Social Exclusion and the Deconstructed State: Time Perception,
Meaninglessness, Lethargy, Lack of Emotion, and Self-Awareness
Jean M. Twenge
San Diego State University
Kathleen R. Catanese and Roy F. Baumeister
Case Western Reserve University
The authors hypothesize that socially excluded individuals enter a defensive state of cognitive decon-
struction that avoids meaningful thought, emotion, and self-awareness, and is characterized by lethargy
and altered time flow. Social rejection led to an overestimation of time intervals, a focus on the present
rather than the future, and a failure to delay gratification (Experiment 1). Rejected participants were more
likely to agree that “Life is meaningless” (Experiment 2). Excluded participants wrote fewer words and
displayed slower reaction times (Experiments 3 and 4). They chose fewer emotion words in an implicit
emotion task (Experiment 5), replicating the lack of emotion on explicit measures (Experiments 1–3 and
6). Excluded participants also tried to escape from self-awareness by facing away from a mirror
The desire to be accepted by other people is one of the most
basic and pervasive human drives (Baumeister & Leary, 1995).
When that drive is thwarted through social exclusion or rejection,
people react in a variety of negative ways. People who have been
ostracized report decrements in physical health and increases in
stress and anxiety (K. D. Williams, 2001). People who feel ex-
cluded or rejected often become more aggressive as a result
(Kirkpatrick, Waugh, Valencia, & Webster, 2002; Leary, Kowal-
ski, Smith, & Phillips, 2003; Twenge, Baumeister, Tice, & Stucke,
2001). Self-defeating behavior often increases among socially
excluded people (Twenge, Catanese, & Baumeister, 2002), and
rejected people experience declines in self-esteem (Leary, Tambor,
Terdal, & Downs, 1995). Prisoners who have been subjected to
solitary confinement show an increase in psychotic behaviors
(McGuire & Raleigh, 1986).
Why does social exclusion cause these negative outcomes?
Early theorizing proposed that heightened states of emotional
distress would mediate between social exclusion and negative
behavior. Although intuitively plausible, the emotional distress
theory has not received much support. We have found that social
exclusion produces few differences in emotion but large differ-
ences in behavior (Baumeister, Twenge, & Nuss, in press; Twenge
et al., 2001, 2002; however, Buckley, Winkel, & Leary, 2002, and
K. D. Williams, Cheung, & Choi, 2000, did find significant effects
for mood using somewhat different manipulations). Even when the
effects on emotion do reach significance in our research, they have
failed to mediate the relationship between exclusion and negative
behaviors. Even K. D. Williams (2001), who has found some
significant effects of ostracism on anxiety and other mood reports,
observed that victims of ostracism often seem to respond in a
numb and neutral manner rather than with overt displays of emo-
tion: “It was as though they had been hit with a stun gun” (p. 159).
In this article, we hypothesize that social exclusion will lead to
feelings of inner numbness. People may respond with empty,
neutral, and even bored feelings when their need to belong is
thwarted, rather than the acute emotional distress that at first
seemed plausible. In fact, such numbness could ward off the
emotional distress that might otherwise arise by defensively iso-
lating affect and keeping negative feelings out of awareness (e.g.,
Massong, Dickson, Ritzler, & Layne, 1982). This state has been
characterized as one of cognitive deconstruction, which is marked
not only by a lack of emotion but also by an altered sense of time,
an immersion in the present rather than past or future, a relative
absence of meaningful thought, and lethargy, all of which may be
driven by the attempt to escape from aversive self-awareness
(Baumeister, 1990, 1991; see also Vallacher & Wegner, 1985,
1987). In other words, people may use the deconstructed state as a
defense against the negative experience of social rejection.
Suicide, Exclusion, and Deconstruction
To construct a theoretical approach, we consulted another liter-
ature—research on suicide—in which emotional distress was in-
tuitively plausible, but findings failed to confirm hypotheses. It
seemed logical to assume that people who kill themselves (or even
attempt to do so) must be suffering from acute unhappiness.
Contrary to that view, most findings suggest that the presuicidal
state is marked by flat affect. For example, suicidal people find it
more difficult to recall emotion-laden memories (J. M. Williams &
Jean M. Twenge, Department of Psychology, San Diego State Univer-
sity; Kathleen R. Catanese and Roy F. Baumeister, Department of Psy-
chology, Case Western Reserve University.
Roy F. Baumeister is now at the Department of Psychology, Florida
During part of the completion of this research, Jean M. Twenge was
supported by National Institute of Mental Health National Research Ser-
vice Award Postdoctoral Grant MH12329. We thank Janet Cacho, Dina
Cuervo, and Jay Rudeen for serving as experimenters and Sander Koole for
his invaluable help with the computer program in Experiment 5.
Correspondence concerning this article should be addressed to Jean M.
Twenge, Department of Psychology, San Diego State University, 5500
Journal of Personality and Social Psychology
2003, Vol. 85, No. 3, 409–423
Copyright 2003 by the American Psychological Association, Inc.
0022-3514/03/$12.00 DOI: 10.1037/0022-35126.96.36.1999
Broadbent, 1986) and perform better on tasks with affectively
neutral stimuli (Geller & Atkins, 1978).
Building on those observations, Baumeister (1990) proposed
that the presuicidal state is characterized by a defensive reaction
called cognitive deconstruction. Most suicide attempts are pre-
ceded by some failure or setback that reflects badly on the self. The
person seeks to avoid the aversive self-awareness and the acute
emotional distress that would ensue from thinking about the im-
plications of this recent failure. This is accomplished by having a
narrowly concrete focus on the immediate present rather than a
broadly meaningful thought pattern. Vallacher and Wegner (1985)
described this state as low levels of action identification. This
here-and-now focus may be successful in warding off the intense
emotion that would accompany meaningful self-awareness, but it
causes other undesirable effects. Inhibitions are undermined be-
cause most of them involve meaningful prescriptions about behav-
ior, and so the deconstructed state can cause a variety of impulsive
and disinhibited behavior (of which suicide attempts are one
important form). Without meaning, time seems to drag, and the
person remains stuck in a relatively empty present moment, cut off
from past and future. Deprived of reasons for action, the person
may become passive, lethargic, and idle (which may at least reduce
the number of actual suicide attempts).
Several parallels between suicide research and the impact of
social exclusion suggested that the deconstructed state might be
relevant to both. First, being rejected or excluded from social
groups is often a negative experience that could reflect badly on
the self, which is just the sort of experience that typically precedes
suicide attempts. Excluded people may therefore be wishing to
avoid self-awareness and the accompanying thoughts about what
might be wrong with them (to have caused others to reject them).
Second, as already noted, the absence of emotion was a surprising
but repeated finding in both literatures.
Third, there have been some signs of lethargy among rejected or
ostracized individuals. K. D. Williams’s (2001, chapter 7) allusion
to the “stun gun” effect of being ostracized was an attempt to
integrate a pattern of observations on how research participants
looked and acted after the people sitting on either side of them had
studiously ignored them while talking to each other. They
slumped, stared at their feet, showed no emotion, ignored every-
thing around them, and even sat there doing nothing when the
experiment was ended and everyone else got up to leave. In other
research, K. D. Williams, Cheung, and Choi (2000) reported an
increase in conformity among people who had been ostracized.
The authors interpreted this as a bid to win acceptance by acting
like other people, but it could also reflect passivity: The person
conforms rather than acting in an independent, self-assertive
Fourth, just as suicide is self-destructive, self-destructive behav-
ior has been found to follow from social exclusion. Laboratory
studies by Twenge et al. (2002) found increases in an assortment
of self-defeating behaviors among participants who had been so-
cially excluded: They made a higher proportion of unhealthy
choices, procrastinated, and took foolish risks. Some of these
patterns also reflect the impulsive aspect of the deconstructed state.
Last, there is a direct link between social exclusion and suicide,
which has been apparent for over a century. Durkheim (1897/
1963) showed that suicide rates are highest among people who are
not well integrated into society as a whole, and subsequent work
has continued to support this conclusion (e.g., Trout, 1980). Many
suicide attempts are directly traceable to recent experiences of
social exclusion, such as loss of job or marriage, and suicide rates
are elevated in ethnic groups or occupational categories dwindling
in size (see Baumeister, 1990, for review).
On the basis of these parallels, we hypothesized that social
exclusion might well produce the deconstructed state identified in
presuicidal individuals. If social exclusion thwarts a basic human
drive and challenges one’s self-worth, then people might prefer to
escape self-awareness and emotional distress by hiding out in a
mental state marked by numbness, lack of meaningful thought, and
a narrow focus on concrete, immediate stimuli. We hypothesize
that both social rejection (being rejected by a group of peers) and
social exclusion (hearing that one will be alone later in life) will
lead to the deconstructed state. These are somewhat different
experiences; rejection is more unambiguously personal, but may
be confined to a specific incident, whereas exclusion is less per-
sonal but longer lasting. Although these two experiences may
differ, we hypothesize that their behavioral and emotional effects
will be similar.
Predictions: Exclusion and Deconstruction
The present investigation used a series of experiments that
manipulated social exclusion and then measured various features
of the deconstructed state. It would be excessive to propose that a
simple laboratory manipulation of social exclusion compares with
a presuicidal state. However, these manipulations might make
self-awareness aversive enough for people to seek refuge in emo-
tional numbness and an absence of meaningful thought. Our pre-
dictions were therefore as follows.
Present Orientation Versus Future Orientation
One of the main components of the deconstructed state of
suicidal patients is a focus on the present instead of the future.
Suicidal people find it difficult to think about the future (Neuringer
& Harris, 1974), and they cannot make elaborate predictions about
the future (Yufit & Benzies, 1973). When given a sentence-
completion task, these patients choose fewer future-tense verbs, as
compared with control participants (Greaves, 1971). Many seem
unable to envision the future and even unable to name any poten-
tial consequences of killing themselves (Weiss, 1957). They focus
on the present and do not wish to deal with the future.
Time span is correlated with meaning. As Vallacher and Wegner
(1985, 1987) showed, meaningful thought at high levels of action
identification encompasses long time spans, extending into the past
and future, whereas less meaningful forms of thought focus on
narrow slices of time, especially the immediate present. If social
exclusion causes a shift toward the deconstructed state, then ex-
cluded participants should be more focused on the present rather
than the future.
Disordered Time Perception
According to some theories, this focus on the present represents
some people’s defensive attempts to stop time and not think about
a hopeless future (Hendin, 1982). This leads to a distorted time
perception in which the present seems to last longer than usual.
TWENGE, CATANESE, AND BAUMEISTER
When asked to estimate how much time had elapsed during 30-s
and 60-s intervals, suicidal patients overestimated the amount of
time that had passed (Neuringer & Harris, 1974). In contrast,
control participants estimated the time intervals fairly accurately.
Several other studies found similar results (e.g., Blewett, 1992;
Brockopp & Lester, 1970; Greaves, 1971; Tysk, 1984; Wyrick &
Time perception can be distorted in either direction, of course.
Under some circumstances, people may underestimate time inter-
vals. The “flow” state identified by Csikszentmihalyi (1990) is
typically described as a loss of the sense of the passage of time, so
that people in flow are often surprised to discover how late it has
grown while they were immersed in their activities. Flow is the
opposite of deconstructed numbness, even though both can be
described as some kind of immersion in the present. In flow,
awareness is absorbed in some deeply satisfying activity, and so
each moment is rich. In deconstruction, the present serves as an
escape from meaningful activity, and so it is experienced as
relatively empty, even oppressively boring. The two states have
opposite effects: Someone in flow finds that time flies, whereas
someone in a deconstructed state finds that time drags.
Hence, we predicted that social exclusion should distort the
perception of time flow in the same way that deconstruction does.
That is, excluded individuals should overestimate the duration of
experimentally controlled intervals. In that way, they would re-
semble severely bored people for whom time drags by slowly. The
opposite distortion, in which they would underestimate the dura-
tion of intervals, is more characteristic of meaningful absorption in
stimulating activity, and that seemed very unlikely among socially
The deconstructed state also includes a tendency to reject mean-
ing and higher order explanations. Suicidal people are cognitively
rigid and use a narrow perspective as a way to cope with their
situation (Baumeister, 1990). In addition, they see little meaning in
life and believe that life is not worth living. One study found a
connection between suicidal tendencies and lack of perceived
meaning in life (Edwards & Holden, 2001). Rogers (2001) has
asserted that a failure to create meaning underlies most suicide
attempts. Social exclusion may produce a similar mental state, as
a present or future without close relationships may seem mean-
ingless. K. D. Williams (2001) theorized that ostracism threatens
meaningful existence because being ignored by others simulates
the invisibility and worthlessness of death (pp. 63–64). Across
several studies, K. D. Williams and his colleagues found that
ostracized people reported that their sense of meaningful existence
had been threatened (K. D. Williams, Bernieri, Faulkner, Grahe, &
Gada-Jain, 2000; K. D. Williams et al., 2002; K. D. Williams,
Shore, & Grahe, 1998).
Meaningful thought is an important basis for self-awareness and
emotion, as these depend on interpreting one’s current situation
and comparing it with standards. Rejection may threaten meaning-
fulness because it strikes a blow against one’s anticipated future
life as surrounded with friends and family. At a simpler level,
meaningful thought may be aversive in the wake of rejection
because the person is tempted to ask why he or she was rejected,
and many possible answers would reflect badly on the self. Evad-
ing meaningful thought is therefore important for the strategy of
warding off aversive self-awareness and emotional distress.
In the present investigation, we included a brief measure of
perceived meaningfulness of life, and we predicted that social
exclusion—even a laboratory manipulation that was separate from
all the meaningful aspects of the person’s life outside the labora-
tory—would cause participants to shift toward perceiving less
meaning in their lives.
Suicidal people often display chronic passivity and lethargy,
which constitute another characteristic of the deconstructed state.
Suicide notes often express acceptance and passive submission
(Henken, 1976), and suicidal patients are generally more passive
(Gerber, Nehemkis, Farberow, & Williams, 1981; Mehrabian &
Weinstein, 1985). These patients also exhibit an external locus of
control and thus perceive personal action as unnecessary, because
they feel their fate is out of their hands (Gerber et al., 1981; Melges
& Weisz, 1971; Topol & Reznikoff, 1982). As Baumeister (1990)
observed, passivity further enables those in the deconstructed state
to escape from self-awareness.
In addition, passivity and lethargy may result from the decon-
structed state because many actions and decisions require mean-
ingful thought, which is aversive in the wake of rejection. That is,
a rejected person may minimize emotional distress by avoiding
meaning, but the basis for intelligent and planful action is under-
mined as well. (Impulsive or aimless activity, automatic responses,
and simple compliance with clear external demands would not be
prevented, however, because these do not require meaningful
choice.) Moreover, self-conscious action tends to implicate the self
as a responsible agent, so people who wish to avoid self-awareness
may shun such action. As noted above, social ostracism leads to
lethargic behavior (K. D. Williams, 2001), although up to now
those reports have been anecdotal. In the present investigation, we
measured lethargy during a writing task and a reaction-time task.
Lack of Emotion
Presuicidal individuals tend to report an absence of emotion
(e.g., Geller & Atkins, 1978; J. M. Williams & Broadbent, 1986),
which at first seems counterintuitive. After all, if one is not upset,
why try to kill oneself? Baumeister (1990) proposed that the
presuicidal state is actually accompanied by defensive efforts to
shut down one’s emotional responses to avoid the acute distress
that might accompany meaningful thought about one’s circum-
stances, which for presuicidal people are often quite negative.
As already indicated, the lack of emotion observed in our
previous studies of social exclusion (e.g., Baumeister et al., 2002;
Twenge et al., 2001, 2002) came as a surprise and prompted us to
revise our assumptions about what mediates the behavioral effects
of thwarting the need to belong. We were reluctant to conclude that
the lack of emotion meant that participants were fully indifferent to
the manipulations of social rejection and exclusion. Instead, we
began to think that they entered into the deconstructed state as a
way of warding off emotion and defending themselves against
A simpler explanation for the lack of self-reported emotion in
our studies is that participants have simply been reluctant to admit
EXCLUSION AND DECONSTRUCTION
to feeling upset. Possibly they thought they would lose face if they
acknowledged how upset they were after being rejected by others.
The present investigation included some self-report measures of
emotion, but we also included implicit measures of emotional
response. If the lack of emotion is merely an artifact of self-report
bias and impression management concerns, then participants might
show heightened emotionality on implicit measures. In contrast, if
they really have entered into a numb state as a defensive way of
shutting down their emotions, then they should show little or no
emotion on the implicit measures and might even avoid emotional
Escape From Self-Awareness
The deconstructed state includes escape from emotional distress,
but the emotional distress itself stems from self-awareness in
connection with recent failures or setbacks (Baumeister, 1990,
1991). In fact, suicide attempts may occur because the person
cannot escape from self-awareness by other means, so death be-
comes desirable as a form of oblivion (Baumeister, 1990). Con-
sistent with that view, presuicidal individuals seem to struggle with
aversive self-awareness. Suicide notes contain a greater percentage
of first-person singular pronouns (and not of first-person plural
pronouns, which would imply interpersonal connection) compared
with other documents, simulated suicide notes, and even notes
written by people who are facing involuntary death (Henken,
1976). Use of first-person pronouns is a well-established measure
of self-awareness (e.g., Wegner & Giulano, 1980). Thus, suicidal
people seem to experience a high level of self-awareness, perhaps
so high that they seek escape through death.
Therefore, our final prediction was that social exclusion would
lead to avoiding self-awareness. Being rejected or otherwise ex-
cluded would focus attention on the individual self as not part of
the group and possibly as having socially undesirable traits. Be-
cause most people desire to gain social acceptance and maintain
satisfying relationships with others, an experience of social exclu-
sion could signify that something is wrong with the self. Contem-
plating one’s possible shortcomings would presumably be an aver-
sive exercise in self-awareness. Hence, excluded people should
want to defensively avoid self-awareness.
Experiment 1 provided a direct test of the effect of social
rejection on time perception, delay of gratification, and emotion.
The deconstructed state includes a relative lack of emotion and a
perception that time is dragging. Thus we predicted that rejected
people would report relatively little emotion and would exhibit
several departures from a normal time orientation. Specifically,
they would overestimate the duration of time intervals and would
exhibit a present rather than a future time focus. The present
orientation should also work against delay of gratification, as delay
of gratification requires the person to forgo immediate rewards for
the sake of a better future.
In our procedure, groups of participants first engaged in a
structured conversation designed to help them get to know each
other. After this, all participants were asked to name the two
people with whom they would most like to work in pairs. By
random assignment, half the participants were told that no one had
expressed an interest in working with them, which constituted a
palpable and seemingly unanimous social rejection; the other half
heard that everyone chose them.
Participants then completed a battery of measures. First, they
were asked to judge the length of two time intervals. The experi-
menter told participants that the stopwatch would run for a certain
length of time, and they should give their best judgment of how
much time had passed. If rejected participants show symptoms of
the deconstructed state, they should perceive time as dragging and
will think more time has passed than actually has. (In contrast,
anything resembling the subjective “flow” experience would be
reflected in underestimation of time intervals, signifying that time
passes very rapidly.) Participants also completed a long mood
measure (to test for flattened affect) and a future time orientation
scale. Last, they were given a measure of delay of gratification, in
the form of a hypothetical choice between two jobs that differed in
short-term and long-term rewards.
The participants were 54 undergraduates (33 men, 21 women) partici-
pating as part of a course requirement for introductory psychology. They
were 76% White and 24% racial minority, and their average age was 18.8
Materials and Procedure
single-sex groups of 4–6 people. They were given nametags on which they
wrote their first names. They were given both written and oral instructions
to learn each other’s names and then talk for about 15 min using a set of
questions as a guide (the questions were taken from the Relationship
Closeness Induction Task developed by Sedikides, Campbell, Reeder, &
Elliot, 1999). After 15 min, the experimenter led the participants to
separate rooms, where participants nominated the two group members they
wanted to interact with again: “We are interested in forming groups in
which the members like and respect each other. Below, please name the
two people (out of those you met today) you would most like to work
with.” Instead of using these nominations, however, participants were
randomly assigned to be accepted or rejected by the group. Accepted
participants were told that everyone had picked them, and rejected partic-
ipants heard that no one picked them. In both cases, they were told that they
would work alone because of the unusual outcome. This procedure was
adapted from Leary et al. (1995) and was used in Twenge et al. (2001).
Judgment of time intervals.
Participants then judged the length of two
time intervals. The experimenter said, “I’m going to let the stopwatch go
for a certain amount of time, and when I stop it, I want you to tell me how
much time passed.” They were asked not to count or use their fingers to
keep track of the time. The first time interval was 40 s, and the participants
estimated the amount of time that had passed. The second interval was 80 s.
The experimenter recorded each participant’s estimates and did not give
any feedback about the estimates. Because of experimenter error, some
participants did not complete the time interval judgments, leaving a sample
of 39 participants for the first time interval and 33 participants for the
second time interval.
All participants then completed a battery of ques-
tionnaires, beginning with a one-item mood measure ranging from 1
(negative) to 7 (positive). The next page asked the participant to rate his or
her current mood on 41 adjectives using 7-point Likert scales. These
included 8 adjectives describing positive affect (e.g., happy, calm) and 33
describing negative affect (e.g., angry, nervous, fearful, ashamed).
Participants arrived at the lab in
TWENGE, CATANESE, AND BAUMEISTER
Delay of gratification.
the following scenario (modified from Kuhlen & Monge, 1968):
Participants then completed a form describing
A friend of yours of your own age has had two jobs offered to him/her.
One job has a relatively high starting salary, but little promise of
advancement or better income. The other job offers a starting salary
that is considerably lower but with the possibility of substantial
advancement and a much higher later income. Which job would you
advise him/her to accept?
The two possibilities offered were: “A. the job with the higher immediate
salary or B. the job starting with the lower salary, but with the possibility
of much higher later income.” Participants then responded to the question
“How certain are you that this is what you would advise?” on a 7-point
Likert scale ranging from 1 (not at all certain) to 7 (very certain). This was
converted into a continuous scale ranging from 1 (very certain of nondelay
response) to 14 (very certain of delay of gratification response). This
served as a measure of delay of gratification; choice A favors a short-term
gain at the expense of long-term gain, whereas choice B favors long-term
gain over short-term gain. Choice B is thus scored as higher delay of
Time orientation self-report.
Participants also responded to two self-
report measures of time orientation. The first was 10 items measuring
present versus future time orientation using a 7-point Likert scale. The
items were adapted from Kuhlen and Monge (1968) and Gjesme (1979),
including items such as “I can only think about the present,” “I find it
difficult to think about the future,” “I am most concerned about how I feel
in the present,” and “I feel a strong tendency to enjoy myself today and let
the future take care of itself.” This scale had an alpha reliability of .73 in
this sample. This measure was scored in the direction of high scores
indicating a future orientation, and low scores indicating a more present
orientation. The second time orientation measure consisted of 20 items
taken from the Zimbardo Time Perception Inventory (ZTPI; Zimbardo &
Boyd, 1999), which was originally developed as a trait measure. The items
used here were reworded to serve as a state measure of time perception. For
example, many items were changed to the present tense. These items
yielded subscales based on the original ZTPI, including future orientation,
past negative, past positive, hedonistic, and present fatalistic.
Results and Discussion
Would people rejected by their peers show distorted time per-
ception, less orientation toward the future, less ability to delay
gratification, and flattened affect? As Table 1 shows, all of these
predictions were confirmed.
Disordered Time Perception
We predicted that rejected participants would experience a spe-
cific pattern of distortions in time perception, including a sense
that time was moving slowly. When asked to judge a 40-s time
interval, rejected participants estimated that more than a minute
had passed (M ? 63.71), whereas accepted participants were fairly
accurate in their estimate (M ? 42.50), F(1, 37) ? 8.58, p ? .006.
The results were similar for the 80-s interval, with rejected par-
ticipants again significantly overestimating the amount of time that
had passed (M ? 100.06), whereas accepted participants were
more accurate (M ? 71.13), F(1, 31) ? 7.40, p ? .01. These
overestimations are consistent with the deconstructed state, in
which time drags by slowly.
Future Versus Present Orientation
Rejected participants were also more present oriented and less
future oriented on a self-report scale, F(1, 50) ? 4.37, p ? .05.
They were more likely to agree that they only wanted to think
about the present and found it difficult to think about the future.
Present orientation is another important feature of the decon-
structed state, particularly because present activities often draw
meaning from the future (such as goals or anticipated fulfillments).
To defensively avoid meaning, one severs the present from the
Social exclusion did not produce significant differences on any
of the scales of the other time perception inventory, the ZTPI. This
may have occurred because the ZTPI was written originally as a
trait measure and not as a state measure. Alternatively, the decon-
structed state may not affect the categories of time perception
measured by the ZTPI.
Delay of Gratification
The social exclusion manipulation also elicited different re-
sponses on the measure of delay of gratification. Accepted partic-
ipants overwhelmingly (94%) reported that they would advise a
friend to take the job that had a lower starting salary but more
possibilities for advancement (vs. a job with a higher starting
salary and little promise of advancement). Such an approach
The Effect of Social Exclusion on Time Perception, Future Orientation, Delay of Gratification, and Mood, Experiment 1
F dfM SDM SD
40-s time interval
80-s time interval
Future orientation scale
Delay of gratification (% choosing delay)
Delay of gratification confidence rating (1–14)
One-item mood rating (1–7)
?2(1, N ? 51) ? 4.74*
* p ? .05.
ZTPI ? Zimbardo Time Perception Inventory.
** p ? .01.
EXCLUSION AND DECONSTRUCTION
represents a willingness to delay gratification and an adaptive
ability to integrate the present and the future. In contrast, rejected
participants were significantly less likely to make this choice
(73%), ?2(1, N ? 51) ? 4.74, p ? .03. Compared with accepted
participants, significantly more rejected participants chose to take
the immediate rewards despite the long-term costs. True, the
majority of rejected participants recognized that delaying gratifi-
cation was the best choice, but a significant minority favored
short-term rewards over long-term benefits.
The certainty of their choice, calculated on a 14-point scale, was
also significantly different between groups, F(1, 49) ? 7.23, p ?
.01. Even when rejected participants chose the delay option, there
was a trend toward less confidence in their choice (M for re-
jected ? 11.92, M for accepted ? 13.00, only among participants
who chose the delayed gratification option), F(1, 39) ? 3.08,
p ? .09.
Lack of Emotion
Accepted and rejected participants differed significantly on the
one-item mood measure but showed no significant differences on
the detailed mood measure. On the one-item, 7-point holistic
measure, rejected participants averaged 4.40 and accepted partic-
ipants averaged 5.05, F(1, 52) ? 5.53, p ? .03. Thus the difference
was significant but small. Meanwhile, the more detailed mood
measure uncovered no mood differences at all, in either positive or
negative affect (see Table 1). This was true even when the scales
were broken down into more specific scales; there were no signif-
icant differences in sadness, fear, embarrassment, or anger. This is
consistent with previous research in our labs, which showed no
significant differences in the Positive and Negative Affect Scale
(PANAS; Baumeister et al., 2002; Twenge et al., 2001) or the
Brief Mood Introspection Scale (BMIS; Twenge et al., 2002).
Leary and colleagues (e.g., Buckley et al., in press) have found
mood effects after social rejection, but in their methods the par-
ticipants do not meet their rejecters in person; this may make the
experience less traumatic and thus not produce the defensive
response of numbness. Thus, rejection produced a small difference
on a holistic measure of mood and no differences on a compre-
hensive, detailed mood list. Being rejected unanimously by a
group of peers apparently did not elicit very much emotional
distress, which is surprising and counterintuitive unless one pro-
poses that they entered into a state of numbness that may be
attractive precisely because it prevents distress.
Given the lack of a main effect for mood (on the detailed
measure), it is not surprising that mood did not mediate the effect
of rejection on any of the dependent variables. Exclusion (accep-
tance vs. rejection condition) was still significantly correlated with
all of the dependent variables when these analyses were controlled
for the detailed scales of positive and negative mood—r(35) ? .45,
p ? .01 for the 40-s time interval; r(29) ? .47, p ? .01 for the 80-s
time interval; r(47) ? .28, p ? .05 for future orientation; and
r(46) ? .36, p ? .02 for the delay of gratification certainty ratings.
Experiment 2 had three goals. First, we sought to measure the
effects of social rejection on feelings of meaninglessness. Al-
though K. D. Williams (K. D. Williams, Bernieri, et al., 2000;
K. D. Williams et al., 1998, 2002) found that ostracism threatened
meaningful existence, two of these studies were not experimental,
and all manipulated ostracism (a somewhat different concept and
experience). We sought to measure feelings of meaning in life after
random assignment to an experience of social acceptance or re-
jection. After being accepted or rejected, participants indicated
their agreement or disagreement with the statement “Life is mean-
ingless.” Cognitive deconstruction is essentially an avoidance or
rejection of meaningful thought, so we predicted that social ex-
clusion would lead to greater perceptions that life is meaningless.
Second, we wanted to replicate the lack of emotion results of
Experiment 1 using a different measure of mood, as a way of
reducing the possibility that the null results were an artifact of an
insensitive measure. Experiment 2 used a more general measure of
mood (modified from K. D. Williams, Cheung, & Choi, 2000),
with items such as “bad–good.”
Third, Experiment 2 included a manipulation check, asking
participants if they felt rejected or accepted. This enabled us to
confirm the effectiveness of the group rejection manipulation.
Neither Experiment 1 nor our previous studies using that method
have confirmed that the group rejection condition actually makes
people feel that they have been rejected.
The participants were 96 undergraduates (52 men, 44 women) partici-
pating to fulfill a course requirement in introductory psychology. They
were 67% White and 33% racial minority, and their average age was 19.1
Measures and Procedure
Participants experienced the same social rejection manipulation as in
Experiment 1, hearing either that everyone or no one chose them after a
group interaction. After receiving the false feedback, they completed
several questionnaires. The first was a mood measure and manipulation
check. The three mood items were drawn from K. D. Williams, Cheung,
and Choi (2000) and asked participants to rate their mood on a Likert scale
from 1 to 9 with the anchors of bad-good, sad-happy, and tense-relaxed.
They also completed a manipulation check item, “rejected–accepted.”
Participants then rated their agreement with two questions, again on a
9-point Likert scale: “How true is the statement: ‘Life is meaningless’?”
and “How true is the statement ‘I am in control of my life’?” The anchors
for the Likert scale were 1 ? not at all true and 9 ? very much true.
Results and Discussion
As Table 2 shows, the manipulation caused participants to feel
more rejected: Those rejected by the group reported significantly
greater feelings of rejection than those in the accepted condition,
F(1, 94) ? 14.67, p ? .001. This finding confirms that the
rejection manipulation was effective. This should also enhance
confidence in the results of Experiment 1, as that study used the
Meaning in Life
As Table 2 shows, rejected participants were significantly more
likely than accepted participants to agree with the statement that
TWENGE, CATANESE, AND BAUMEISTER
“Life is meaningless,” F(1, 94) ? 10.56, p ? .002. As the mean
(2.44 out of 9.00) illustrates, however, it is more accurate to say
that rejected participants did not disagree with the statement as
strongly as accepted participants did. Seventy-nine percent of
rejected participants indicated disagreement with the statement by
circling a number below the midpoint of 5, but 21% agreed with
the statement (circling a 5 or higher). None (0%) of the accepted
participants agreed with the statement; they all circled numbers
between 1 and 4. Thus, socially rejected people found less mean-
ing in life, as predicted by deconstruction theory and K. D.
Williams’s (2001) theory of ostracism and threats to meaningful
existence. Such a response may be part of a larger shift toward a
deconstructed state that avoids meaning to prevent oneself from
recognizing the possible implications of the social rejection.
Feelings of Control
In contrast to the effect of rejection on meaningfulness, rejected
participants did not report feeling a lack of control (see Table 2).
This suggests that social rejection does not affect all attitudes
about life, but (in this case) meaninglessness in particular.
Lack of Emotion
The results of this experiment also confirmed the lack of emo-
tion displayed by the participants in Experiment 1. The accepted
and rejected groups did not differ on the composite, three-item
measure of mood (see Table 2). There were also no significant
differences on the individual items. Thus, participants did not
show increased negative mood or decreased positive mood as the
result of social rejection. This is consistent with deconstruction
theory, which predicts that socially rejected people will display
flattened affect. The results should be interpreted with caution,
however, as one cannot confirm a hypothesis with a null effect.
We also found that mood did not mediate the effects. The
bivariate correlation between condition (acceptance vs. rejection)
and feelings of meaninglessness was r(94) ? .32, p ? .002. When
controlled for mood, it was essentially unchanged, r(93) ? .31,
p ? .002. Mood was not correlated with feelings of meaningless-
ness in a bivariate analysis, r(94) ? .15, ns, or when controlled for
condition, r(93) ? .13, ns.
In Experiment 3, we measured the effect of social exclusion on
lethargic passivity. One hallmark of the deconstructed state is that
meaningful action is undermined by the avoidance of meaningful
thought and also by a reluctance to accept responsibility onto the
Lethargy and passivity are notoriously difficult to measure,
especially among laboratory participants who often arrive full of
energy and eager to explore the experience or at least get it over
with. We measured lethargy by asking participants to write defi-
nitions for 10 common proverbs in 10 min. Such a task clearly
requires meaningful thought, as the person must move from the
stimulus sentence to the broader meaning and then reformulate that
meaning in his or her own words. The lethargic response would be
to write less and not to exert the energy necessary to give long,
detailed definitions of the proverbs. The active response would be
to generate a large number of words in the process of defining the
proverbs. If excluded participants are feeling lethargic and numb
as part of the deconstructed state, we would expect them to
generate fewer words.
To strengthen generalizability, we used a different manipulation
of social exclusion. Participants were given a personality inventory
and heard false feedback about the results. In the crucial “future
alone” condition, they were told that they were likely to end up
alone later in life. The first control group heard that they would
have good relationships throughout life (future belonging). In a
second control condition, participants heard that they would be
accident prone later in life, which is a negative outcome unrelated
to social exclusion (future misfortune).
We measured both self-esteem and mood as possible mediators.
We used yet another measure of mood, the PANAS (Watson,
Clark, & Tellegen, 1988) to test for mood differences among the
conditions. In addition, we administered the State Self-Esteem
Scale (SSES; Heatherton & Polivy, 1991) to test for mediation by
state self-esteem. Perhaps social exclusion leads to decreased
self-esteem, which then leads to the depressive deconstructed state.
The participants were 43 undergraduates (23 men, 20 women) partici-
pating as part of a course requirement for introductory psychology. They
were 74% White and 26% racial minority, and their average age was 18.4
Materials and Procedure
Participants were first asked to fill out a personality questionnaire (the
Eysenck Personality Questionnaire [EPQ]; Eysenck & Eysenck, 1975).
The Effect of Social Exclusion on Feelings of Meaninglessness, Sense of Control, Mood, and
Acceptance Versus Rejection, Experiment 2
M SDM SD
Agree that “life is meaningless”
Agree that “I am in control of my life”
Three-item mood measure
Manipulation check: Rejected (low) versus accepted (high)
** p ? .01. *** p ? .001.
EXCLUSION AND DECONSTRUCTION
Participants were randomly assigned to one of three social exclusion
conditions: future alone, future belonging, or future misfortune. To gain
credibility, the experimenter first gave an accurate assessment of the
participant’s extraversion score, providing correct feedback about whether
the score was high, medium, or low on this scale. The experimenter used
this as a segue into reading a randomly assigned “personality type” de-
scription. In the future alone condition, the participant was told the fol-
You’re the type who will end up alone later in life. You may have
friends and relationships now, but by your mid-20s most of these will
have drifted away. You may even marry or have several marriages,
but these are likely to be short-lived and not continue into your 30s.
Relationships don’t last, and when you’re past the age where people
are constantly forming new relationships, the odds are you’ll end up
being alone more and more.
In contrast, people in the future belonging condition were told the
You’re the type who has rewarding relationships throughout life.
You’re likely to have a long and stable marriage and have friendships
that will last into your later years. The odds are that you’ll always
have friends and people who care about you.
Last, a future misfortune condition was included, in which people were
told the following:
You’re likely to be accident prone later in life—you might break an
arm or a leg a few times, or maybe be injured in car accidents. Even
if you haven’t been accident prone before, these things will show up
later in life, and the odds are you will have a lot of accidents.
This condition was intended to describe a negative outcome that was not
connected with relationships or social exclusion. This method was used
previously in Baumeister et al. (2002) and Twenge et al. (2001, 2002).
After hearing the prediction about their future lives, participants com-
pleted the PANAS (Watson et al., 1988) and the SSES (Heatherton &
Polivy, 1991). They were then presented with a list of 10 common proverbs
and asked to “explain in your own words what you think it means.” The
proverbs included “People who live in glass houses shouldn’t throw
stones,” “Don’t cry over spilled milk,” “Don’t throw the baby out with the
bathwater,” and “A bird in hand is worth two in the bush.” Participants
were given 10 min to work on this task. We then counted the total number
of words generated by each participant.
Results and Discussion
Would social exclusion cause people to be lethargic and not
generate as many words in response to a definition task? We found
that social exclusion did indeed lead to lethargy (see Table 3).
Participants in the future alone condition generated significantly
fewer words compared with the other two conditions, F(2,
40) ? 4.53, p ? .02. This effect was not due to simply hearing bad
news. Participants in the future misfortune condition, who heard
that their lives would involve frequent accidents, generated just as
many words as participants in the future belonging condition.
Planned contrasts showed that the future alone group wrote sig-
nificantly fewer words than the other two groups (using Tukey’s
honestly significant difference [HSD], p ? .05), but the future
belonging and future misfortune groups were not significantly
different from each other. Thus, at least on a task that required
considerable meaningful thought, social exclusion resulted in a
significant reduction in total output.
We attempted to code the interpretations for accuracy using two
trained raters, but no significant differences were found. This
might be dismissed as an uninterpretable null finding, possibly
reflecting measurement difficulties. Alternatively, it may mean
that the deconstructed state did not make people any less accurate
in their ability to think—merely less inclined to do so.
There were no significant differences among the conditions in
mood or state self-esteem (see Table 3). The null effect for mood
replicated the results of Experiments 1 and 2, which also found no
differences in mood between the accepted and rejected groups
(except for the small difference on the one-item measure). Even
with a different manipulation of social exclusion and a different
mood measure, we found that social exclusion does not produce
significant differences in mood.
The link between exclusion and lethargy was also not mediated
by mood or by state self-esteem. To perform correlational analy-
ses, we grouped the future belonging and future misfortune groups
together to compare with the future alone group. The correlation
between exclusion and words generated was still significant when
controlled for negative mood, r(40) ? .44, p ? .005, and when
controlled for positive mood, r(40) ? .40, p ? .01. In addition,
neither mood scale was significantly correlated with number of
words generated. The effect was also not mediated by state self-
esteem; the correlation between exclusion and words was signifi-
cant with state self-esteem controlled, r(38) ? .44, p ? .005, and
the correlation between state self-esteem and words generated was
Our control group in this study heard that they would be acci-
dent prone later in life, and this group did not display the lethargy
seen in the future alone participants. Thus, not all types of negative
feedback produced a passive response. One might ask if other
types of negative feedback about one’s future (e.g., failing in a
career or having financial trouble) would also show null effects:
Are these misfortunes just as “bad” as social exclusion, where
The Effect of Social Exclusion on Word Generation, Mood, and State Self-Esteem, Experiment 3
Future belonging MisfortuneFuture alone
M SDM SDM SD
* p ? .05.
TWENGE, CATANESE, AND BAUMEISTER
perhaps being accident prone is not? This question is almost
impossible to answer: One would have to compare setbacks from
different domains that have been equated on an objective scale of
how “bad” they are. It is unlikely that such an objective scale
In Experiment 4, we examined lethargy in another way: by
measuring participants’ reaction times. The lethargic response
would be a slower, sluggish reaction time. We returned to the
group rejection manipulation used in Experiments 1 and 2, with
participants hearing either that no one or everyone in a group had
chosen them as a partner for further interaction. Participants then
played a reaction-time game, ostensibly with another person but
actually against the computer. They were told to click on a square
when it turned red and a beep sounded. The program continued
for 25 trials, but we did not necessarily expect the lethargic
impairment to continue throughout the exercise. Novel tasks de-
pend on controlled processes, including the self’s executive func-
tion, whereas after a few trials the response can become familiar
and well learned so that responses can be fairly automatic. Cog-
nitive deconstruction is essentially a state of the self, and so it
should mainly affect controlled processes while leaving the auto-
matic ones untouched. The specific prediction was therefore that
rejection would lead to slower responding on the early trials but
not on the later trials.
The participants were 100 undergraduates (49 men, 51 women) partic-
ipating as part of a course requirement for introductory psychology. They
were 74% White and 26% racial minority, and their average age was 18.6
years. Data on the aggressive behavior of these respondents were reported
in Twenge et al. (2001) and Twenge and Campbell (2003).
Materials and Procedure
As in Experiments 1 and 2, participants arrived at the lab in single-sex
groups of 4–6 people. They talked for 15 min and were then led to separate
rooms, where they nominated the two people with whom they desired
further interaction. Each participant was told either that everyone had
picked him or her for the group task (accepted condition) or that no one had
picked him or her (rejected condition).
Participants were then told that they would play the reaction time game
with a nongroup member (someone who arrived too late to participate in
the group discussion) because of the unusual vote. In actuality, the Macin-
tosh computer was programmed to mimic a person’s responses (a detailed
description of this program is presented in Bushman & Baumeister, 1998).
Participants were told that they would have to press a button as fast as
possible on each trial. A green square on the computer screen would first
turn yellow, and then red; participants were told to click on the square as
soon as it turned red. The program also produced an audio beep when the
square turned red. The computer program recorded the reaction time of the
participant on each trial.
Results and Discussion
Rejected participants (M ? 495.61 s) showed slower reaction
time on the first trial, compared with accepted participants (M ?
326.92 s), F(1, 98) ? 4.35, p ? .04. Thus, when the task was novel
and unfamiliar, rejection led to slower reaction times. This fits the
view that rejection causes lethargy that can interfere with con-
The second trial showed a trend in the same direction as the first
turn, with rejected participants (M ? 279.84) reacting slower than
accepted participants (M ? 244.92), F(1, 98) ? 2.27, p ? .13.
There were no significant differences on trials 3–25. The second
trial reaction times were also faster overall (comparing the first and
second turns), t(200) ? 4.68, p ? .001, a trend that continued in
the later trials. The faster reaction time on these trials supports the
argument that automatic processes have begun to take over this
function, leading to fewer differences between the experimental
conditions. Thus, once the task was routine, rejected participants
reacted just as fast as accepted participants. The implication is that
rejection causes a lethargy in executive function that slows down
controlled responses to an unfamiliar task but does not affect
automatic responses to a familiar task.
Experiments 1–3 found that social exclusion did not produce
differences in self-reported mood across three different measures
listing emotion words (although there was a small effect on a
one-item mood rating in Experiment 1). All of these studies used
explicit mood measures, in which participants were explicitly
asked to rate their current mood on a variety of adjectives. It is
possible that socially excluded people, in particular, are reluctant
to admit to being in a negative mood. They may wish to preserve
the illusion (to themselves and possibly to the experimenter) that
they are unaffected by the social exclusion.
Experiment 5 was therefore designed to measure mood more
implicitly. Participants first received the future prediction exclu-
sion manipulation used in Experiment 3. In the crucial future alone
condition, participants were told that they were likely to end up
alone later in life. We simplified the design by including only one
control group, who received no feedback at all. After the manip-
ulation, participants sat in front of a computer programmed to
display a line of Xs and then flash briefly. Participants were told
that the computer would flash a word during this brief time (it
actually flashed a blank screen) and that they should circle the
word they thought they saw from a series of choices. There
were 18 trials with four choices; each question had one emotion
word and three neutral words as choices—for example, (a)
TREAD; (b) BREAD; (c) SCARED; (d) FARED. The number of
emotion words chosen served as the measure of emotional
Competing predictions could be made. If the lack of emotional
response to rejection in previous work was an artifact of self-report
methods, it remains plausible that rejection would cause substan-
tial emotional distress. In this case, rejected participants would
choose more negative emotion words and fewer positive emotion
words. Alternatively, if rejection is emotionally neutral and unim-
pactful, then there would be no difference between conditions.
Meanwhile, the cognitive deconstruction theory predicts that re-
jected participants would choose fewer emotion words (of either
valence) than other participants, consistent with the view that
deconstruction is essentially a move to shut down one’s emotions.
EXCLUSION AND DECONSTRUCTION
Participants were 30 (10 men, 20 women) undergraduates who com-
pleted the experiment in return for course credit. They were 43% White and
57% racial minority, and their average age was 20.0 years.
Participants completed the EPQ and were given accurate feedback about
their extraversion score. They were then randomly assigned to either (a)
hear the future alone prediction that they would be alone later in life
(worded exactly the same as in Experiment 3), or (b) move directly to the
next part of the experiment without hearing a future prediction. Participants
then sat down at the computer and the experimenter explained the task. A
Macintosh computer was used to present the stimulus material. We mod-
ified this technique from Hass, Katz, Rizzo, Bailey, and Moore (1992) and
Koole, Smeets, von Knippenberg, and Dijksterhuis (1999). Participants
were told that the computer would display a row of Xs and would then flash
a word for a split second before returning to the screen with the row of Xs
(in actuality, the computer showed a blank screen). On each of 18 trials,
participants were asked to circle the word they thought they saw from
among four choices with one emotion word and three other choices—for
example, (a) RECANT; (b) PLEASANT; (c) REPLANT; (d) PRESENT.
There were nine negative emotion words (down, scared, low, uneasy,
upset, sad, nervous, anger, depressed) and nine positive emotion words
(happy, content, pleasant, joyful, calm, good, pleased, glad, secure) drawn
from Hass et al. (1992).
Results and Discussion
Future alone participants circled fewer emotion words overall
(negative plus positive; M ? 4.80) compared with no-feedback
participants (M ? 6.93), F(1, 28) ? 5.00, p ? .04. The results for
negative words (M for future alone ? 1.67; M for no feed-
back ? 2.47) and positive words (Ms ? 3.13 and 4.47, respec-
tively) were in the same direction, with future alone participants
circling fewer words, although those differences failed to reach
These findings help rule out the view that the lack of emotional
distress after rejection is simply an artifact of self-report measures.
In this experiment, the task was implicit and more subtle. Endors-
ing emotion words in this procedure was presumably just a sign of
accessibility. There was no sign that negative emotion (or positive
emotion for that matter) was more accessible among people who
anticipated a socially isolated future than among people who had
received no feedback about their futures. On the contrary, the
significant difference obtained in this study suggests that the
prospect of social exclusion made all emotional responses less
likely. These findings seem most consistent with the deconstruc-
tion hypothesis: Social exclusion is threatening, and people re-
spond to it by attempting to shut down their emotions.
Experiment 6 investigated another crucial feature of the decon-
structed state: the attempt to escape from self-awareness. People
shift to low levels of meaningful thought partially because they are
reluctant to focus thoughtful attention on themselves; such atten-
tion would lead to distressing and painful reflections on one’s
flaws or inadequacies. Social exclusion raises the threatening
possibility that something is wrong with the self (which causes
others to reject you), and so people might defensively prefer not to
dwell on those thoughts.
The procedure for measuring whether people avoid or seek
self-awareness was adapted from Greenberg and Musham (1981).
The room contained two chairs, one placed facing a mirror and the
other with its back to the mirror and facing a blank wall. Partici-
pants were asked to sit down and wait for another experimenter.
We then surreptitiously recorded which chair they chose: the one
facing the mirror, with its increase in self-awareness, or the one
facing away from the mirror, with its avoidance of self-awareness.
Facing a mirror is a common technique for increasing self-
awareness (Diener & Wallbom, 1976), so the choice of the chair is
a clear reflection of avoiding or embracing self-awareness. We
predicted that socially excluded participants would be more likely
to avoid self-awareness by choosing the chair facing away from
Forty undergraduate students (26 men, 14 women) participated to fulfill
introductory psychology course requirements. Data from 2 participants
were discarded because 1 expressed suspicion and 1 did not follow direc-
tions (thus, there were originally 42 participants). The sample was 72.5%
White and 27.5% racial minority, and the average age was 19.1 years.
Participants were told that the study involved completing personality
questionnaires and learning about one’s personality, and they completed a
demographic survey and the EPQ. Participants were randomly assigned to
one of four social exclusion conditions: future alone, future belonging,
future misfortune, or no feedback (similar to the method used in Experi-
ment 3). They then completed the BMIS (Mayer & Gaschke, 1988) and the
SSES (Heatherton & Polivy, 1991).
At this point, participants were told that this experiment was over.
Because this experiment was so short (it usually took people only 15–20
min to complete), participants were told that it was being combined with
another “unrelated” experiment that would only take a few minutes. The
second experiment, they were told, would be led by another experimenter
who was having trouble getting people to sign up for his experiment
because it was so short. This experiment would take place in a different lab
across the hall. The experimenter did not present the participants with an
explicit opportunity to decline participation in the second experiment. All
participants had signed up for a 30-min session, and none refused to
complete the second experiment.
The participant was then led to a separate laboratory in the same building
to increase the believability of the cover story. The experimenter directed
participants to a room that appeared to be in use: The door was ajar and the
lights were on, but the experimenter was not there. The room contained two
chairs placed back to back with a sign taped between them reading “Do Not
Remove.” Each chair was equidistant from the wall in front of it, and
approximately equidistant from the door through which the participants
entered the room (see Figure 1). One of the walls had a 36-in. ? 48-in.
full-length mirror, and the other wall was blank. The positioning of the
mirror was counterbalanced: half the time it was on the right side of the
room and the other half the time it was on the left side.
When standing in the doorway, the participants could view the entire
room. The experimenter acted surprised that the second experimenter was
not waiting in the room, and commented to the participant that he probably
stepped out to use the bathroom and should be back any minute. The
TWENGE, CATANESE, AND BAUMEISTER
experimenter then said to the participant, “Why don’t you have a seat, I’m
sure [the experimenter] will be back in a minute.”1The participants’ choice
to sit either facing toward or away from the mirror was the main dependent
variable of seeking or avoiding self-awareness.
The experimenter then left the room, ostensibly to search for the missing
experimenter. She returned to the room saying that she didn’t see him
anywhere, but while the participant waited, he or she could fill out a
questionnaire so they wouldn’t get behind. The questionnaire assessed the
strength of the participants’ seating choice (1 ? not at all strong to 9 ?
extremely strong). Participants also reported their extraversion score, the
prediction they had received for the future (if any), and how much they
thought this prediction might describe their future (1 ? not at all to 7 ?
very much). These questions served as the manipulation check for the
social exclusion manipulation. Finally, participants recorded what was on
the wall in front of them (to confirm the experimenter’s record) and wrote
an open-ended explanation of their choice to sit either facing or away from
the mirror. Participants were then carefully and thoroughly debriefed,
thanked, and dismissed.
Participants’ memory for their predicted future and their belief
that this prediction might describe their future served as a manip-
ulation check for social exclusion. All participants correctly re-
called their prediction for the future. Participants were also asked
to rate their belief that this prediction would describe their future
on a Likert scale (1 ? not at all, 7 ? very much). A one-way
analysis of variance on this rating (comparing the three conditions
who heard future predictions) showed a significant difference
between conditions, F(2, 28) ? 10.99, p ? .0001. Tukey HSD post
hocs revealed that people in the future belonging condition
(M ? 5.50, SD ? 1.27) were more likely to think that this
prediction would describe their future than those in the future
alone (p ? .0001) and future misfortune (p ? .001) conditions.
People in the future alone (M ? 2.50, SD ? 1.08) and future
misfortune (M ? 3.10, SD ? 2.02) conditions did not differ
significantly in their belief that this prediction would describe their
Would socially excluded people avoid self-awareness by choos-
ing to sit facing a wall rather than risk facing their own reflection
in a mirror? As Table 4 shows, those in the future alone condition
were significantly more likely to avoid self-awareness by seating
themselves facing away from the mirror. All but 1 future alone
participant chose to face away from the mirror (90% chose to
avoid self-awareness by facing away). In contrast, participants in
the other three conditions were evenly divided in their seating
choices. The difference between the future alone condition and the
three control conditions combined was significant, ?2(1, N ?
40) ? 5.76, p ? .02. Thus, the future alone participants consis-
tently chose to avoid self-awareness by looking at a featureless
blank wall rather than their own reflection in a mirror. Participants
in the control conditions apparently chose their seats randomly.
The overall chi-square among the four conditions (future alone,
future belonging, misfortune control, and no-feedback control)
was also significant, ?2(3, N ? 40) ? 9.31, p ? .03. In addition,
the future alone group was significantly different from the misfor-
tune control group, ?2(1, N ? 20) ? 5.50, p ? .05, and from the
no-feedback control group, ?2(1, N ? 20) ? 7.50, p ? .02
(planned contrasts with Bonferroni corrections applied). Thus, the
negative feedback had to be specific to relationships to cause
avoidance of self-awareness; hearing that one would be physically
hurt (future misfortune) did not increase the desire to face away
from a mirror (in fact, 60% of these participants chose to face the
mirror instead of the wall). This group was not much different
from those who heard no prediction at all, 70% of whom chose to
face the mirror (suggesting that facing a mirror rather than a wall
is people’s usual preference). Overall, the three control groups
1The experimenter was not blind to condition or to the hypotheses of
this study. To reduce experimenter bias, the experimenter adhered to a
strict script by precisely plotting her motions and verbal responses to create
consistency across all conditions. By being consistent across all partici-
pants, the experimenter hoped to reduce any unconscious nonverbal or
verbal behaviors that might confirm expectancies. The experimenter al-
ways stood in the same place before allowing the participant to enter the
room with the mirror. Instead of leading the participant into the room, the
experimenter always stood just beyond the threshold of the doorway while
she commented that the second experimenter was absent. This motion
blocked entrance to the room and left all participants standing in the
doorway. This ensured consistency across conditions. It also prevented
overeager participants from entering the room before they were instructed
to do so and allowed them to view the set up of the room from behind the
experimenter. Most importantly, this motion ensured that all participants
entered the room from the same starting point and were never led into the
room by the experimenter (which could potentially reveal experimenter
expectancies). The experimenter always stepped aside to her right when
finally allowing the participant to enter the room. Because the mirror was
counterbalanced (it was sometimes on the right side of the room and
sometimes on the left), the experimenter’s stepping aside to the right
should not create any particular response set in the participants. The
experimenter always entered the room behind the participants once the
participants had fully entered.
The experimenter also adhered to a strict verbal script. She never
engaged in unnecessary conversation with participants while leading them
to the second laboratory. She always gave the same verbal instructions to
participants to sit down and wait for the experimenter. If the participant
asked the experimenter where he or she should sit, the experimenter always
responded, “Doesn’t matter . . . wherever you want.” If participants lin-
gered and did not sit down, she said, “You can take a seat now.” After the
participant chose a seat, the experimenter then told the participant that she
would leave to find the other experimenter.
Diagram of experimental room for Experiment 6.
EXCLUSION AND DECONSTRUCTION
were not significantly different from each other—future misfor-
tune versus no feedback, ?2(1, N ? 20) ? 0.22; future misfortune
versus future belonging, ?2(1, N ? 20) ? 1.82; no-feedback versus
future belonging, ?2(1, N ? 20) ? 3.20. However, the future
belonging group did not differ significantly from the future alone
group, ?2(1, N ? 20) ? 1.25. This result was not expected.
Perhaps hearing feedback about one’s future relationships, whether
good or bad, produces a tendency to avoid self-awareness. As the
other contrasts demonstrate, negative feedback in and of itself does
not produce the effect.
Despite a very clear behavioral preference to face the wall rather
than to face the mirror (90% vs. 10%), people who had been
excluded were no more likely than others to self-report a stronger
preference for their choice, F(3, 38) ? 2.32 (rated on a 9-point
Likert scale: future alone M ? 4.40; future belonging M ? 3.00;
future misfortune M ? 5.00; no feedback M ? 4.90). Thus, future
alone and control participants were equally certain about their
seating choice. We also coded the participants’ open-ended rea-
sons for choosing the chair into three categories: no reason, prox-
imity (which was counterbalanced and thus cannot account for the
effect), and mention of wanting to look at or avoid the mirror.
There were no significant differences among the conditions in their
reasons for choosing one chair instead of another.
Mood was assessed using the BMIS (Mayer & Gaschke, 1988),
and results showed that there were no mood differences among
conditions on the valence subscale, F(3, 38) ? 0.18, ns, or the
arousal subscale, F(3, 38) ? 1.20, ns. Thus, the manipulation of
social exclusion did not produce differences in mood. This is
consistent with the previous experiments.
Also consistent with the previous experiments, mood did not
mediate the effect. After controlling for both the valence and
arousal subscales of the BMIS, social exclusion (future alone vs.
the control groups) significantly predicted the self-awareness de-
pendent measure, r(37) ? .38, p ? .02. In addition, neither mood
valence nor arousal was significantly correlated with avoiding
self-awareness after exclusion condition was controlled.
The social exclusion manipulation also had no impact on par-
ticipants’ level of state self-esteem. Scores on the SSES did not
differ significantly by condition for either the total state self-
esteem score, F(3, 38) ? 0.04, ns, or any of the three subscales
(performance self-esteem, social self-esteem, or appearance self-
esteem). Nevertheless, we tested for mediation. After controlling
for state self-esteem scores, social exclusion (future alone vs.
control groups) significantly predicted avoidance of self-
awareness, r(37) ? ?.41, p ? .02. Exclusion predicts the avoid-
ance of self-awareness when state self-esteem is controlled, and
thus state self-esteem does not mediate the relationship. However,
state self-esteem was significantly correlated with self-awareness
avoidance when social exclusion was controlled, r(37) ? ?.38,
p ? .02. This suggests that state self-esteem and social exclusion
each had independent effects on the avoidance of self-awareness.
What happens inside the psyche of a person who is rejected by
others? Intuition and theory had previously led us to expect that
emotional distress would be a major, powerful result and that the
behavioral consequences of social exclusion would be mediated by
that distress. Yet, the data failed to support these theories and
intuitions. Across multiple studies, our manipulations of social
exclusion produced either weak or nonsignificant effects on emo-
tional distress. Moreover, emotion consistently failed to show any
sign of mediating the behavioral effects. It remains plausible that
delayed emotional distress may be caused by rejection, but emo-
tion certainly does not appear fast enough to affect behavior.
The present investigation was stimulated by the search for an
alternative to emotional distress as an inner effect of social exclu-
sion. We began by noting parallels to the research literature on
suicide, which had also encountered many failures to confirm the
intuitively and theoretically plausible prediction that the psyches
of suicidal individuals would be brimming with intense, negative
emotions. Although suicidal individuals were not found to be
happy, cheerful souls, neither were they filled with anguish. The
most common finding was that they showed emotional numbness,
as if their emotional system had somehow shut down (see
Baumeister, 1990, for a review). The responses of our laboratory
participants had also indicated numbness, and the present findings
confirmed this. In Experiment 5, for example, socially excluded
participants chose fewer emotion words in an implicit measure of
emotion disguised as a subliminal perception task. This is very
similar to results with suicidal patients, who found it more difficult
to access emotional memories and seemed the most comfortable
with neutral stimuli (Geller & Atkins, 1978; J. M. Williams &
Broadbent, 1986). This lack of emotional distress was also evident
in Experiments 1–3 and 6, which showed null effects for exclusion
on four different self-report measures; the only exception was a
small difference on a one-item mood measure in Experiment 1.
The Effect of Social Exclusion on Avoiding Self-Awareness, Experiment 6
Experimental conditionChair facing mirrorChair facing wall
Control conditions combined
10% (n ? 1)
53% (n ? 16)
70% (n ? 7)
60% (n ? 6)
30% (n ? 3)
90% (n ? 9)
47% (n ? 14)
30% (n ? 3)
40% (n ? 4)
70% (n ? 7)
?2(1, N ? 40) ? 5.76*
?2(3, N ? 40) ? 9.31*
* p ? .05.
TWENGE, CATANESE, AND BAUMEISTER
Several other labs working on rejection and ostracism have
found significant mood effects (e.g., Buckley et al., in press; Snapp
& Leary, 2001; K. D. Williams, Cheung, & Choi, 2000), and it
seems prudent to discuss why our results differ. It is possible that
the future alone manipulation (unique to us), which predicts an
entire future of exclusion, produces stronger feelings of defense
and deconstruction than the other types of manipulations. In ad-
dition, our group rejection procedure differs from that of Buckley
et al. (in press). In those studies, participants never met their
rejecting peers in person (instead, communication took place on
paper or by speaking into a microphone without verbal feedback in
return). Our participants met and talked with their peers in person
before the rejection, which might have elicited a stronger defense
response. Further research should attempt to determine if mood
effects differ systematically on the basis of the specific rejection,
exclusion, or ostracism manipulation used.
The main purpose of the present investigation was to test a
series of predictions based on cognitive deconstruction. Decon-
struction is a mental state found in suicidal individuals. It is
assumed to reflect an attempt to avoid awareness of the self’s
deficiencies (as suggested by recent events—in this case, being
rejected by others) along with the emotional distress that might
accompany facing up to one’s shortcomings and failures. The
central reasoning was that social exclusion would thwart people’s
desire to be accepted by others, and consequently rejected people
would enter the defensive state of cognitive deconstruction (in
which the emotional distress and aversive self-awareness could be
kept at bay). A series of predictions about the behavior of socially
excluded individuals was confirmed. We now review the main
Findings Consistent With the Deconstructed State
Time distortion is one sign of the deconstructed state. Time
perception was altered among socially rejected participants in
Experiment 1. They significantly overestimated the duration of
experimentally timed intervals, in sharp contrast to the fairly
accurate estimates in the control group. Overestimating the dura-
tion of intervals occurs when time is passing very slowly, such as
when one is severely bored or life is generally empty.
The altered sense of time was also reflected in the lack of future
orientation. On one measure, rejected participants in Experiment 1
were more likely to report that they were focused on the present
and found it difficult to think about the future (although another
measure, the ZTPI, did not show significant differences). The
blockage of thoughts about the future, like the slowed passage of
time, has been observed among suicidal people and is a defining
feature of deconstruction. The meaningful construction of human
life links different episodes across time, and the present often
draws meaning from the future (e.g., when people explain their
current activities in terms of goals or other anticipated future
outcomes). Hence, much of the meaning of present events is
eliminated when the future is blocked out of awareness.
The loss of meaning was confirmed with a different measure in
Experiment 2. Socially rejected people did not disagree as strongly
with the statement “Life is meaningless.” These findings confirm
and extend K. D. Williams’s (2001) theory that ostracism threatens
meaningful existence. Even when participants were randomly as-
signed to briefly experience rejection by a group of strangers, their
views of life were affected. The possibility that life has no meaning
became more probable. This fits the view that social exclusion
produces a state incompatible with meaning—thus, a decon-
structed state. If such a small brush with social rejection can make
life seem more meaningless, long-term experiences of ostracism
and exclusion are likely to produce deep-seated feelings of mean-
inglessness and despair.
Delay of gratification is an important key to success in life, but
it requires a meaningful connection between the present (where
sacrifices have to be made) in favor of the future (where goals are
met and rewards anticipated). The deconstructed state is marked by
disinhibited, impulsive behavior, which is presumed to reflect this
lack of meaningful guidance of present behavior by future consid-
erations. Experiment 1 measured delay of gratification by asking
people how they would advise someone choosing between two
jobs, one of which was better in the short run, whereas the other
was better in the long run. Social exclusion led to a significant shift
toward the job that was better in the short run (immediate gratifi-
cation). Nearly all control participants favored the job that was
better in the long run, but a sizeable minority of socially rejected
participants voted in favor of the job with the short-term benefits.
These findings converge with those reported by Twenge et al.
(2002), in which excluded participants chose the short-term plea-
sures of magazines, candy bars, and video games over the long-
term benefits of good health and doing well on an upcoming test.
Socially excluded people also displayed lethargy and passivity,
other hallmark behaviors of the deconstructed state. In Experi-
ment 3, excluded participants wrote fewer words when asked to
define common proverbs, thus choosing to exert less effort at a
cognitive task. Rejected participants in Experiment 4 demonstrated
slowed reaction time on the first trial of a game, again showing
lethargy in response to a novel task. Taken together, these studies
show that social exclusion leads individuals to slow down, con-
serve energy, and generally behave in a lethargic and passive
fashion. These results are consistent with those of K. D. Williams,
Cheung, and Choi (2000), who found greater conformity among
ostracized individuals. However, they are inconsistent with the
results of K. D. Williams and Sommer (1997), who found that
ostracized women worked harder on a subsequent group task.
Future research should determine which rejection and ostracism
situations lead to passivity and which do not.
We examined the role of mood in several studies. With the
exception of the one-item measure in Experiment 1, none of the
explicit measures of mood were affected by the social exclusion
and rejection manipulations. In Experiment 5, participants com-
pleted an implicit mood measure, guessing which word they saw
momentarily displayed on a computer screen. Excluded partici-
pants chose fewer emotion words, consistent with our prediction
that exclusion leads to emotional numbness.
The ostensible purpose of the deconstructed state is to avoid
emotion and self-awareness. We have already mentioned that the
present findings confirmed the absence of emotion. Experiment 6
was designed to measure avoidance of self-awareness. Consistent
with predictions, socially excluded participants showed a marked
aversion for seats that faced a mirror (a self-focusing cue). Only
one single participant in the rejection condition took the seat that
faced the mirror. In contrast, control subjects seemed essentially
indifferent as to whether they faced toward or away from the
mirror, with about half the control participants choosing each seat.
EXCLUSION AND DECONSTRUCTION
Social exclusion caused avoidance of mirrors, and this suggests
that excluded people are averse to self-awareness.
The present investigation has identified some intrapsychic con-
sequences of social exclusion. Previous experiments found the
emotional impact of exclusion to be muted, but the present find-
ings indicate that there are in fact some discernible intrapsychic
effects. These fit the pattern of a state of deconstruction, in which
meaningful thought and self-awareness are avoided, attention is
limited to the present, and action is slow and sparse. Deconstruc-
tion is a defensive state of mind, designed to ward off the negative
consequences of rejection. This state may be mildly aversive, but
it is preferred over the alternative of intense emotional distress and
detailed ruminations about all one’s misdeeds and faults that might
have caused the rejection.
Baumeister, R. F. (1990). Suicide as escape from self. Psychological
Review, 97, 90–113.
Baumeister, R. F. (1991). Escaping the self: Alcoholism, spirituality,
masochism, and other flights from the burden of selfhood. New York:
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for
interpersonal attachments as a fundamental human motivation. Psycho-
logical Bulletin, 117, 497–529.
Baumeister, R. F., Twenge, J. M., & Nuss, C. K. (2002). Effects of social
exclusion on cognitive processes: Anticipated aloneness reduces intelli-
gent thought. Journal of Personality and Social Psychology, 83, 817–
Blewett, A. E. (1992). Abnormal subjective time experience in depression.
British Journal of Psychiatry, 161, 195–200.
Brockopp, G. W., & Lester, D. (1970). Time perception in suicidal and
nonsuicidal individuals. Crisis Intervention, 2, 98–100.
Buckley, K. E., Winkel, R. E., & Leary, M. R. (in press). Reactions to
acceptance and rejection: Effects of level and sequence of relational
evaluation. Journal of Experimental Social Psychology.
Bushman, B. J., & Baumeister, R. F. (1998). Threatened egotism, narcis-
sism, self esteem, and direct and displaced aggression: Does self-love or
self-hate lead to violence? Journal of Personality and Social Psychol-
ogy, 75, 219–229.
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience.
New York: Harper & Row.
Diener, E., & Wallbom, M. (1976). Effects of self-awareness on antinor-
mative behavior. Journal of Research in Personality, 10, 107–111.
Durkheim, E. (1963). Suicide. New York: Free Press. (Original work
Edwards, M. J., & Holden, R. R. (2001). Coping, meaning in life, and
suicidal manifestations: Examining gender differences. Journal of Clin-
ical Psychology, 57, 1517–1534.
Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual of the Eysenck
Personality Questionnaire. San Diego, CA: EdITS.
Geller, A. M., & Atkins, A. (1978). Cognitive and personality factors in
suicidal behavior. Journal of Consulting and Clinical Psychology, 46,
Gerber, K. E., Nehemkis, A. M., Farberow, N. L., & Williams, J. (1981).
Indirect self-destructive behavior in chronic hemodialysis patients. Sui-
cide and Life-Threatening Behavior, 11, 31–42.
Gjesme, T. (1979). Future time orientation as a function of achievement
motives, ability, delay of gratification, and sex. Journal of Psychology,
Greaves, G. (1971). Temporal orientation in suicidals. Perceptual and
Motor Skills, 33, 1020.
Greenberg, J., & Musham, C. (1981). Avoiding and seeking self-focused
attention. Journal of Research in Personality, 15, 191–200.
Hass, R. G., Katz, I., Rizzo, N., Bailey, J., & Moore, L. (1992). When
racial ambivalence evokes negative affect, using a disguised measure of
mood. Personality and Social Psychology Bulletin, 18, 786–797.
Heatherton, T. F., & Polivy, J. (1991). Development and validation of a
scale for measuring state self-esteem. Journal of Personality and Social
Psychology, 60, 895–910.
Hendin, H. (1982). Suicide in America. New York: Norton.
Henken, V. J. (1976). Banality reinvestigated: A computer-based content
analysis of suicidal and forced-death documents. Suicide and Life-
Threatening Behavior, 6, 36–43.
Kirkpatrick, L. A., Waugh, C. E., Valencia, A., & Webster, G. D. (2002).
The functional domain specificity of self-esteem and the differential
prediction of aggression. Journal of Personality and Social Psychol-
ogy, 82, 756–767.
Koole, S. L., Smeets, K., von Knippenberg, A., & Dijksterhuis, A. (1999).
The cessation of rumination through self-affirmation. Journal of Per-
sonality and Social Psychology, 77, 111–125.
Kuhlen, R. G., & Monge, R. H. (1968). Correlates of estimated rate of time
passage in the adult years. Journal of Gerontology, 23, 427–433.
Leary, M. R., Kowalski, R. M., Smith, L., & Phillips, S. (2003). Teasing,
rejection, and violence: Case studies of the school shootings. Aggressive
Behavior, 29, 202–214.
Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs, D. L. (1995).
Self-esteem as an interpersonal monitor: The sociometer hypothesis.
Journal of Personality and Social Psychology, 68, 518–530.
Massong, S. R., Dickson, A. L., Ritzler, B. A., & Layne, C. C. (1982).
Assertion and defense mechanism preference. Journal of Counseling
Psychology, 29, 591–596.
Mayer, J. D., & Gaschke, Y. N. (1988). The experience and meta-
experience of mood. Journal of Personality and Social Psychology, 55,
McGuire, M. T., & Raleigh, M. J. (1986). Behavioral and physiological
correlates of ostracism. Ethology and Sociobiology, 7, 187–200.
Mehrabian, A., & Weinstein, L. (1985). Temperament characteristics and
suicide attempters. Journal of Consulting and Clinical Psychology, 53,
Melges, F. T., & Weisz, A. E. (1971). The personal future and suicidal
ideation. Journal of Nervous and Mental Disease, 153, 244–250.
Neuringer, C., & Harris, R. M. (1974). The perception of the passage of
time among death-involved hospital patients. Life-Threatening Behav-
ior, 4, 240–254.
Rogers, J. R. (2001). Theoretical grounding: The “missing link” in suicide
research. Journal of Counseling & Development, 79, 16–25.
Sedikides, C., Campbell, W. K., Reeder, G. D., & Elliot, A. J. (1999). The
relationship closeness induction task. Representative Research in Social
Psychology, 23, 1–4.
Snapp, C. M., & Leary, M. R. (2001). Hurt feelings among new acquain-
tances: Moderating effects of interpersonal familiarity. Journal of Social
and Personal Relationships, 18, 315–326.
Topol, P., & Reznikoff, M. (1982). Perceived peer and family relation-
ships, hopelessness, and locus of control as factors in adolescent suicide
attempts. Suicide and Life-Threatening Behavior, 12, 141–150.
Trout, D. L. (1980). The role of social isolation in suicide. Suicide and
Life-Threatening Behavior, 10, 10–23.
Twenge, J. M., Baumeister, R. F., Tice, D. M., & Stucke, T. S. (2001). If
you can’t join them, beat them: Effects of social exclusion on aggressive
behavior. Journal of Personality and Social Psychology, 81, 1058–1069.
Twenge, J. M., & Campbell, W. K. (2003). “Isn’t it fun to get the respect
that we’re going to deserve?”: Narcissism, social rejection, and aggres-
sion. Personality and Social Psychology Bulletin, 29, 261–272.
TWENGE, CATANESE, AND BAUMEISTER
Twenge, J. M., Catanese, K. R., & Baumeister, R. F. (2002). Social Download full-text
exclusion causes self-defeating behavior. Journal of Personality and
Social Psychology, 83, 606–615.
Tysk, L. (1984). Time perception and affective disorders. Perceptual and
Motor Skills, 58, 455–464.
Vallacher, R. R., & Wegner, D. M. (1985). A theory of action identifica-
tion. Hillsdale, NJ: Erlbaum.
Vallacher, R. R., & Wegner, D. M. (1987). What do people think they’re
doing: Action identification and human behavior. Psychological Re-
view, 94, 3–15.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali-
dation of brief measures of positive and negative affect: The PANAS
scales. Journal of Personality and Social Psychology, 54, 1063–1070.
Wegner, D. M., & Giulano, T. (1980). Arousal-induced attention to self.
Journal of Personality and Social Psychology, 38, 719–726.
Weiss, J. M. (1957). The gamble with death in attempted suicide. Psychi-
atry, 20, 17–25.
Williams, J. M., & Broadbent, K. (1986). Autobiographical memory in
suicide attempters. Journal of Abnormal Psychology, 95, 144–149.
Williams, K. D. (2001). Ostracism: The power of silence. New York:
Williams, K. D., Bernieri, F., Faulkner, S., Grahe, J., & Gada-Jain, N.
(2000). The Scarlet Letter study: Five days of social ostracism. Journal
of Personal and Interpersonal Loss, 5, 19–63.
Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). CyberOstracism:
Effects of being ignored over the Internet. Journal of Personality and
Social Psychology, 79, 748–762.
Williams, K. D., Govan, C. L., Croker, V., Tynan, D., Cruickshank, M., &
Lam, A. (2002). Investigations into differences between social- and
cyber-ostracism. Group Dynamics: Theory, Research, and Practice, 6,
Williams, K. D., Shore, W. J., & Grahe, J. E. (1998). The silent treatment:
Perceptions of its behaviors and associated feelings. Group Processes
and Intergroup Relations, 1, 117–141.
Williams, K. D., & Sommer, K. L. (1997). Social ostracism by one’s
coworkers: Does rejection lead to loafing or compensation? Personality
and Social Psychology Bulletin, 23, 693–706.
Wyrick, R. A., & Wyrick, L. C. (1977). Time experience during depres-
sion. Archives of General Psychiatry, 34, 1441–1443.
Yufit, R. I., & Benzies, B. (1973). Assessing suicidal potential by time
perspective. Life-Threatening Behavior, 3, 270–282.
Zimbardo, P. G., & Boyd, J. N. (1999). Putting time in perspective: A
valid, reliable individual-differences metric. Journal of Personality and
Social Psychology, 77, 1271–1288.
Received November 18, 2002
Revision received April 14, 2003
Accepted April 23, 2003 ?
EXCLUSION AND DECONSTRUCTION