hiltz et al., 1989; Lea & Spears, 1991; Matheson & Zanna, 1989).
Cues filtered in
Another possibility is that significant social information can be communicated in CMC even though it lacks visual and auditory channels. Several observations suggest that, under some conditions, relatively simple cues can contribute in powerful and sophisticated ways to convey social information and expressive meanings, to help regulate the interaction, and to influence attitudes and decision making. For example, power and status can be communicated through linguistic style and behavior in computer conferences (Selfe & Meyer, 1991). Different politeness strategies can function to maintain or reduce the psychological distance between communicators and can therefore be used to control the interaction (Hiemstra, 1982). Changes to the identities of users through the provision of pseudonymity or anonymity can crucially affect the conduct and outcome of group discussions in CMC (Hiltz et al., 1989; Lea & Spears, 1991; Spears et al., 1990). Indeed, experienced users can become adept at manipulating their on-line identities through these means (Finholt & Sproull, 1990; Myers, 1987). Users of a bulletin board often develop rules of social etiquette that define acceptable communication behavior for their user group. These norms and standards are sometimes made explicit to newcomers by posting the rules in a welcome message (Tsichritzis & Gibbs, 1985).
Paralinguistic cues, emoticons and emojis.
The availability of paralinguistic cues is a feature of CMC that is well known to regular users. Although the term is more commonly associated with nonverbal communication in face-to-face interaction (Edinger & Patterson, 1983), paralanguage is also present in written communication where it takes the form of typographical marks and other features of the text that, although they have no lexical meaning, nevertheless signify socially shared meanings. Spitzer (1986) documented some of the varieties of paralanguage, including the keyboard tricks that are used to produce codes that are interpreted only within the social conventions of CMC. Apart from these rather esoteric marks, there is a range of more generalized paralinguistic codes that are used to express emotion and meaning in written text. These include familiar marks such as the ellipsis and the exclamation mark. Reading these signs not only contributes to the understanding of the transmitted message, however, they also serve to define the message style from which information about the sort of person who has written the message may be inferred.
For example, one set of widely understood codes called emoticons simply relies on the reader interpreting the marks as a human face turned through 90°. Numerous variations include the smile 🙂 signifying intended humor and the frown 🙁 signifying annoyance or frustration. More recently, emojis or picture icons are popularly used to communicate a wide range of emotions and other meanings.
Experiments into social cognition have revealed some of the processes by which we tend to make significant inferences about people’s personality, emotional state, and behavioral intentions on the basis of minimal cues (Fiske & Taylor, 1984). The overattribution of enduring traits occurs particularly where the behavioral cues are consistently observed and other relevant information is in short supply (Antaki, 1989). Apparently, we even apply some of the same processes to nonanimate objects, such as computers, when they exhibit human like behavior or dialogue (Cook & Salvendy, 1989). In CMC, the presence of typing errors in a message may indicate that the sender was in a hurry when writing, but if errors are consistently observed over a series of messages they may be interpreted to mean that the person is careless or incompetent. Similarly, consistent overuse of typographical marks may lead one to conclude that the sender is a spontaneous, lively, or uninhibited person. These attributions come about because we integrate a wide range of behavioral cues into cognitive schemata about people. These cognitive frameworks define the range of behaviors that we expect from “incompetent people,” “lively people,” and so on (Cantor, Mischel, & Schwartz, 1982).
Group and Category Cues
These meager cues about personality may assume greatest importance in interactions between strangers where memories of previous interactions are unavailable. By activating the relevant social categories and schemata, the individual constructs an impression of the stranger from the available cues, but impressions so formed are unlikely to be very accurate because they rely on social stereotypes (Fiske & Taylor, 1984; Hamilton, 1981). Nevertheless, these impressions are likely to influence the conduct of the CMC significantly. Essentially, the point we wish to make is that the reduction in the number of cues available in CMC does not point to a reduction in the social context of the CMC. Instead, communicators will use whatever cues are available to construct impressions of each other; a relative lack of cues will place greater reliance on social categorization processes to interpret the available information so as to form an adequate social context. An implication of this argument is that relatively simple manipulation of the available cues will exert powerful effects on the impressions formed of interactants. Elsewhere, we have drawn the distinction between interpersonal cues, that is, cues that relate to the communicators as unique individuals, and social cues that relate to the communicator’s group and organizational identities (Spears & Lea, 1992). The establishment of an appropriate group and organizational context for the communication, we suggest, will affect the interpretation of available communication cues in accordance with this distinction, just as the group context has been shown to affect communicative behavior and decision making generally in CMC. Indeed, we consider that the de-individuated conditions in which organizational CMC takes place (visual anonymity and physical isolation) enhances social context effects in this medium (Lea & Spears, 1991, 1992; Spears & Lea, 1992).
Another possibility is that experienced electronic mail users will differ from novice users in the ways in which they come to use these cues. With experience, users may become more attentive to the presence of cues in the text, or they may learn how to make more sophisticated social judgements from them, just as we become more adept at making attributions from nonverbal communication in face-to-face interaction. Some measure of support for this prediction can be derived from evidence that new users of computer-conferencing systems go through an initial learning period in which they master the socially accepted conventions associated with CMC (Tsichritzis & Gibbs, 1985). Experienced users are also more likely to employ intentionally those paralinguistic cues that depend upon locally defined meanings for their interpretation (Spitzer, 1986).
These various issues were explored in two studies. Study 1 was a preliminary investigation into social perception in CMC and the role of paralanguage therein. The aim was to investigate whether the presence of small amounts of different types of paralinguistic cues in electronic mail messages affected the social perceptions of communicators. If the filtered-cues model is correct, and CMC is a medium that drastically reduces or eliminates interpersonal cues, there should be no significant differences in the subjects’ perceptions of message senders as a function of the type of cue present in the message. Alternatively, if paralanguage provides sufficient cues that contribute to impression formation, significant differences in the perception of message senders should be observed depending on which cues, if any, were present. A further hypothesis was that experienced electronic mail users should make greater and more sophisticated use of paralinguistic cues than novice users. This should be revealed by significant differences in either the extent or pattern of interpersonal evaluations, as a function of the type of cue present in the message.