USER
Do not answer yet. This is just another part of the text I want to send you. Just receive and acknowledge as "Part 1/2 received" and wait for the next part.
[START PART 1/2]
Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The Causes and Consequences of a Need for Self-Esteem: A Terror Management Theory . Public Self and Private Self, 189–212. https://doi.org/10.1007/978-1-4613-9564-5_10 :
"Throughout the past few thousand years, historical accounts, philosophical treatises, and works of fiction and poetry have often depicted humans as having a need
to perceive themselves as good, and their actions as moral and justified. Within the
last hundred years, a number of important figures in the development of modem psychology have also embraced this notion that people need self-esteem (e.g., Adler,
1930; Allport, 1937; Homey, 1937; James, 1890; Maslow, 1970; Murphy, 1947;
Rank, 1959; Rogers, 1959; Sullivan, 1953). Of these, Karen Homey most thoroughly discussed the ways people try to attain and maintain a favorable self-image.
The clinical writings of Homey, and other psychotherapists as well, document the
ways in which people attempt to defend and enhance self-esteem; they also suggest
that difficulty maintaining self-esteem, and maladaptive efforts to do so, may be
central to a variety of mental health problems. In this chapter, we will first review
the research supporting the existence of a need for self-esteem. Then we will present
a theory that accounts for this need and specifies the role it plays in a variety of
phenomena including self-presentation.
Empirical Support for a Need for Self-Esteem
It was not until the early 1950s that researchers began to obtain quantitative evidence concerning the need for self-esteem. Since then, research on psychopathology
has shown low self-esteem to be associated with a variety of psychological problems,
including alcoholism, anxiety, depression, neuroticism, and schizophrenia (see
Wylie, 1979, for a review). Such findings suggest that people do need self-esteem for
healthy psychological functioning; however, this research is correlational and there-
190 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
fore subject to a number of alternative explanations. For example, it may be that psychological difficulties lead to low self-esteem, or that whatever factors contribute to
such difficulties also cause low self-esteem.
The Self-Serving Bias in Causal Attribution
Clearer support for the existence of a self-esteem need has been found in experimental research, beginning with studies demonstrating a self-serving bias in individuals'
causal attributions for their own successes and failures. These studies typically
entail randomly assigning subjects to experience either a favorable or unfavorable
outcome on a test and then obtaining their estimates of the extent to which potential
causal factors such as ability, effort, luck, and task difficulty were responsible for
their particular outcomes. The one highly consistent finding has been that individuals who experience success assign greater responsibility to factors within themselves (e.g., internal factors, such as ability) and less responsibility to factors outside
themselves (e.g., external factors, such as luck) than do individuals who experience
failure (e.g., Johnson, Feigenbaum, & Welby, 1964; Miller, 1976; Stephan, Rosenfield, & Stephan, 1976; also see Bradley, 1978, and Zuckerman, 1979, for reviews).
Thus, as would be expected if people do indeed have a need for self-esteem,
individuals seem to take credit for success but deny responsibility for failure.
Eliminating cognitive alternative explanations. In the last 10 years, however, a set
of alternative explanations for this self-serving bias has received considerable attention. Miller and Ross (1975) proposed a number of cognitive mechanisms that could
lead to a self-serving attributional bias in the absence of a self-esteem motive.
Despite the fact that the notion of a need for self-esteem was the basis for all of the
research on the self-serving bias up to that point in time, it was argued that such cognitive explanations are preferable because they rely only on the widely accepted
information-processing framework for understanding human behavior (see also
Nisbett & Ross, 1980). The most compelling of these cognitive explanations is that
if individuals believe they are generally competent, they are likely to infer that when
they perform competently it is because of their abilities, and when they perform
poorly it is because of external factors.
However, since 1975, a number of studies have provided support for the selfesteem explanation of the self-serving bias. McFarland and Ross (1982) conducted
a study in which subjects were led to attribute success or failure to either their level
of ability or the characteristics of the test. Success-internal subjects reported more
positive affect, less negative affect, and higher self-esteem than did success-external
subjects; failure-internals, on the other hand, reported less positive affect, more
negative affect, and lower self-esteem than did failure-external subjects. Correlational research (e.g., Arkin & Maryuma, 1979; Feather, 1969) and research using
hypothetical outcomes (e.g., Nicholls, 1976; Weiner, Russell, & Lerman, 1978,
1979) have found similar effects. These studies show that the self-serving pattern of
attributions does indeed increase the favorability of the consequences of outcomes
for affective experience and self-esteem. If people are motivated to maximize positive affect and minimize negative affect, a motivational influence on attributions for
Need for Self-Esteem 191
performance outcomes is likely. Furthermore, self-serving attributions do seem to
help the individual maintain self-esteem.
Other research provides more direct support for the self-esteem explanation by
demonstrating that the intensity of the affective consequences of an outcome alters
the extent to which people's attributions are self-serving. Stephan and Gollwitzer
(1981) found that subjects led to believe a placebo pill they had taken would produce
autonomic arousal were less prone to make self-serving attributions for a prior success or failure than were no placebo subjects. In a similar vein, Fries and Frey (1980)
found that subjects were less likely to derogate a test after failure if they could attribute arousal caused by the failure to a nonthreatening source.
In another study, Stephan and Gollwitzer (1981) provided subjects with false feedback concerning their level of physiological arousal after a success or failure
experience. Subjects led to believe they were highly aroused were more self-serving
in their performance attributions than were low-arousal feedback subjects. Finally,
Gollwitzer, Earle, and Stephan (1982) demonstrated that unlabeled residual arousal
from physical exercise also increased the self-serving nature of subjects' attributions. Presumably, this residual arousal intensified subjects' perceptions of
outcome-related affect, thus motivating them to become more self-serving in their
attributions. To the extent that such affective consequences are attributed to the
implications of the outcome for self-esteem, these studies show that the self-serving
bias is mediated by self-esteem concerns.
Other Self-Esteem Maintenance Strategies
The resolution in favor of a motivational explanation for the self-serving bias is quite
similar to the resolution of the earlier dissonance versus self-perception controversy.
A number of studies have clearly shown that negative affect does playa role in attitude change after individuals feel responsible for engaging in behavior that has foreseeable negative consequences (see Fazio & Cooper, 1983, for a review).
Interestingly, the findings of attitude change that have been attributed to a dissonance reduction process can be interpreted as resulting from a need to defend selfesteem (see Wicklund & Brehm, 1976, for a review of the dissonance literature). In
virtually all of these studies, the attitude change can be viewed as a way for the
individual to deny that he or she has done something that is eitherimmoral or stupid
(see Aronson, 1968; Bowerman, 1978; Schlenker, Forsyth, Leary, & Miller, 1980;
Steele & Liu, 1983).
In recent years, a substantial number of other studies have also yielded findings
supportive of the existence of a need for self-esteem. Phares and Lamiell (1974) and
Berglas and Jones (1978) reasoned that if individuals try to protect self-esteem by
attributing failures to external factors, when failures are anticipated, they may set
up plausible external attributions by engaging in performance-inhibiting behavior
prior to and during performance. Evidence for the use of this strategy, which Berglas
and Jones (1978) have aptly labeled self-handicapping, has been obtained in a number of studies (Berglas & Jones, 1978; Frankel & Snyder, 1978; Kolditz & Arkin,
1982; Phares & Lamiell, 1974; Snyder, Smoller, Strenta, & Frankel, 1981; Tucker,
Vuchinish, & Sobell, 1981). Such findings attest to the strength of self-esteem
192 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
needs, because they show that people will actually undermine their own chances for
success to be sure that they have justification for a self-esteem-protecting external
attribution should they subsequently fail.
Research has also demonstrated other types of efforts to establish self-esteemprotective attributions, by showing that individuals will report a variety of states and
traits when they believe these factors could serve as excuses for subsequent failure
(Greenberg, Pyszczynski, & Paisley, 1984; Pyszczynski & Greenberg, 1983; Smith,
Snyder, & Handelsman, 1982; Smith, Snyder, & Perkins, 1983; Snyder, Smith,
Angelli, & Ingram, in press). For example, Smith et al. (1982) and Greenberg et al.
(1984) have shown that, for individuals who are highly concerned about possible
failure on an upcoming test, reports of test anxiety are reduced if they are told that
test anxiety does not affect performance on the particular test. Thus, individuals
report more test anxiety when it can serve as an excuse for subsequent failure than
when it cannot.
Studies have also shown that, under certain conditions, once a shortcoming on a
specific dimension becomes salient, individuals compensate by overevaluating
themselves on unrelated dimensions (e.g., Baumeister & Jones, 1978; Greenberg &
Pyszczynski, 1985; Pyszczynski, Greenberg, & Beck, 1986) or, if the dimension is
very important to the individual, by engaging in positive self-descriptions regarding
the threatened dimension (e.g., Gollwitzer & Wicklund, 1985; Gollwitzer, Wicklund, & Hilton, 1982; Wicklund & Gollwitzer, 1981). Thus, people alter their selfperceptions to maintain self-esteem it is threatened.
Individuals also alter their perceptions of and comparisons with others to maintain
self-esteem. Research on self-serving attributive projection has shown that people
overestimate consensus for their poor performances and underestimate consensus
for their good performances, especially when the performance is on an ability
dimension of personal importance (Campbell, in press). In other words, when people perform poorly on an ego-involving task, they overestimate how many others
would also perform poorly; when people perform well on an ego-involving task they
underestimate how many others would also perform well. In addition to a selfserving bias in perceived consensus, Pyszczynski, Greenberg, and LaPrelle (1985)
have found a self-serving bias in information search such that, after receiving a performance outcome, individuals search more extensively for social comparison information if they expect to find that others performed poorly than if they expect to find
that others performed well. It has also been found that when individuals are exposed
to social comparison information, they prefer to compare themselves with others
who have performed worse on a salient dimension than with others who have performed better (see Wills, 1981, for a review). Perhaps as a consequence of this
downward comparison process, people believe they are better than average on a wide
variety of dimensions (Felson, 1981).
Guided by a self-evaluation maintenance model (Tesser, 1980; Tesser & Campbell, 1983; Tesser & Moore, in this volume), Tesser and his colleagues have demonstrated a number of other ways in which individuals seem to adjust the nature of their
social comparisons to protect self-esteem. This line of research has yielded some
compelling support for the operation of self-esteem maintenance processes in social
Need for Self-Esteem 193
behavior (see Tesser and Moore, Chapter 5, this volume, for a more complete
account). Tesser and Campbell (1980) had subjects perform two different tasks with
a confederate posing as another subject; on one of the tasks they performed equally
but on the other task the confederate outperformed the subject. Subjects reduced the
personal importance (Le., relevance) of the dimension on which the confederate
outperformed them, especially when the confederate had been described as similar
to the subjects; thus, they minimized the threat to self-esteem of being outperformed
by a similar other. Similarly, in a correlational study of high school students, Tesser
and Campbell (1982) found that low performance in school relative to similar but
not dissimilar classmates is associated with low ratings of the personal relevance of
school performance. Further evidence that individuals deny the personal relevance
of threatened attributes has been obtained by Greenberg, Pyszczynski and Solomon
(1982), Greenberg and Pyszcynski (1985) and Tesser and Paulhus (1983).
Tesser (1980) obtained evidence of a related process in male sibling relations. For
males whose siblings were generally more competent than themselves, the closer
the sibling was in age (and therefore the more appropriate for social comparison),
the less the males perceived themselves as similar to their siblings. This inverse relationship was not found for males who were not less competent than their siblings.
Presumably, males with superior siblings close in age attempted to deny the
appropriateness of a threatening social comparison; consistent with this reasoning,
these individuals also reported more friction in their relationships with their siblings
than any of the other subjects in the study.
A laboratory study by Pleban and Tesser (1981) obtained further support for this
process. They found that, when the performance dimension was low in relevance;
the more a confederate outperformed a subject, the more the subject perceived the
confederate as similar to himself or herself. This finding can be viewed as an attempt
to bask in reflected glory, just as identification with a football team increases with
success of the team (Cialdini et aI., 1976). However, when the performance was
high in relevance and, therefore, similarity would imply that a potentially threatening social comparison would be appropriate, the more the confederate outperformed
a subject, the less the subject perceived the confederate as similar to himself or herself. As in the sibling study, when threatened with a self-esteem-damaging social
comparison, individuals deny similarity to the potential comparison other.
Evidence for Self-Esteem Maintenance or Public Impression Management?
To summarize, a substantial body of evidence has accumulated suggesting that, in
order to maintain self-esteem, we alter our self-referent causal attributions, our selfreports of states and traits, our performance-related behavior, our attitudes and
beliefs, and our social perceptions and comparisons. Despite this impressive array
of support for a need for self-esteem, an alternative explanation based on impression
management warrants consideration. Indeed, much of the aforementioned evidehce
can be explained by a need to protect public image rather than private self-image.
Instead of needing self-esteem, perhaps individuals simply need to maintain a positive image in the eyes of others. This reasoning has been used to explain the evidence
194 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
for the self-serving bias (Bradley, 1978), dissonance reduction (e.g., Schlenker et
aI., 1980; Tedeschi, Schlenker, & Bonoma, 1971; Tedeschi & Rosenfeld, 1981),
compensation effects (Baumeister & Jones, 1978), and self-handicapping (Kolditz &
Arkin, 1982). Research has shown that individuals do engage in a variety of
behaviors to manage impressions for others (see Baumeister, 1982; Jones & Pittman, 1982; Schlenker, 1980; and Tedeschi, 1981 for reviews). Therefore, given the
public nature of most laboratory research, it is possible that many of the findings
attributed to a need for self-esteem actually resulted from self-presentational concerns about public image.
However, a number of recent studies have demonstrated the self-serving attributional bias under conditions in which attributions were unlikely to be influenced by
such self-presentational concerns (Greenberg et aI., 1982; House, 1980; Ries,
Rosenfeld, Melburg, & Tedeschi, 1981; Weary et aI., 1982). For example, Greenberg et aI. (1982) created private success and failure by giving groups of subjects an
intelligence test they could score themselves, and instructing them to conceal their
scored answer sheets and keep them when they left the study. Subjects were then
asked on an anonymous questionnaire to attribute causal responsibility to potential
factors. In this way their attributions would have no clear implications for public
image unless others knew whether they had succeeded or failed. Under these private
conditions, a strong self-serving bias was found. There were also some indications
that under public conditions the bias was weaker. Along with a private self-serving
attributional bias, this study, and a study by Frey (1978) as well, found a self-serving
tendency for subjects to privately evaluate a test as more valid if they had succeeded
than if they had failed.
These studies suggest that a need for self-esteem, apart from a desire for public
esteem, seems to underlie the self-serving bias. Recent research has shown that a
number of other self-esteem maintenance strategies also function to protect one's
private self-image. With regard to anticipatory attributional defenses, Greenberg et
aI. (1984) found that self-reports of test anxiety to set up an excuse for possible
future failure occurred under anonymous conditions and, furthermore, did not
occur when the incentive for success was high. The latter finding suggests that the
self-reports of test anxiety were privately believed because they did not occur when
subjects did not want their chances for success hindered; if the self-reports of test
anxiety were simply to serve as potential public excuses, they would have been used
even when subjects were primarily concerned with succeeding on the test.
It has also been shown that a general increase in the favorability of self-evaluations
in response to a specific threat to self-esteem occurs even under conditions in which
such compensatory self-inflation could serve no public impression management
function (Greenberg & Pyszczynski, 1985). Similarly, research has recently found
attitude change after counterattitudinal behavior under bogus pipeline conditions
and under private conditions, thereby supporting an intrapsychic explanation of dissonance phenomena over a public impression management explanation (Baumeister
& Tice, 1984; Stults, Messe, & Kerr, 1984). Finally, Tesser and Paulhus (1983) have
shown that individuals deny the ego relevance of a test they have failed even when
they are led to believe that only they know they failed and that the experimenter
believes they succeeded.
Need for Self-Esteem 195
The findings of these studies cannot be accounted for by public impression
management explanations such as those posited by Bradley (1978), Tedeschi et al.
(1971) and others. Clearly, individuals do engage in a variety of strategies to maintain self-esteem even when public image is not at stake. However, this point is compatible with a number of self-presentation theories that propose that individuals
have a need to present a positive image to themselves as well as to others (e.g., Baumeister, 1982; Goffman, 1955; Schlenker, 1980). In fact, such a proposition is
indistinguishable from the notion that people need self-esteem. These theorists simply conceptualize self-esteem maintenance as analogous to, and perhaps a special
case of, maintaining a positive image for an audience. In contrast, we describe below
a theory that conceptualizes public image maintenance as a component of selfesteem maintenance.
Threats to Public Image Threaten Private Self-Esteem
The evidence reviewed above suggests that, even when public image is not at stake,
individuals use a number of strategies to protect self-esteem when it is threatened.
However, it has also been found that, when an outcome does threaten public image,
private efforts to protect self-esteem are especially vigorous (Apsler, 1975; Frey,
1978; Greenberg & Pyszczynski, 1985). Apsler (1975) found that, after females
were embarrassed by engaging in a series of silly behaviors in the presence of an
audience, they were especially likely to engage in a helpful act, even if no witnesses
to their silly behavior would know of their helpfulness. Apparently, after damage to
their public image, people need to engage in positive, socially desirable behaviornot to restore a positive image for those others, but to restore a positive image for
themselves (self-esteem). Similarly, Gollwitzer and Wicklund (1985) have found
that individuals threatened by undesirable personality feedback will describe themselves especially positively to someone unaware of the negative feedback, even if the
subject believes the other person prefers people who are self-deprecating. Clearly,
in this study, the subjects were not acting to gain public esteem but to restore selfesteem.
In addition, Frey (1978) demonstrated that the self-serving tendency to privately
evaluate a test as more valid after success than after failure was stronger if the performance outcomes were public than if they were private. Similarly, Greenberg and
Pyszczynski (1985) found that individuals greatly inflated the favorability of their
self-images on a private measure of self-esteem after public failure but not after private failure. Finally, Tesser and Paulhus (1983) found that subjects who thought the
experimenter believed they had performed poorly on a certain ability dimension privately reduced the personal relevance of the dimension, even if they knew they had
actually performed well and the experimenter's belief was erroneous. Thus, even
when only public image is threatened, individuals privately engage in self-esteem
defense.
To summarize, the empirical research on self-esteem maintenance shows that
individuals do engage in a variety of strategies to privately defend self-esteem when
it is privately threatened. On the other hand, it has also been shown that, when public esteem is threatened, individuals are particularly likely to engage in private self-
196 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
esteem maintenance strategies. To explain these phenomena, as well as other evidence concerning self-esteem maintenance, and the evidence of the deleterious concomitants oflow self-esteem, we require a theory that explains: (a) what self-esteem
is; (b) why we need it; (c) how self-esteem is affected by public esteem; and (d) how
the need for self-esteem affects social behavior.
A Terror Management Theory of the Need for Self-Esteem
Now that we have established the existence of a need for self-esteem and have
reviewed a variety of its manifestations, we outline such a theory. Our theory is
based largely on the writings of Ernest Becker, especially The Birth and Death of
Meaning (1962), The Denial of Death (1973), and Escapefrom Evil (1975). Becker
attempted to synthesize the ideas of a very diverse array of theorists in order to
understand the dynamics of human social behavior. In doing so, he found that Alfred
Adler, Norman Brown, Peter Berger and Thomas Luckmann, Charles Horton
Cooley, Sigmund Freud, Erving Goffman, Karen Homey, Soren Kierkegaarde,
George Herbert Mead, Friedrich Nietzsche, Otto Rank, and Harry Stack Sullivan
had all arrived at certain very compatible insights concerning the role of self-esteem
in social behavior. Becker built upon these insights in constructing his own theoretical conception of the human animal; consequently, the current theory has benefited
from them as well.
The Cultural Animal
Becker (1962) proposed that the need for self-esteem is uniquely human, and exists
because of our capacities for symbolic, temporal, and self-reflective thought.
Although these attributes have greatly enhanced our ability to survive in a wide variety of environments, they have also led to some uniquely human problems (see M.
B. Smith, 1978). Specifically, we have the capacity to wonder why we exist and to
consider the possibility that the universe is an uncontrollable, absurd setting in
which the only inevitability is our own ongoing decay toward absolute annihilation,
which, to make matters worse, could occur at any moment because of anyone of a
variety of chance events.
Becker (1973, 1975) argued that we would be paralyzed with terror if we could not
deny such a conception; therefore, over time, as our cognitive capacities increased,
we developed cultural world views that imbued the universe with order, predictability, meaning, and permanence. As Becker noted, all cultures provide a description
of how the world was created, a prescription for leading a good, meaningful life, and
some hope of immortality (see Rank, 1950). Thus, each culture provides the
individual with a relatively benign world view that allows for the denial of his or her
ultimate vulnerability and mortality.
From this perspective, humans are not unique because they are social animals, but
because they are cultural animals. Humans live within a shared symbolic conception
Need for Self-Esteem 197
of the universe that is ultimately determined by culture, and yet is believed to be an
absolutely accurate representation of reality by individuals within the culture. As
the source of meaning and value, the culture provides the individual with a basis for
valuing himself or herself. The individual can have a sense of worth to the extent that
she or he satisfies the cultural criteria for being good (valuable). Thus, self-esteem
consists of viewing oneself as valuable within the context of the universal drama conveyed by the culture.
Self-Esteem as a Cultural Anxiety-Buffer
I have seen the moment of my greatness flicker,
And I have seen the eternal FOotman hold my coat, and snicker,
And in short, I was afraid.
T. S. Eliot
But why do humans seem to have such a desperate and pervasive need to view themselves as valuable (i.e., to have self-esteem)? From birth through early childhood,
the only basis of safety and security is the care from the parents. They satisfy needs
and minimize anxiety. Indeed, human infants are completely dependent on their parents for survival. As infants develop into children, they begin to acquire selfconsciousness; along with this, the warmth and care of the parents becomes increasingly conditional. Therefore, they develop an understanding that as long as they are
good boys or good girls, they will receive good outcomes and be protected from bad
outcomes by their apparently omnipotent parents; they also learn that if they are bad
boys or bad girls, they risk loss of the ultimate care and protection of the parents
and, perhaps, annihilation by them. Consistent with this reasoning, Rochlin (1965)
has observed a transformation in children from dread of being abandoned to dread
of being worthless. As a consequence of this perceived contingency, a positive selfconcept becomes associated with feelings of warmth and security, and a negative
self-concept becomes associated with terror (Becker, 1962, 1973; Sullivan, 1953).
From this point on, in order to avoid feelings ofterror, individuals must believe they
are good (i.e., valuable); therefore, individuals need self-esteem to function with
minimal anxiety.
For the child, this sense of equanimity consists of believing that he or she is of
primary value to the parents. However, as the child's cognitive capacities increase,
he or she begins to realize that there are outcomes from which the parents cannot
protect him or her. The child may also begin to realize that his or her parents are
vulnerable creatures who will eventually die and cannot protect the child from a
variety of aversive experiences, including their own death. Thus, the child's basis
for equanimity is undermined and new means of dealing with the terror of annihilation must be developed. At this point, it is not sufficient to be loved by the
mortal parents; to restore equanimity, a superior basis of value and protection must
be found. Fortunately, the culture provides such a basis by providing values, standards and roles, conceptions of the world as just (Lerner, 1980), and the possibility
of immortality.
198 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
Essentially, once the task of imparting a sense of absolute value and, consequently,
immortality, becomes too great for the parents, it is transferred to the religious and
secular concepts, symbols, and authorities of the culture. Of course the parents
greatly facilitate this transformation by virtue of their intentional and unintentional
conveyance of the world view espoused by their culture. The child learns that, to
minimize terror, he or she must believe he or she is valuable and deserving within
the context of the culture to which he or she subscribes; thus, for the adult human,
self-esteem is a cultural-anxiety buffer.
Consistent with this conceptualization of self-esteem as an anxiety-buffer, a number of researchers have found that individuals who are chronically low in self-esteem
are especially prone to anxiety (e.g., Bledsoe, 1964; French, 1968; Lipsitt, 1958;
Rosenberg, 1965; Rosenberg & Simmons, 1972; Truax, Schuldt, & Wargo, 1968;
Winkler & Myers, 1963). The research reviewed earlier, showing that a state of
arousal labeled as negative affect mediates self-serving attributions and test evaluations following failure and attitude change following counterattitudinal behavior,
also supports this idea (for self-serving beliefs, see Fries & Frey, 1980; Gollwitzer,
Earle, & Stephan, 1982; Stephan & Gollwitzer, 1981; for dissonance phenomena,
see Fazio & Cooper, 1983, for a review). From the present perspective, these findings occurred because, by threatening self-esteem, failures and counterattitudinal
behaviors weakened individuals' cultural anxiety-buffers and thereby engendered
anxiety. We are suggesting, then, that the psychological significance of such events
goes far beyond their implications for the specific domain in which the threats
occur. This point is supported by the research demonstrating that under certain conditions individuals compensate for failures by generally inflating the favorability of
their self-images (e.g., Greenberg & Pyszczynski, 1985).
The Two Components of Self-Esteem
To summarize, because we can conceive of our ultimate vulnerability and mortality
and can anticipate a variety of horrifying experiences, we have the potential to be
paralyzed by terror at any moment in our lives. By elevating us above the rest of the
living world, and providing a view ofthe world as orderly, predictable, meaningful,
and permanent, culture allows for the possibility of minimizing our terror by denying our essential creatureliness (i.e., our impotence, vulnerability, and mortality).
This possibility is realized to the extent that we can feel we are valuable members
of the culture.
From this perspective, self-esteem is an anxiety-buffering sense of personal value
(or heroism, as Becker refers to it) that consists of two components: first, faith in a
particular cultural drama that portrays human life as meaningful, important, and
enduring; and second, belief that one plays a significant part in that drama. Each
component is essential for self-esteem and, thus, for adequate terror management.
It is fairly obvious that events that suggest we have shortcomings can threaten our
self-esteem; however, events that suggest that our cultural drama is not absolutely
valid threaten self-esteem as well. If the standards by which we imbue ourselves
Need for Self-Esteem 199
with value are questioned, our personal claims of value are questioned as well.
Therefore, we propose that a substantial portion of our social behavior is directed
toward either or both of the following two goals: sustaining faith in a cultural drama
that provides the basis for self-esteem and maintaining a sense of value within that
cultural world view. It may be fruitful, then, to consider the possible role of each of
these two goals in social behavior. In considering these issues we find it useful to distinguish between general maintenance processes and defenses against threat. In
order to keep the anxiety-buffer provided by self-esteem, one must continually
reaffirm one's value and one's faith in the absolute validity of one's world view. This
facilitates one's ability to cope with specific threats when they arise. When threats
do occur, the urgency of maintaining one's anxiety-buffer is greatly enhanced and a
wide range of defensive strategies aimed at defusing the threat or repairing the
damage it produced may be used.
Sustaining faith in the cultural drama. General maintenance. Volumes have been,
and still could be written on the innumerable ways in which faith in a given culture
is developed and maintained. We will not even attempt here to do this topic justice,
except to mention a few basic modes of cultural affirmation. Most socialization and
education of children serves to instill the values and world view of the culture. Formal and informal historical and religious teachings may be particularly directed
toward conveying a cultural conception of reality that provides order, meaning, and
the possibilities of significance and immortality.
Cultural symbols (e.g., in the United States, government officials, churches,
monuments, flags, currency, religious and historical artifacts) and cultural rituals
(e.g., in the United States, singing the national anthem, going to church, visiting
historical locations and theme parks, following news and sports events, fashion, and
entertainment) also play major roles in maintaining our faith in the reality, significance, and permanence of the cultural drama, because they objectify it and
demonstrate social consensus.
Defenses against threat. This theory implies that any experience that suggests that
our cultural drama is wrong, or that other versions of reality are equally valid,
threatens self-esteem and is therefore a source of anxiety. Consequently, such
experiences motivate us to eliminate the threat; by so doing, we can sustain faith in
the basis of our self-esteem.
Such threats can result from environmental events; for example, when one's culture predicts the world should end on a certain day but it does not. As Festinger,
Riecken, and Schachter (1956) have observed, such dramatic disconfirmations of a
central aspect of one's cultural world view often lead to renewed efforts to convince
oneself of the validity of the threatened belief. Threats to one's cultural drama more
commonly result, however, from the knowledge that others do not subscribe to the
same cultural drama. For example, outgroups often have very different beliefs and
values from the cultural mainstream. We suggest that the pervasive tendency of
ingroup members to display negative attitudes, beliefs, and behaviors toward out-
200 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
group members is an attempt to defuse the threat to one's own beliefs implied by the
existence of the outgroup.
Quite consistent with our theory, a substantial body of research has shown that
this bias against outgroup members results largely from the belief that they have
different cultural values and beliefs (e.g., Byrne & Wong, 1962; Goldstein & Davis,
1972; Moe, Nacoste, & Insko, 1981; Rokeach, 1968; Rokeach & Mezei, 1966; Silverman, 1974; Stein, Hardyck, & Smith, 1965). It has even been found that, with
the possible exception of intimate contacts, outgroupers who seem to share the
ingroup's values may be liked just as much as ingroupers (e.g., McKirnan, Smith, &
Hamayan, 1983). It is when individuals are confronted with others who view the
world quite differently that they are threatened and consequently react negatively.
Our theory suggests that such negative reactions are increasingly likely the more
compelling the alternative conception of reality appears to be and the more committed the outgroupers are to their views. The awareness of such people is threatening
because they call into question the absolute validity of one's own cultural drama; the
individual cannot maintain a sense of absolute personal value if the basis of such a
judgment is merely one of a wide variety of subjective world views, none any more
correct than the others. History is replete with examples of cultural efforts to eradicate such threats, ranging from derogating to proselytizing to annihilating. In fact,
most, if not all wars can be viewed as battles to determine whose cultural terrorshield is the right one (e.g., the Crusades, the wars between Moslems and Jews,
Hindus and Moslems, Protestants and Catholics, and the United States and the
Soviet Union). Consistent with this position, we suggest that, although political and
economic considerations have certainly played a role in many armed conflicts, it is
the ideological threat upon which leaders focus to motivate masses of people into
battle.
Although threats to the cultural drama at the intergroup level may be particularly
dramatic, such threats may be significant at the interpersonal level, as well. Even
within a given culture, particularly a large heterogeneous culture with highly
differentiated roles, value discrepancies among its members are likely. Although,
for simplicity'S sake, we have referred to the cultural drama upon which self-esteem
is based as if it were the same for everyone who is technically a member of a particular culture, clearly this is not the case. For example, for a professional football
player, performance on the field is highly valued; the world of football is a substantial part of the basis of his possibilities for self-esteem. In contrast, the ability to read
Latin may seem to be a completely worthless skill to him. On the other hand, for a
philosophy professor, understanding Latin may be a source of self-esteem; he or she
may believe that football is a completely inane and absurd activity. Both individuals
may share certain cultural values, such as dedication and integrity, yet their respective views regarding what is important may undermine each other's basis of selfesteem.
Consequently, it is rather unlikely that these two people would like each other,
unless they altered their views. In general, then, our theory implies that, even
within the context of the overall culture, others whose views are dissimilar may
threaten our basis of self-esteem and therefore engender negative reactions. Consistent with this idea, a large body of research has shown that the more dissimilar
Need for Self-Esteem 201
another person is, in terms of important attitudes, beliefs, and values, the less an
individual will like and help the person and the more willing an individual will be
to hurt the person (for reviews, see Byrne, 1971; Rokeach, 1968).
It also has been shown that people try to convert persons with deviant opinions
and, ifthat fails, they reject such persons (Schachter, 1951). Reactions of this type
to those who do not display uniformity, conformity, or obedience follow directly
from our theory. Rejection of deviants occurs even when their deviance seems to be
innocuous. One of the most common examples in U.S. culture is rejection of those
who do not maintain their appearance in accord with cultural prescriptions for a
particular role or situation. If one is working for an accounting firm, one is required
to wear the right type of clothing, and it must be unwrinkled and in good condition. Official dress requirements are much more lax for some occupations, such
as college professors. However, if a new faculty member should be seen conducting his duties wearing shorts and unmatched socks, his or her colleagues are likely
to be bothered by this, and may even request that the person wear more "dignified" apparel.
Such deviant behavior poses no direct, concrete threat to the well-being of the
individuals disturbed by it. The threat exists at a symbolic level. Such behaviors (as
well as less and more extreme acts of deviance) threaten the values underlying the
cultural prescriptions that have been violated. If individuals derive self-esteem from
viewing their occupation as a highly valued cultural role, then someone with the
same occupation who has the appearance of a transient or a "common laborer"
poses a threat to their cultural anxiety-buffer. The threat may also be more general
in that someone who does not maintain an appropriate appearance is, in a small way,
challenging the absolute rightness ofthe cultural way oflife (the hippies, as a group,
seemed to be viewed as such a threat in the United States in the 1960s). Minor
threats of this nature can usually be defused through derogation (e.g., the person can
be labeled oddball, nerd, geek, eccentric, neurotic, etc.). Unfortunately, more serious threats may engender such behaviors as ridicule, efforts at conversion, brainwashing, ostracism, beatings, and killings.
Maintaining a sense of personal value within the cultuml drama. General maintenance. We are proposing that self-esteem is entirely a cultural creation. People
cannot have a sense of self-worth without meeting the requirements of value
prescribed by the cultural drama to which they subscribe. Such prescriptions consist
of both general standards and more specific role expectations (for both occupational
and social positions). The general standards specify certain competencies and moral
attitudes that all members of a given culture need to demonstrate. Other requirements for value, however, differ among members of the culture, depending on their
ascribed and chosen roles. Similarly, Wicklund and Gollwitzer (1982) have proposed and found support for the notion that individuals vary in their requirements
for self-completion (i .e., self-definitional needs) as a function of their chosen selfdefining goals (e.g., to be a musician). From the present perspective, these selfdefining goals are ways to attain and maintain cultural value; in other words, they are
roles from which individuals attempt to secure their self-esteem.
To the extent that self-esteem derives from living up to shared cultural standards
202 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
and role expectations, approval from others signifies that we have met these standards and thus are indeed valuable. As Festinger (1954) suggested, such reliance on
social reality is especially likely to occur when objective sources of information
about one's value are unavailable, as is the case with most culturally valued attributes. Therefore, the affection, attention, and approval that people receive from
others within their culture are very important sources of a sense of personal value.
Thus we suggest that much self-presentational behavior is motivated by a desire to
maximize the favorability of our own self-evaluations; the more approval and the
less disapproval that we receive from others, the easier it is for us to privately
believe that we are valuable individuals.
The human propensities for conformity, uniformity, and obedience, which have
been demonstrated in a variety of contexts, are consistent with this idea, especially
given the negative reactions to deviance discussed earlier (e.g., Asch, 1958; Gergen
& Wishnov, 1965; Jones, Gergen, & Jones, 1963; Milgram, 1974; Stires & Jones,
1969). Other research has shown that individuals respond positively to those who
praise and like them and negatively to those who disapprove of them (e.g.,
Berscheid & Walster, 1978; S. C. Jones, 1973; Kenny & Nasby, 1980; Shrauger,
1975).
Perhaps the most potent form of approval occurs when one is loved; therefore,
mutual love relationships may be primary sources of self-esteem. In such cases,
individual A imbues individual B with great value; if B wants and needs A, then A
can perceive himself or herself as valuable as well. In other words, the more a person
loves his or her partner, the more his or her partner's love adds to his or her sense
of personal value. This may account for the intensity of familial and romantic love
in most, if not all cultures. Interestingly, in Western culture, love is often lauded as
magical, transcendent, and eternal (see Rubin, 1973), thus making it a particularly
suitable basis for minimizing existential terror. As Becker (1973) noted, the one
problem is that one's sense of equanimity then becomes dependent on the romantic
partner. Perhaps this is why people will do almost anything to preserve good relationships; from this perspective, the negative feeling when such a relationship is
threatened, which we call jealousy, is anxiety engendered by the threat of loss of an
important source of self-esteem.
Children also playa major role in maintaining a sense of personal value. As many
theorists have suggested (e.g., Becker, 1962; Rogers, 1959; Sullivan, 1953), the
child's sense of value is derived from the parent's (or parents') love; but offspring
provide parents with a tremendous sense of value as well. Parents can take credit for
the existence of these creatures; therefore any value they perceive in their children
and their behaviors (from being cute to saving the world) imparts value to the parent(s). In addition, parents know they are needed by their children and are almost
constantly reminded of this. Finally, parents can derive a sense of permanent value
and immortality to the extent that they view their children as extension of themselves who can eventually have their own children, ad infinitum. These ideas can
help explain the intensity of parental concern for their children's well-being,
prosperity, and adherence to the parents' cultural drama.
In general, then, individuals can feel valuable to the extent that they feel needed.
Therefore, a person can build and maintain self-esteem by being helpful to others,
Need for Self-Esteem 203
particularly others who have been deemed by the culture to be particularly worthy
of help. Helping imparts a sense of value both because of the approval it generates
from others and because of one's private sense of living up to cultural standards of
goodness.
There are also more tangible bases of value in most cultures. For example, in the
United States a wide variety of material symbols of value exist, most of which indicate one's financial status (e.g., jewelry, clothes, automobiles). For most members
of American culture, material wealth is a prime indicator of success in the cultural
drama. Such a basis of self-esteem may be particularly appealing because it allows
for visible, tangible, and enduring symbols of one's worth. In small and large ways,
Americans may build their own pyramids.
Perhaps individuals can also establish a sense of worth vicariously through identification with real and fictional cultural heroes. By identifying with someone special and important, especially someone who has defied death, one can attain a
feeling of being significant and immortal, even if only momentarily. This may be
particularly likely to occur when a culture is having difficulty providing self-esteem
for its people. Under such circumstances, people may flock to an individual with
heroic qualities and a clear alternative world view to establish or restore a sense of
their own value. This may help explain why certain historical figures gained so many
dedicated followers (e.g., Gandhi, Hitler, Jesus).
Similarly, it may be that the appeal of much entertainment derives from the portrayal of heroism (see Bettelheim, 1977; Campbell, 1968). Heroes in literature,
film, television, and sports also allow us to share the glory. Perhaps one of the
clearest examples of this is the Spanish bullfight, in which the heroic matador, as the
representative of the culture, defies death by "vanquishing" the terrifying bull;
however, this analysis applies equally well to a variety of spectator sports, films, and
works of fiction. Such portrayals of heroism may also help individuals frame their
own lives within a heroic context (e.g., in the United States, one could be the Bob
Hope or Richard Pryor of teachers, the Dr. 1. of shortorder cooks, the John Wayne
of computer troubleshooters, or the Clint Eastwood of politics).
Similarly, participatory forms of entertainment, such as gambling, video games,
board games ("Dungeons and Dragons" may be a rather extreme example of this),
theater groups, and participatory sports provide individuals with contexts in which
they may attain a temporary sense of heroism (i.e., self-esteem). Thus, leisure
activities may be enjoyable primarily because they allow individuals to bolster their
cultural anxiety-buffers.
Each of these modes of attaining self-esteem probably varies with regard to the
strength and durability of the sense of personal value that it confers. An inspiring
movie may provide a couple of hours' worth, bringing one's partner to orgasm a couple of days' worth, and saving someone's life a couple of months' worth. However,
the effects are always transitory, making self-esteem maintenance a complex and
taxing problem. People adapt quickly to whatever they have already attained; therefore, a sense of personal value is a tenuous, day-to-day proposition (see Brickman &
Campbell, 1971). Consequently, people have a virtually constant need for reminders
or new indicators of achievement and being loved. The athlete needs to come
through in the next pressure situation, the millionaire businessman needs to keep
204 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
accumulating money, the Don Juan needs new romantic conquests, and the parents
need continual affection from their offspring.
Defenses against threat. Our theory implies a number of conditions that may
threaten self-esteem and consequently require some form of defense to minimize
anxiety. Clearly, self-esteem will be threatened whenever individuals become aware
that some aspect of themselves may undermine their efforts to fulfill a valued role,
be desired by others, gain social approval, avoid disapproval, or accumulate symbols
of worth. A particular activity will therefore be ego-involving to the extent that it has
potential for providing information concerning attributes relevant to the individual's
ability to be a valued participant in his or her particular cultural drama. Our previous
discussion has shown that, when individuals are anticipating, or have experienced,
failures on tasks, they alter their beliefs, attitudes, and behaviors in a variety of ways
to protect self-esteem.
Aside from specific failures, events that generally heighten self-awareness may
also make potentially threatening shortcomings salient. Recall that Becker proposed
that the capacity for self-awareness is largely responsible for the human potential for
existential terror; it is when we are self-aware that our creatureliness and our isolated existence is salient. Therefore, it is when we are most self-aware that we
should be most concerned with being valued participants in the ongoing cultural
drama. Consistent with this reasoning, a substantial body of research has shown
that, under conditions of heightened self-awareness, individuals are much more
likely to act in accord with salient internalized cultural standards (see Buss, 1980;
Carver & Scheier, 1981; and Wicklund, 1975, for reviews) and also to defend selfesteem from the threat of failure (Hull & Levy, 1979; Kernis, Zuckerman, Cohen,
& Sparafora, 1982).
Our analysis so far implies that, although the basis of self-esteem threat consists of its implications for cultural value, the threat can consist of private awareness of a culturally relevant shortcoming; therefore private failures can threaten
self-esteem and private ways to minimize the threatening implications of such
failures can protect self-esteem. Thus, our theory is consistent with the previously
discussed research demonstrating private efforts to defend self-esteem. However,
the theory can also account for research showing heightened private self-esteem
defense when a shortcoming is known to others; the opinion of others provides a link
to the shared cultural drama from which self-esteem is derived. Being of value
means living up to shared cultural standards; if others who are fellow participants
in, and validators of the cultural drama do not believe we are meeting those standards, our own beliefs that we are doing so are threatened. Self-esteem is therefore
threatened whenever individuals become aware that some component of their value
as perceived by others within the culture falls short of their own perception of that
component. This explains why public awareness of a shortcoming leads to especially
vigorous private attempts to defend self-esteem, even if, as in the Tesser and Paulhus
(1983) study, the individual knows that such public awareness is based on erroneous
information.
It follows, then, that self-esteem concerns are aroused not only in achievement
Need for Self-Esteem 205
settings, but in virtually any social situation. Theorists have often noted that
individuals try to establish and maintain a particular positive social identity in the
public arena (e.g., McCall & Simmons, 1978). Along the same lines, Goffman
(1955) has discussed the need to protect the "sacred self" by maintaining face in all
social encounters. More recently, Wicklund and Gollwitzer (1982) have argued that
public acknowledgment is necessary to anchor one's self-definition in social reality.
Quite similarly, Baumeister (1982) has proposed that one of the two major functions
of self-presentation is to establish and maintain a preferred self-concept (selfconstruction). All of these ideas are quite compatible with our proposition that selfesteem is a sense of value that is culturally created and maintained. Whereas each
of these approaches is unique in some ways, they all explicitly or implicitly
acknowledge that the favorability of social image will directly influence the favorability of private self-image; from our perspective, this is because self-esteem is a
function of perceived cultural esteem, which, in turn, is reflected largely in the
appraisals of others. Thus, self-presentational behavior is not only designed to
garner specific rewards and avoid specific punishments from a particular present
audience-it is also designed to maintain and bolster self-esteem.
This idea is highly consistent with the research on interpersonal attraction and
self-presentation mentioned in the above discussion of general maintenance. It can
also help explain a wide variety of everyday experiences. When a teacher addresses
his or her first class, the teacher's concerns about doing well and the associated anxiety are far greater than would be expected if outcomes from the particular audience
were all that is at stake-clearly a potential threat to self-esteem is involved. When
a lonely man, alone in a bar, refuses to approach attractive women because he is
"afraid of rejection;' clearly what he is avoiding is not merely negative outcomes
from the particular denizens of the bar, but a threat to self-esteem.
In fact, a wide variety of social behaviors, some trivial, some very serious, are
influenced by our needs to protect self-esteem. Toward the trivial end, concerns
about self-esteem may keep people from buying porno magazines or from dancing
in public. Toward the serious end, they may keep someone in dire need of counseling
from seeking such nelp. Finally, an extreme example can be seen in Japanese culture,
where a traditional response when one's public image has been severely undermined
is suicide; an honorable death, which is deemed worthy by the culture, may be
preferable to life without the cultural anxiety-buffer.
In each of these instances, and innumerable others, we can see the powerful
influence self-esteem needs exert on social behavior. Phenomenologically, people in
these situations modify their behaviors to minimize negative feelings that are
described variously as anxiety, embarrassment, guilt, humiliation, nervousness, and
shame. Although a variety of theorists have assumed these reactions occur (e.g.,
Goffman, 1967), and a few have briefly addressed their source (e.g., Berger & Luckmann, 1967), theories directed toward explaining these phenomena are rare. Following Becker (1962, 1973), our theory posits that these feelings are a leakage of the
basic existential terror from which self-esteem protects us. Threats to public image
threaten one's value in the cultural drama, thereby undermining one's basis of
equanimity.
206 Jeff Greenberg, Tom Pyszczynski, and Sheldon Solomon
Conclusion
Summary
We have reviewed the research supporting the existence of a need for self-esteem
and have attempted to account for these [mdings with a theory that explains what
self-esteem is, why we need it, and how the need for self-esteem affects social
behavior. Briefly, we propose that culture reduces the terror engendered by awareness of our vulnerability and mortality by providing a shared symbolic conception
of reality that imputes order, predictability, significance, and permanence to our
lives. This cultural drama provides the possibility of leading a meaningful and
enduring existence; equanimity is attained only when a person believes that she or
he is a valued participant in such a cultural drama. This attitude, which is referred
to as self-esteem, serves the essentially defensive anxiety-buffering function of
imbedding the individual within a transcendent cultural drama. Stated simply, selfesteem gives people a basic sense of security that is needed very badly.
Strengths of the Theory
We have attempted to show how this theory can provide a powerful explanatory
framework for a wide variety of social psychological findings. We also believe that,
because the theory focuses on the relationship between the individual and culture,
it can account for a broad range of phenomena that have not been, or cannot be
addressed by other social psychological conceptions of human behavior. It has been
noted that social psychology has traditionally been an ahistorical and acultural discipline (see Gergen, 1973; McGuire, 1973; Sampson, 1978). Therefore, many
issues concerning past and ongoing human events have been ignored or dismissed as
beyond the bounds oflegitimate psychological discourse and, thus, have been left to
historians, sociologists, anthropologists, political scientists, and economists. In contrast, by exploring the psychological functions of culture, the terror management
theory suggests that historical, cultural, and economic behaviors cannot be understood without considering the psychological needs of the individual. By doing so,
perhaps the theory can aid the slow process of integration of the social sciences
toward a full understanding of human behavior.
Research implications. We also believe that the theory has considerable potential
for generating empirical research. We are currently examining the proposed anxietybuffering property of self-esteem by bolstering or threatening self-esteem in a variety of ways and assessing subsequent affect and behavior in potentially anxietyprovoking situations. Another direction we are taking is assessment of the effects of
heightened salience of creatureliness and mortality on propensities to protect selfesteem and adhere to cultural values. Interestingly, Paulhus and Levitt (in press)
have recently found that subtle exposure to affect-laden words, such as death, coffin,
guts, and blood, led individuals to evaluate themselves in an especially favorable
manner.
We also hope to assess how direct threats to self-esteem may lead people to bolster
such things as cultural values, group identification, and the tendency to like similar
Need for Self-Esteem 207
others and reject dissimilar others. Conversely, we are also considering the possibility that heightened group identification decreases the threat to self-esteem of a particular task outcome, and that threats to group identification or other aspects of
one's cultural drama intensify the need to bolster self-esteem.
Another direction we hope to take is to assess the role of public knowledge in selfesteem threat. For example, our theory implies that performance on a task that is initially viewed by the individual as trivial may become ego-involving if others seem
to value good performance, especially if the others are similar to or valued by the
individual. Similarly, when individuals perform potentially embarrassing acts, the
extent of embarrassment should covary directly with the perceived similarity and
value of the audience. We are also studying the effects of entertainment activities on
self-esteem. We could go on, but the main point is that the theory can generate a variety of testable hypotheses, and therefore, judgments of its validity will ultimately
depend on the outcome of ongoing and future empirical work.
Final Thoughts
One general implication from this theory is that a variety of mental health problems
may result from the individual's inability to maintain a cultural anxiety-buffer,
because of loss of faith either in one's ability to maintain a valued role within one's
cultural drama or in the cultural drama itself. Thus cultures can be evaluated by
examining how well they fulfill the responsibility of providing a compelling conception of reality that allows the greatest number of individuals within the culture to
derive self-esteem, with the least expense to others inside and outside the culture.
If this idea could be kept in mind, along with acknowledgment of the ubiquitous need
for a sense of personal heroism (Becker, 1973), perhaps societies could evolve more
effective and benign ways for all of us to manage our basic terror}
Acknowledgments. Thanks to <PRESIDIO_ANONYMIZED_PERSON>, Nancy Hobbs, Donna Morganstern, Abram
Rosenblatt, John Thibaut, Challenger Vought, and Catherine Wylie for their contributions to this work"
Greenberg, J., & Arndt, J. (2012). Terror management theory. Handbook of Theories of Social Psychology: Volume 1, 398–415. https://doi.org/10.4135/9781446249215.n20 :
"ABSTRACT
Terror management theory was developed to explain
the motivational underpinnings of phenomena such
as self-esteem defense and prejudice. The theory is
rooted in a long tradition of thought regarding
human awareness of death and its role in psychological functioning. The theory posits that to manage
the potential for terror engendered by the awareness of mortality, humans sustain faith in worldviews
which provide a sense that they are significant
beings in an enduring, meaningful world rather than
mere material animals fated only to obliteration
upon death. The theory is supported by a wide range
of studies showing that self-esteem and worldviews
provide protection against anxiety and death-related
cognition, reminders of mortality instigate worldview
bolstering and self-esteem striving, and threats to
the worldview and self-esteem increase the accessibility of death-related thought. The research has also
led to a dual defense model of responses to conscious and unconscious death thoughts. We then
focus on two of many topics informed by the theory:
attitudes and behavior regarding physical health,
and political preferences and intergroup conflict. We
then consider factors that mitigate destructive forms
of terror management. Finally, we briefly summarize
the contribution of terror management work so far
and where it’s heading.
TERROR MANAGEMENT THEORY
Terror management theory was born out of a
fundamental dissatisfaction with the field of
social psychology circa 1980 shared by three
graduate students at the University of Kansas:
Sheldon Solomon, Tom Pyszczynski, and the
senior author of this chapter. Firmly in the
grip of social cognition, the field portrayed
people as dispassionate information processors guided by schemas and heuristics, operating in a historical, cultural, motivational,
and affective vacuum. Raised in working
class families characterized by joy and anger,
sibling rivalry and love, passion and sarcasm,
in neighborhoods centered around churches
and synagogues, bars and ballparks, full of
people driven by regional, ethnic, and occupational pride and conflict, social psychology at the time struck us as describing some
affectless androids the three of us had
never met.
In viewing people as highly motivated
creatures driven by their needs to protect
their self-esteem and assert the superiority of
TERROR MANAGEMENT THEORY 399
their own group over other groups, we were
like young salmon swimming upstream
against the cognitive revolution. We focused
our research on how motivation influenced
people’s perceptions of themselves and
others, particularly in the context of defending self-esteem. We contributed to research
on self-serving biases and self-handicapping,
which showed that people bias their cognitions to protect their self-esteem. But we
completed our graduate studies believing that
we had no idea what drove people toward
their prides and prejudices.
Spread around the country by the whims
of the difficult job market (some things never
change), the three of us began searching outside the discipline’s journals for answers.
I recall getting a call from Sheldon, back in
1982, in which he proclaimed that he had
found a guy with the answers. It was a
deceased cultural anthropologist named
Ernest Becker, and the answers were in his
book, The Denial of Death, which had won
the 1973 Pulitzer Prize for general nonfiction. I quickly read the book. I found it
frightening, brutal, and revelatory. Building
on a wide array of sources, and taking an
existential psychoanalytic perspective, the
book seemed to explain all the human trends
we had observed growing up and which were
so well documented by social psychological
research: conformity, obedience, self-serving
biases, aggression, and prejudice. It explained
everything from the rise of Nazism in
Germany to the complexities of sex.
Sheldon, Tom, and I began discussing the
book’s many insights, as well as those provided by his earlier book, The Birth and
Death of Meaning (1971), and his last book,
Escape from Evil (1975). We began using
Becker’s ideas to teach social psychology in
a more integrated way, as if the creature
being discussed when we covered self-esteem
was the same one we discussed when we
covered social influence, aggression, prejudice, and close relationships. We were invited
by Roy Baumeister to participate in a symposium entitled “Public and Private Self” at the
October 1984 Society of Experimental Social
Psychology (SESP) conference in Snowbird,
Utah, and to contribute a chapter to an associated edited volume. We decided at SESP to
introduce the core of Becker’s analysis to our
colleagues in social psychology, though Tom
couldn’t make it. On a snowy morn in front
of an unlit fireplace in our room at the
Snowbird Ski Lodge, Sheldon and I rather
excitedly hacked out a simple summary of
the theory, which we there dubbed “terror
management theory” (TMT).
The symposium room was moderately full
as Sheldon started his presentation, but as he
began discussing Marx, Kierkegaard, Freud,
and Otto Rank, a good half of the audience
jostled their way out of the room. From the
back of the room, I spotted a few remaining
dignitaries such as John Darley, Ned Jones,
who was visibly shaking his head throughout
the talk, and our graduate school advisor and
one fan, Jack Brehm. At talk’s end, instead of
the high fives we had envisioned, we received
stunned silence, shock, and dismay. I was
comforted at least that the practice of stoning
had gone out of fashion.
Undeterred by the chilly initial reception,
we forged ahead with the chapter, which
became the first written presentation of TMT
(Greenberg et al., 1986). Simultaneously, we
worked on a paper more fully presenting the
theory and explaining its potential value as a
broad explanatory framework. We were
hoping such a paper would be well received
at American Psychologist, a publication
which clamors for broad, integrative ideas.
The paper was rejected unceremoniously
with two reviews; one a single paragraph, the
other a single sentence: “I have no doubt that
this paper would be of no interest to any psychologist, living or dead.”
Socialized into our profession very well,
we did not accept these concise but argument-free bases for rejection. After a year or
so of back and forth, editor Leonard Eron
explained that the ideas may have had some
validity but would not gain currency until
backed by empirical research. Actually, up to
that moment, we had not thought of TMT as
something to study empirically. But it struck
400 HANDBOOK OF THEORIES OF SOCIAL PSYCHOLOGY
us – duh – that this was precisely what we
had been trained to do – to derive testable
hypotheses from theories, and then test
them.
We believe that TMT explains some fundamental things about human beings and
their social behavior. But TMT was not like
your standard variety social psychological
theory. Most theories in social psychology
are mini-theories focused on a specific set of
processes pertinent to a particular topic
within the field: stereotype threat (prejudice);
the elaboration likelihood model (persuasion); the culture of honor (aggression); selfverification theory (self). TMT is about the
role of the unconscious fear of death in just
about everything we humans do. But soon we
were able to derive hypotheses from this
broad existential psychodynamic theory and
strategies for testing them in collaboration
with our graduate students.
Realizing resistance was going to be
strong, we put together six studies, a larger
set than was customary at that time, before
submitting to the Journal of Personality and
Social Psychology. The paper was accepted
(Rosenblatt et al., 1989); the reviews were
along the lines of, “This can’t be right, I
don’t like it, but I can’t explain their findings,
so take it.” We thought at the time, “Fair
enough.”
Since then, the empirical support for TMT
has blossomed into a literature consisting of
over 400 studies and counting, conducted in
16 countries. This body of work, which has
included numerous theoretical expansions
and refinements, has benefited greatly from
the second (and now third) generation of
TMT researchers, former students of the
original trio including the junior author of
this chapter. In the last decade, researchers
from independent labs throughout the world
have also made valuable contributions to this
continually expanding literature. We believe
that the growth of this area of study reflects
the generative value of a broad theory that
integrates widely varied human endeavors by
exploring deeply rooted forces that drives
humans to behave the way they do. Indeed,
the junior author of this chapter became
enthralled as an undergraduate at Skidmore
College with the broad existential perspective TMT offered and the promise of subjecting such ideas to empirical scrutiny. In due
course, we will provide a brief overview of
the literature that has developed over the past
20 years with an emphasis on recent directions pertinent to contemporary concerns.
But first we should step back to acknowledge
the rather extensive roots of the theory.
THE DISTAL AND PROXIMAL ROOTS
OF TMT
Although met with surprise and skepticism
when introduced to social psychologists in
the 1980s, a decent case can be made that
TMT is an ancient theory that can be traced
back to one of the first narrative texts
from around 3000 BC, the Sumerian Epic of
Gilgamesh. A central theme of this story,
which influenced the major religions that
later emerged out of the Middle East, is the
main character’s deep concerns about death
and his consequent search for immortality.
Gilgamesh, devastated by the death of his
friend Enkidu, realizes that he too will die.
He roamed the desert, lamenting: “How can
I rest, how can I be at peace? Despair is in
my heart. What my brother is now, that shall
I be when I am dead … I am afraid of
death…” He then embarked on a journey to
find immortality.
From that early time on, the notion that we
humans fear death and urgently desire to
deny or transcend it in some way has been a
prominent theme in literature, religious writings, and philosophy. Indeed, Schopenhauer
declared death the muse of philosophy; we
couldn’t even begin to review its role in
philosophical thought here. But we should
note that the first person to put the basic
points of TMT together seems to have been
the famed Greek historian Thucydides.
Around 400 BC, Thucydides was focused
on understanding the vicious intergroup
TERROR MANAGEMENT THEORY 401
conflicts that plagued Ancient Greece. He
proposed that the fear of inevitable death
drove people to seek immortality in three
ways: through heroic, noble actions restoring
justice, which qualified them for a divinely
awarded afterlife; through memories of their
heroic deeds; and through identification with
death transcending group identifications. As
Ahrensdorf (2000: 591) put it: “Thucydides
contends that they will seek to escape by
somehow overcoming their mortal condition,
by living on after their death – either through
their city, or through their glory, or in an
afterlife – and by winning the gods’ favor
through the vehement affirmation of their
own nobility or piety, or justice.” Thucydides
also noted that the increasing salience of
mortality once conflicts begin intensifies the
desire to heroically vanquish the enemy.
Jumping forward about 2000 years to the
modern English literary tradition, poets from
Shakespeare to Wordsworth, Keats, and
Shelley, to Dickinson and Emerson, among
hundreds of others, have acknowledged the
role of the fear of death and desire to escape
it in the human psyche. Similarly, novelists
such as Swift, Dickens, Dostoevsky, and
Tolstoy down to twentieth-century writers
such as James Baldwin, Don Delillo, James
Joyce, Phillip Roth, Milan Kundera, and Kurt
Vonnegut have explored how the fear of
death drives various forms of human behavior. Here, Baldwin concisely captures the gist
of TMT:
Life is tragic simply because the earth turns and
the sun inexorably rises and sets, and one day, for
each of us, the sun will go down for the last, last
time. Perhaps the whole root of our trouble, the
human trouble, is that we will sacrifice all the
beauty of our lives, will imprison ourselves in
totems, taboos, crosses, blood sacrifices, steeples,
mosques, races, armies, flags, nations, in order to
deny the fact of death, which is the only fact we
have (James Baldwin, The Fire Next Time, 1963).
Awareness of the role of death in the human
psyche has not been limited to the great philosophers, poets, and novelists, but can also
be found extensively in visual art (e.g., Van
Gogh, Klimt), music (e.g., Schubert and
Mahler), and film (e.g., Woody Allen and
Ingmar Bergman). And it is not limited to
representatives of “high culture;” indeed, it
seems increasingly prevalent in contemporary pop culture also. As we’ve been writing
this chapter in early 2009, two films, David
Fincher’s Fitzgerald-inspired The Curious
Case of Benjamin Button, and Frank Miller’s
comic-book-inspired The Spirit, as well as
the second 2009 episode of the sitcom Scrubs
have focused squarely on the psychological
problem of mortality. Recognition of this
problem in so many of the products of
Western culture makes it all the more interesting that in the 1980s social psychology
texts made no mention of death and TMT
was treated by most people in the field as
outlandish and irrelevant to understanding
human social behavior.
As noted earlier, TMT was formalized out
of three of Becker’s books, The Birth and
Death of Meaning, The Denial of Death, and
Escape from Evil. These books combine
insights from anthropology, evolutionary
biology, philosophy, psychoanalysis, and
sociology. We urge anyone interested in
understanding human behavior to read them.
Around the time Becker was putting his
synthesis together, evolutionary philosopher
Susanne Langer and psychoanalytic historian
Robert Jay Lifton were coming to the same
view, and a bit later in 1980, so did existential
psychotherapist Irvin Yalom. The primary
influences on Becker’s analysis were
Kierkegaard, William James, Freud, Gregory
Zilboorg, Erving Goffman, Norman Brown,
and especially Otto Rank. Within psychology, Rank, Freud’s disciple and an impressive interdisciplinary scholar back in the
1930s, first granted the fear of death and the
urge toward immortality a central role in
human culture and social behavior. By systematizing these existential psychodynamic
ideas into a coherent explanatory framework,
TMT integrates much of what we know about
human behavior and provides a fertile basis
of generating a wealth of testable hypotheses.
The TMT research in turn has inspired useful
expansions and refinements of the theory.
402 HANDBOOK OF THEORIES OF SOCIAL PSYCHOLOGY
THE BASICS OF TMT
TMT starts with two simple observations.
The first is that humans are animals with
many systems designed to keep them alive,
including a fight-or-flight response to
impending threats to their continued existence. The second is that the cognitive abilities of humans have led to the awareness that
death is inevitable and could come at any
time due to a host of potential causes. The
theory posits that this knowledge, in an
animal built to avert death, creates an everpresent potential for intense anxiety, or terror,
which must be managed continuously. The
terror is managed by sustaining faith in a
view of the world and oneself that denies the
precarious and transient nature of one’s own
existence.
Cultural worldviews have been shaped to
serve this function from ancient times to now.
Presumably there was a point in time when
the cerebral cortex of our species became sufficiently developed to provide self-awareness
and the ability to think in terms of past,
present, and future. These largely adaptive
cognitive capacities also led to the awareness
of mortality. Although fear in response to
imminent threat is often adaptive, constant
anxiety regarding the fragility of continued
existence and its inevitable end would not be.
At this point, our ancestors constructed
and gravitated toward shared conceptions of
reality that most effectively quelled the
potential fear that knowledge of our vulnerability and mortality engendered. These cultural worldviews imbue external reality with
order, stability, meaning, and purpose, and
offer ways in which people can believe they
will endure beyond death either literally
through an everlasting soul, symbolically
through a death transcending identity, or for
most worldviews, in both ways.
At the most basic level, all cultural worldviews allow people to live out their lives
largely within a conception of reality in
which they view themselves as symbolic or
spiritual beings who exist in a meaningful
world, rather than as mere transient animals
fated only to obliteration upon death. Indeed,
the content of consciousness is structured by
the cultural worldview in which a given individual is raised. We think in terms of names,
dates, months, days of the week, hours, minutes, social roles, and categories. But these
are ultimately elaborate window dressing
that lends artificial and largely arbitrary
structure to an ongoing subjective experience
of unique sensations and perceptions that
begins at birth and ends at death. Questions
like “Who are you?” “Where are you?”
“What time is it?” can only be answered with
reference to culturally created constructs.
How do people become imbedded in a
cultural worldview that provides them with
their fundamental psychological security?
Developmentally, the human newborn is the
most helpless and dependent of all living
creatures. It is also a highly distressed creature because, as Rank (1932 /1989) noted, it
was conceived and developed attached to its
mother and was suddenly, violently separated
from its warm home, the womb, at birth.
From exposure to the first words that flow
out of their parents’ mouths, the infant begins
to be socialized into the prevailing cultural
worldview. In early childhood, the only basis
of psychological security for these helpless
creatures is the care and love of the parents.
The parents provide knowledge, comfort,
nutrition, and protection.
As the parents begin to require particular
actions from the child to sustain their love
and approval, the child learns that when she
does the right things, all is good in the world
but when she does the wrong things, all hell
could break lose. To sustain the psychological security of the parental love, the child
internalizes the parental values of good and
bad, and tries to live up to the good, despite
the infringement on the child’s natural desires
and urges this entails. For the good little fiveyear-old sitting on mom’s lap, all is right in
the world. For the bad five-year-old who just
punctured one of dad’s new stereo speakers
with a flying action figure, the world has just
become a scary place. In this way, the child
comes to associate being good and valued
TERROR MANAGEMENT THEORY 403
with feeling secure, and being bad and
worthless with dread. Self-esteem, that feeling of being valuable, thus comes to serve an
anxiety-buffering function.
The plot thickens as the child’s cognitive
development leads her to realize that the
things she fears – the dark, monsters, ghosts,
big dogs – threaten her very existence and
that the parents are limited and can’t always
protect them from everything. Concomitant
with these dawning realizations, the child’s
basis of security gradually shifts to larger
more powerful sources, deities, presidents,
and the culture in general. This transference
process is facilitated by the fact that the parents have been inculcating the child into the
larger worldview, including the parent’s own
bases of psychological security, throughout
the child’s upbringing.
So over childhood, psychological security
becomes predicated on being a good
Christian, American, and so forth, on being
valuable within the context of the individual’s internalized version of the cultural worldview. This sense of enduring significance is
not only linked implicitly with safety and
security but explicitly with the literal immortality of heaven and the symbolic immortality of being part of death-transcending entities
such as the family line and the nation, and
lasting cultural achievements in science, politics, and the arts. In these ways, the significant person in a society can feel like he or she
is an eternal part of an enduring reality.
Summary of TMT and its basic
implications
In sum, the simple formula for effective
terror management is faith in a meaning providing cultural worldview and the belief that
one is a valuable contributor to that meaningful world (the TMT conceptualization of
self-esteem). We initially derived two basic
implications from the theory (Solomon et al.,
1991). First, self-esteem and the worldview
upon which it is predicated serve a critical
anxiety-buffering function. So people strive
for and defend self-esteem in the service of
psychological security and these efforts
depend on the specific cultural worldview to
which the individual subscribes. Second,
because these constructs are ultimately fragile social constructions, people will react
negatively to anyone or anything that undermines faith in their worldview or their selfworth. We believe that this provides a very
basic insight into prejudice and intergroup
conflict. People who criticize one’s own
worldview, or simply adhere to a worldview
very different from one’s own, call into question the validity of one’s own basis of psychological security.
TMT therefore posited that such different
people are inherently threatening and are
reacted to with four defenses. The first, most
widespread, is derogation. If these others are
ignorant or evil, then their alternative beliefs
can be dismissed. The second is assimilation.
If these people are wrong, then I can help
them see the light, which will make me all the
more certain my worldview is the right one.
Missionary activity provides a vivid example
of this strategy. The third is accommodation,
which involves incorporating appealing
aspects of the alternative worldview into
one’s own. In this way, one can preserve faith
in the central components of one’s own
worldview. Rock and rap music initially
expressing anger and challenging mainstream
American culture lose their threatening nature
when converted to muzak or used to sell fast
food burgers. Finally there is annihilation. If
one’s worldview is right and the others’ is
wrong, then one’s own should prevail.
The core tests of hypotheses
derived from TMT
TMT is consistent with a wide range of evidence from anthropology, archeology, and history, but our challenge was to deduce predictions
that put the theory to the test. The central TMT
hypotheses are based on its account of the psychological function of cultural worldviews
and self-esteem. If these structures provide
404 HANDBOOK OF THEORIES OF SOCIAL PSYCHOLOGY
protection against death-related fears, then: (a)
reminding people of their mortality should
increase bolstering of the worldview and striving for self-worth; (b) bolstering these structures should reduce anxiety in response to
threat and defensive reactions to reminders of
death; and (c) threatening the structures should
arouse anxiety and bring death-related concerns
closer to consciousness.
Hundreds of studies have supported these
hypotheses. First, let’s consider hypothesis
(a). Reminding people of their mortality (mortality salience or MS) increases positive reactions to people who validate aspects of
participant’s worldviews and negative reactions to people who challenge aspects of their
worldview or espouse a different one (see
Greenberg et al., 2008 for a review). For
example, MS leads to negative evaluations of
Jews by Christians, of a critic of the US by
Americans, and of foreign products by
Germans. MS even leads to increased aggression against a critic of one’s preferred political
orientation, whether liberal or conservative.
On the other hand, MS increases positive reactions to heroes and celebrities, one’s national
soccer team, and members of one’s own religious group, and leads to more donations to
valued charities. MS also increases striving
for self-worth (see Greenberg et al., 2008). In
people who value such things, MS increases
bold driving, displays of physical strength,
proenvironmental intentions, focus on one’s
appearance, interest in self-esteem enhancing
dating partners, and ratings of the importance
of and desire for fame and wealth.
A substantial number of studies have also
supported hypothesis (b). Bolstering selfesteem reduces anxiety in response to threat
(Greenberg et al., 1992b) and defensive reactions and death-thought accessibility following reminders of mortality (Harmon-Jones
et al., 1997). Similarly, bolstering or defending one’s worldview also reduces defensive
reactions and death-thought accessibility
following reminders of mortality (e.g., Arndt
et al., 1997).
Recent research has also supported
Hypothesis (c). Reminders that humans are
animals, criticism of participants’ worldviews, and threats to participants’ self-esteem
all increase the accessibility of death-related
thoughts and not the accessibility of other
negative thoughts (e.g., Friedman and Rholes,
2007; Schimel et al., 2007).
This body of research has operationalized
reminding people of their mortality in a variety of ways and has extensively compared
reminders of mortality with reminders of
other negative thoughts (see Greenberg et al.,
2008). People have been reminded of death
via two simple questionnaire items, writing
one sentence about death, fear of death
scales, proximity to cemeteries and funeral
homes, and subliminal death primes. Thinking
about mortality has been compared to thinking about being paralyzed, failing, feeling
uncertain, general anxieties, giving a speech
in public, intense pain, dental pain, disease,
unpredictable, intense bouts of pain, life
being meaningless, being socially excluded,
upcoming exams, unexpected events, and
worries after college.
Although some studies have found similar
effects of other aversive thoughts in circumscribed contexts (e.g., McGregor et al., 2001;
van den Bos, 2001) the vast majority have
found quite different effects for reminders of
mortality. TMT theorists (e.g., Greenberg
et al., 2008) have argued that this is because
death is the only inevitable future event, it is
what human biological systems, including
the fight or flight system, are most focused
on avoiding, and it threatens to undermine all
human desires, whether for belonging, cognition, control, or growth. One possibility is
that when these other threats lead to defenses
similar to those instigated by MS, they may
undermine structures that serve terror management, thereby bringing thoughts of death
closer to focal awareness. For example,
Landau et al. (2004a) found that threatening
individuals’ belief in a just world led to elevated death-thought accessibility. These other
aversive experiences may also, under some
conditions, be threatening in their own right.
Given the clear evidence that thoughts of
mortality so often arouse different responses
TERROR MANAGEMENT THEORY 405
than these other aversive cognitions, an
important direction for future inquiry research
is understanding when psychological defenses
are serving terror management and when
they are serving other concerns.
A theoretical refinement: the dual
process model
After the Rosenblatt et al. (1989) paper was
published, a German psychologist, Randolph
Ochsmann, reported having difficulty replicating MS effects. In contrast to two brief
items, Ochsmann was reminding people of
death by leading them through an extensive
guided imagery exercise of imagining their
own death and interment. This discrepancy
led to an important insight into the processes
instigated by MS. Realizing that the initial
terror management studies invariably
included some kind of intervening experience between the MS manipulation and
assessment of the dependent measures (e.g.,
mood scales, experimenter instructions) we
posited that terror management defenses do
not occur when death is prominent in consciousness but rather when death is highly
accessible but no longer in focal attention.
Studies (see Pyszczynski et al., 1999 for a
review) initially supported this idea by showing, for example, that MS only leads to worldview defense if there is a non-death-related
task between the MS induction and the dependent measure. They also showed that immediately after an MS induction, the accessibility of
death-related thought is low. After a delay, in
contrast, death-related thought is no longer in
focal attention but becomes high in accessibility. Given people’s reluctance to consciously
dwell on their own mortality, we hypothesized
that, confronted with such thoughts, people
initially suppress them. A series of studies supported this reasoning, finding for example that
MS led to immediate increases in deaththought accessibility if participants were placed
under high cognitive load. Another series of
studies further clarified that role of consciousness in MS-induced worldview defenses by
showing that worldview defenses emerge in
response to unconscious thoughts of death
(e.g., when people are exposed to subliminal
reminders of mortality).
A dual process model of terror management
emerged from this line of research. Explicit
thoughts of death instigate direct proximal
defenses to remove death-related thoughts
from current focal attention. Such “pseudorational” mechanisms make death seem like a
far-off problem, thereby allowing the individual to stop thinking about it. However, after
thoughts of death have been removed from
focal attention, death-thought accessibility
increases, heightening the potential to experience death-related anxiety. This in turn instigates symbolic distal defenses such as bolstering
one’s worldview or self-worth. These terror
management defenses then bring death thought
accessibility down to baseline levels.
Summary
TMT was originally developed to explain
two facts of human experience, namely that
people have difficulty getting along with
those who are different and that people have
a trenchant need to feel good about themselves. Research shows that both these propensities are strands in the fabric by which
people insulate themselves from deeply
rooted fears of mortality. Further, studies
have supported a dual process model of the
defenses instigated by reminders of mortality. Based on this foundation, the theory has
guided investigation of a variety of aspects of
human experience, many more so than were
envisioned on that snowy day in Utah.
THE GRAVE MATTERS: APPLYING
TMT TO CONTEMPORARY HUMAN
CONCERNS
TMT is relevant to a wide range of domains
of human behavior. Although we have been
accused on occasion of trying to explain
406 HANDBOOK OF THEORIES OF SOCIAL PSYCHOLOGY
everything with TMT, our position is not that
everything stems from the need for terror
management, but that most, if not all, significant domains of human behavior are influenced by terror management concerns.
Indeed, the theory has provided insights into
a wide range of phenomena, including human
ambivalence about sex (e.g., Goldenberg
et al., 2000), reactions to the handicapped
(e.g., Hirschberger, 2006), the operation of
stereotypes (e.g., Schimel et al., 1999), academic performance (Landau et al., 2009),
altruism (e.g., Jonas et al., 2002), parenthood
(e.g., Wisman and Goldenberg, 2005), attachment and close relationships (e.g., Mikulincer
et al., 2003), stigmatization (e.g., Salzman,
2001), reactions to nature (e.g., Koole
and van den Berg, 2005), attitudes toward
women (e.g., Landau et al., 2006a), religion
(Greenberg et al., in press), art (e.g., Landau
et al., 2006b), film (Sullivan et al., 2010),
and even human–android relations (see e.g.,
MacDorman, 2005).
The theory has also generated some useful
insights, following Kurt Lewin, into affairs
of particular practical significance, including
legal affairs (Arndt et al., 2005a), consumer
behavior (Arndt et al., 2004), and mental
health (Arndt et al., 2005b). Given space
limitations, we will focus on two heavily
researched applications of TMT. The first
concerns the decisions that people make with
regard to the health, and the second concerns
the political sphere and our seemingly escalating proclivities for intergroup violence.
We will then conclude by considering factors
that mitigate the more harmful defensive
responses provoked by concerns about
mortality.
Understanding the role of
awareness of death in everyday
health decisions
It seems rather obvious that death does not
bode well for an individual’s health. Despite
this fact, theory and research had largely
ignored the deeper psychological significance
of concerns about mortality for understanding the decisions people make regarding their
physical health. This is surprising given
that many health campaigns remind individuals that noncompliance with health recommendations can hasten their demise. What
are the possible consequences of such
reminders? Goldenberg and Arndt (2008)
recently extended TMT with a terror management health model (TMHM) to address
such questions.
The TMHM starts with the idea that healthrelevant scenarios have varying potential to
activate death-related cognition, and then
builds from the proximal and distal terror
management defenses previously described.
When a health situation leads people to
explicitly think about death, or thoughts of
death are otherwise in focal attention, health
decisions are influenced by a proximal motivation to jettison the threatening cognitions
from focal attention. For example, an individual may take proactive (e.g., exercise
more) or avoidant (e.g., deny perceived risk)
steps to render death less of an immediate
concern. When thoughts of death are activated implicitly (i.e., outside of focal attention), health-relevant decisions are driven
more by a desire to affirm the symbolic value
of the self; for example, by pursuing the
standards upon which an individual’s selfesteem is based, investing in worldview
beliefs, or distancing oneself from the creatureliness of the body. Responses to both
conscious and unconscious death-related
thought can be beneficial or harmful to
health, but responses to unconscious death
thoughts will be moderated by the individual’s worldview and bases of self-worth,
whereas responses to conscious death
thoughts will be moderated by the perceived
efficacy of efforts directed toward reducing
the threat of death (see Goldenberg and
Arndt, 2008, for a review of the supportive
research).
In one study illustrating this basic distinction, participants were reminded of their
mortality (or not) and then immediately
thereafter or after a delay answered questions
TERROR MANAGEMENT THEORY 407
about their intentions to tan. Immediately
after thinking about death, participants
reported lower intentions to suntan, suggesting efforts to reduce vulnerability and protect
health. However, after a delay (when thoughts
of death were accessible but not conscious)
participants increased their intentions to
suntan, suggesting an effort to bolster esteem
by enhancing attractiveness as defined in
contemporary society. Thus, conscious
thoughts of death led to a health over beauty
decision, but unconscious thoughts of death
led to a beauty over health decision.
In drawing attention to these distinctions,
TMHM helps explain a number of consequences of fear appeals, such as a so-called
“boomerang” effect whereby fear appeals
(which often highlight the threat of death)
can have consequences antithetical to those
intended. Whereas the initial response to a
fear appeal may reflect efforts to protect
one’s physical self (which then remove death
from consciousness), the delayed effect of
unconscious thoughts of death may instigate
self-esteem striving that entails more risky
behavior.
The TMHM focuses on three general
propositions. The first is that people’s health
decisions will reflect efforts to remove conscious thoughts of death from focal attention.
Accordingly, research has found that when
confronted with explicit thoughts of death,
people react by trying to avoid the threat or
by trying to reduce its likelihood with health
promoting behaviors (such as increasing
exercise). Of course, the critical question is
when and for whom a threat avoidance or
health promoting response is most likely to
emerge. When an individual perceives a
health response as effective, maintains optimism, approaches health situations with
active coping strategies, or construes vulnerability as low, they respond to conscious
thoughts of death with health-oriented rather
than threat-avoidance strategies, particularly
in death-related health domains (e.g., screenings for breast and skin cancer).
The second direction of research has
explored how the unconscious resonance of
death-related cognition promotes maintaining, not one’s health, but a sense of meaning
and self-esteem. For example, thoughts of
death outside of focal attention increase both
health-defeating (e.g., intentions to suntan)
and health-facilitating (e.g., exercising) outcomes depending on situational and dispositional self-esteem contingencies (see
Goldenberg and Arndt, 2008). As this suggests, MS can lead to seemingly counterintuitive risky effects on health-relevant
outcomes, both for self and others. In one
study, after being reminded of mortality,
Christian medical students were more likely
to adopt a cautious triage approach to
Christian patients complaining of chest pain
but adopted a more cavalier approach to
Muslim patients reporting identical symptoms (Arndt et al., 2009). This suggests that
even people who are more routinely exposed
to issues concerning death may still be vulnerable to terror management effects. But
why would individuals make such decisions?
Such responses help to bolster the symbolic
buffers that offer protection from deeply
rooted existential fear. Consistent with this
analysis, Vess et al. (2009) found that not
only did religious fundamentalists respond to
reminders of death by endorsing faith-based
over medical-based treatments for illness
(both for themselves and others), but doing
so helped to satiate their need for meaning in
life.
This does not imply, however, that unconscious thoughts of death necessarily increase
health risk behavior. If individuals derive
self-esteem from behaviors conducive to
health, they display health-facilitating outcomes across different health domains when
death thought accessibility is high. Moreover,
in a number of areas (e.g., exercise, smoking
cessation, tanning, breast cancer screening),
when self-esteem contingencies are experimentally altered in health-conducive ways,
people respond to increased death thought
accessibility with more protective intentions
and behavior. For example, shifting normative appearance standards toward pale skin
led actual beachgoers, when distracted from
408 HANDBOOK OF THEORIES OF SOCIAL PSYCHOLOGY
activated thoughts of death, to request higher
SPF sun lotions when offered compensatory
samples, especially among those who derive
self-esteem from extrinsically defined
standards (e.g., Arndt et al., 2009). Indeed,
following MS, showing those who smoke to
fit in with others a commercial about how
smoking reduces one’s popularity leads to
increased smoking cessation intentions.
The third direction of TMHM research
builds from Goldenberg and colleagues’
(2000) work on how reminders of our physical, creaturely nature intensify terror management concerns. Confrontations with the
physical body threaten the illusion of symbolic significance, and thus our psychological security. Research has shown both
reminders of death and of the creaturely
aspects of the body influence health-related
attitudes and behavior. For example, they
increase negative reactions to breastfeeding
and pregnant women, increase perceptions of
discomfort regarding mammograms, and
reduce the thoroughness of breast self-exams.
In sum, TMHM research has shown that the
awareness of mortality plays a significant
role in attitudes and behavior pertinent to
cancer detection and prevention as well as to
more general healthful living.
TMT, politics, and intergroup
conflict
The awareness of mortality also contributes
to intergroup strife, in two primary ways.
First, as Becker (1971) noted, the mere presence of groups with worldviews very different from one’s own threatens individuals’
terror-assuaging faith in their worldview. We
have reviewed research showing that reminders of mortality lead to derogation and even
aggression against those who criticize the
participant’s worldview or simply subscribe
to a different one.
However, in a chapter called “the Nexus of
Unfreedom” from The Denial of Death,
Becker (1973) elucidated an even more
destructive consequence of the fear of death.
He argued that the most blood has been
spilled not by those with evil motives but by
those who are serving the great leader, God,
and country. Our mortality requires us to find
something bigger than ourselves for our salvation. When that individual or institution
designates some other group as evil or that
group is perceived as a threat to the terrorassuaging leader, deity, or entity, violent
attempts to eradicate that evil threat commonly result.
In his final book, Becker (1975) also noted
that no matter what our worldview is, residual death anxiety will exist and people will
seek controllable sources of that anxiety to
mask its true cause – scapegoats. So the
greatest sense of death transcendence is provided by participating in a heroic triumph
over evil. In other words, concerns about
mortality will draw people toward worldviews and leaders who provide them most
compellingly with a sense that they are part
of something great and have a mission to
heroically triumph over those who are evil;
rigid worldviews and charismatic leaders that
clearly delineate who is good and who is evil
will do this best. Becker used this analysis to
explain something that otherwise seems inexplicable: the rise of Nazism in Germany culminating in the Holocaust (as well as other
similar historical phenomena). Indeed, much
of the course of human history seems to have
been guided by efforts to heroically triumph
over evil in the service of the deathtranscending ideologies and leaders in which
people have invested.
TMT research has supported these ideas in
a number of ways. Reminders of mortality
draw people toward charismatic leaders and
ideologies that sell a simple vision of the
greatness of the ingroup and the need to rid
the world of evildoers. Using descriptions of
hypothetical gubernatorial candidates, Cohen
et al. (2004) found that MS increased the
appeal of a candidate with a charismatic
style; that is, one who exuded self-confidence
and emphasized the greatness of the state and
nation. As the 2004 American presidential
election approached, Americans were faced
TERROR MANAGEMENT THEORY 409
with a choice that would allow a unique test
of the TMT analysis of attraction to leaders.
One was George W. Bush, a folksy, at-ease
incumbent president who emphasized the
greatness of America and the need to rid the
world of evildoers, such as the “Axis of
Evil.” The other was John Kerry, a Democratic
challenger with a stiff speaking style who
had a complex, sometimes hard to decipher
view of issues and who was portrayed by the
Republican spin machine as a waffling flipflopper. Landau et al. (2004b) posited that
MS, as well as reminders of terrorism by
virtue of activating fears of death, would
increase the appeal of Bush and reduce the
appeal over Kerry. And in a series of studies
prior to the 2004 election, that’s exactly what
they found. Subsequently, with the help of a
Bin Laden tape released days before the election, Bush won re-election even though he
had invaded a country and embroiled the US
in a protracted war based on false claims.
The Cohen study suggests that the Bush
effect resulted from Bush’s charismatic style
and simple good-versus-evil worldview.
However, Bush could have also had an
advantage for terror management purposes as
the current leader of the nation. In addition,
Jost and colleagues (2003) have suggested
that conservative rightwing ideologies may
serve terror management better than liberal
or leftwing ones. However, in the Landau et
al. (2004b) studies, despite increasing the
preference for Bush, MS did not increase
participants’ self-reported political conservatism. In addition, a recent study by Kosloff et
al. (2010) building on Cohen et al. found that
MS leads people to prefer a charismatic
hypothetical leader only if that leader also
espouses policies that match the individual’s
pre-existing political orientation, whether
conservative or liberal.
So MS seems to move people toward those
who espouse more straightforward goodversus-evil ideologies as long as those ideologies fit with the individual’s pre-existing
worldview. Based on these findings,
Pyszczynski et al. (2006) wondered if MS
would also increase the appeal of engaging in
violence against those designated by one’s
culture as evil. In their first study, they
exposed Iranian college students to MS or a
control topic, and then asked them to react to
a fellow student who espoused e
[END PART 1/2]
Remember not answering yet. Just acknowledge you received this part with the message "Part 1/2 received" and wait for the next part.