Despite the theory's merits, it is clear that, although every episode of social anxiety appears to involve self-presentational concerns, people do not feel socially anxious every time they think they will not make a particular desired impression. Self-presentational concern is a necessary but not always sufficient cause of social anxiety, raising the question of what variable allows us to account for the situations in which self-presentational difficulties do and do not cause people to feel anxious?

Our answer is that self-presentational concerns result in social anxiety primarily when those concerns have real or imagined implications for relational devaluation. Relational devaluation occurs when an individual perceives that one or more other people do not regard their relationships with the individual to be as important, close, or valuable as he or she desires. Put differently, people feel socially anxious when they believe that the impressions they make will not lead others to value their relationships with them as much as they desire, and particularly if those impressions may lead others to actually devalue, avoid, or reject them. The adolescent on a first date, the job applicant in the interview, the performer on stage, and the ill-at-ease party-goer are worried not merely about making undesired impressions, but about making impressions that will diminish the degree to which other people value their relationships with the individual. Baumeister and Leary (1995) reviewed evidence showing that people not only choose to spend most of their time with other people, but they form social attachments easily, and strongly resist the dissolution of those relationships (even many seemingly insignificant ones). The universality and strength of the human need to belong suggests that it likely evolved as a fundamental aspect of human nature because it conferred an adaptive advantage. Living in small groups of hunters and gatherers on the African savannah, early humans were likely to survive predators, starvation, injury, and illness only with the mutual support of the other individuals with whom they lived (Barash, 1977; Mithen, 1996; Tooby & Cosmides, 1992). Prehistoric individuals who tried to live away from the clan--through choice, accident, or ostracism--were unlikely to survive, much less reproduce.

Given the vital importance of maintaining social bonds throughout prehistory, a motivational-affective system evolved that helped people to avoid jeopardizing their relationships with other people. This system allowed them to monitor, in an automatic and ongoing fashion, the degree to which they were being accepted and valued vs. rejected and devalued by other people (Leary & Baumeister, in press; Leary & Downs, 1995). This sociometer typically operates with little conscious awareness on the part of the individual. Of course, people sometimes consciously ponder how they are perceived and evaluated by others, but typically the sociometer monitors the social environment, including one's own behavior, at a preattentive level for indications of immediate or potential relational devaluation.

As long as the sociometer detects no threats to relational well-being, the individual will interact with other people in a reasonably composed manner with a minimum of conscious self-reflection. Under such circumstances, people may not even be aware that they are monitoring others' reactions, although the ease with which potentially evaluative cues can evoke a response shows that the sociometer was active all along.

However, when the sociometer detects evidence of a potential problem in the individual's relational sphere, the system evokes a negative emotional response, causing the individual to feel uneasy if not downright distressed--that is, socially anxious. That is, people feel socially anxious when they believe they might make an impression that may result in relational devaluation. In order to help people to avoid devaluation and rejection, the sociometer is sensitive to cues that indicate the mere possibility that one's acceptance by other people may be jeopardized. Forewarned of the possibility, people can behave in ways that lower the likelihood of devaluation. And even when they can not avoid relational devaluation entirely, they will be motivated to take preemptive actions to buttress their relationships before the anticipated damage actually occurs (Baumeister & Jones, 1978). Viewed in this way, social anxiety may be regarded as an early warning signal for events that may lead to relational devaluation.

As an early warning system, the sociometer is inherently biased toward "false positives," sometimes detecting potential relational threats that may, in fact, turn out to be nothing. Just as a detection system for enemy missiles will occasionally provoke false alarms because it is calibrated to maximize detection of all real threats, the sociometer may cause people to feel anxious regarding imagined, potential threats that never occur. Thus, other people sometimes view an individual's social anxiety as unnecessary or overblown, even though the anxious individual experiences the threat as quite real. Furthermore, the sociometer appears to be calibrated to detect decrements in acceptance rather than true rejection per se (Leary, Haupt, Strausser, & Chokel, 1998). As a result, people may feel socially anxious even when they expect that others will value and accept them if their expectations fall below a desired threshold. As noted, relational devaluation occurs when an individual perceives that other people do not regard their relationships with him or her to be as important, close, or valuable as the individual desires. Outright rejection is not necessary.

Several theorists have suggested that emotions serve to interrupt behavior, thereby stopping organisms from continuing to behave in ways that might have disastrous consequences (e.g., Frijda, 1986; Simon, 1967). Thus, in addition to serving as a warning signal that alerts people to threats to their relational well-being, social anxiety interrupts ongoing behavior and induces a conscious assessment of the potential threat. People who feel socially anxious are acutely self-aware (Patterson & Ritts, 1997), thinking about how others are perceiving them and about their ability (or, often, inability) to cope with the situation, often to the point of being unable to devote their full attention to other things (Hartman, 1983; Hope, Gansler, & Heimberg, 1989; Sarason & Sarason, 1986). As troubling as it may be, self-preoccupation is an essential feature of social anxiety. If people are to protect the quality of their interpersonal relationships, they must consciously assess any challenges to relational devaluation that arise.

Because anxiety is inherently aversive, people try to avoid doing things that will make them anxious and they take action to reduce anxiety when it occurs. As a result, social anxiety also motivates people to take preemptive or remediative steps to protect their social bonds.

An analysis of social anxiety from the standpoint of relational devaluation makes it clear that nonphobic social anxiety is not only functional but essential to interpersonal relations. Although anxiety is inherently unpleasant, people's interpersonal interests are protected by their capacity to experience social anxiety. People who are never socially anxious do not work to control others' perceptions and evaluations of them and, as a result, tend to behave in ways that offend and alienate others (see Miller & Leary, 1992, for a related discussion of embarrassment).

Baumeister and Leary (1995) suggested that most instances in which people are ignored, shunned, excluded, or otherwise devalued center around four themes. Stated differently, people are most likely to devalue their relationships with those who make one of four kinds of impressions on them. First, people are devalued when they appear to be inept, incompetent, or unskilled. Competence is particularly important when one's value to other people depends on being able to perform certain tasks. The primitive hunter who misses the kill, the athlete who misses the shot, and the stock broker who misses the financial projection are less likely to be valued as members of their respective groups (as well as by the constituents of those groups) than the more highly skilled hunter, athlete, or financial analyst. Second, relational appreciation and devaluation are often influenced by a person's physical appearance. Physically attractive people are liked better than unattractive ones, and people devalue relationships with unattractive individuals (Feingold, 1992). Third, people's acceptance may be jeopardized when they violate important group rules or standards. Minor violations of social norms lead people to be seen as inconsiderate or unsocialized; violations of important ethical guidelines result in being seen as immoral. In either case, people who deviate from group standards are typically devalued, and extreme deviants may even be ostracized (Schachter, 1951). Finally, people may be ignored, avoided, or rejected when they are simply unappealing as social interactants. We do not value our relationships with people whom we view as abrasive, socially inept, boring, or otherwise unpleasant as much as our relationships with more desirable individuals.

Given the role that competence, attractiveness, adherence to group norms, and social desirability play in acceptance and rejection, people are understandably motivated to be perceived as competent, physically attractive, rule-abiding, and otherwise desirable. Because these domains have the greatest implications for relational appreciation and devaluation, self-presentational doubts regarding these kinds of images are most often associated with social anxiety. People worry most about projecting images of incompetence, unattractiveness, immorality, and social undesirability.

Various approaches have been promoted for the treatment of trait social anxiety and social phobia. Sociometer theory suggests that treatments for trait social anxiety and social phobia will be maximally effective if they focus on clients' concerns regarding relational devaluation. Indeed, relaxation-based approaches aside, existing psychological treatments for social anxiety may have their effects by influencing clients' perceptions of and reactions to relational devaluation. Specifically, cognitive therapies either lower clients' desire for acceptance or enhance their personal sense of social acceptability, and behavioral treatments (such as skills training) increase clients' ability to behave in ways that lead to affirming, accepting reactions from other people.

Viewing social anxiety as a response to potential relational devaluation suggests that treatments should focus on clients' concerns regarding their acceptability to other people rather than on self-acceptance. People feel socially anxious because they are worried about how other people value them, so bolstering their private self-images, although possibly effective, is at best an indirect way to lower social anxiety. On the other hand, convincing socially anxious clients that other people value them should be effective (Haemmerlie & Montgomery, 1982, 1984).

Psychotherapists should also try to determine whether a particular client's social anxiety is reasonable given their circumstances. Although many people are excessively concerned with being accepted and underestimate the degree to which they are valued, some individuals are anxious because they accurately detect a great deal of relational devaluation as they go about their social lives. Not only should unkempt, abrasive, socially inept individuals feel socially anxious (assuming that their sociometers are working properly), but even well-adjusted, socially desirable people may become highly anxious when they find themselves in unsupportive social environments. In the first instance, therapeutic efforts should be directed toward improving clients' relational acceptability as opposed to trying to convince them of their inherent worth as people or persuading them not to be concerned with other people's reactions. In the second instance, the client may be reassured by the simple knowledge that his or her anxiety is a reasonable, functional reaction to the social context, accurately reflecting the absence of regular contact with people who value their relationships with him or her.

Given that the proximal cause of concerns regarding relational devaluation involves one's impressions, treatments might focus specifically on the client's self-presentational concerns. Elsewhere, we have discussed different kinds of self-presentational concerns that may lead people to feel socially anxious and recommended treatment approaches depending on the nature of the client's difficulties (Leary, 1987; Leary & Kowalski, 1995a, 1995b).

People appear to be innately prepared to detect and respond to threats involving their acceptance by other people. Social anxiety may be conceptualized as the emotional output of an early warning system that is designed to detect potential relational devaluation well in advance of actual rejection so that the individual may take steps to protect relationships that may be in jeopardy. Because the degree to which others value their relationships with the individual depend primarily on their impressions of him or her, indications that the individual is unable to make desired impressions raise the specter of relational devaluation and evoke social anxiety.