A Contemporary Learning Theory Perspective on the
Etiology of Anxiety Disorders
It’s Not What You Thought It Was
Northwestern University and The Family Institute at
The authors describe how contemporary learning theory
and research provide the basis for perspectives on the
etiology and maintenance of anxiety disorders that capture
the complexity associated with individual differences in the
development and course of these disorders. These insights
from modern research on learning overcome the shortcom-
ings of earlier overly simplistic behavioral approaches,
which sometimes have been justifiably criticized. The au-
thors show how considerations of early learning histories
and temperamental vulnerabilities affect the short- and
long-term outcomes of experiences with stressful events.
They also demonstrate how contextual variables during
and following stressful learning events affect the course of
anxiety disorder symptoms once they develop. This range
of variables can lead to a rich and nuanced understanding
of the etiology and course of anxiety disorders.
lives, making anxiety disorders the most common category
of diagnoses in the fourth edition of the Diagnostic and
Statistical Manual of Mental Disorders (DSM–IV; Ameri-
can Psychiatric Association, 1994; see Kessler, Berglund,
Demler, Jin, & Walters, 2005). Most of us might intuitively
guess that anxiety disorders would typically develop during
or following a frightening or traumatic event or during a
period of significant stress when many of us experience
some anxiety. Yet it is also obvious that not everyone
undergoing traumas or highly stressful periods develops an
anxiety disorder. Any good etiological theory must be able
to account for this and many other apparent mysteries
involved in who does and who does not develop an anxiety
disorder. Consider the following examples, which illustrate
the kinds of issues we address in this article:
Keywords: anxiety disorders, etiology, maintenance, learn-
pproximately 29% of the U.S. population is esti-
mated to have or to have had one or more diag-
nosable anxiety disorders at some point in their
Emily and Marian both had had traumatic experiences with dogs.
Emily was hiking with her own dog when another dog attacked
her and bit her on the wrist. She was terrified. The wound became
badly infected and very painful, requiring medical treatment.
Marian was walking in the fields when she became terrified by
three large, growling dogs that chased her to a fence. One began
tearing at her pant legs, but their owner fortunately intervened before
she was physically injured. Why did Marian but not Emily go on to
develop dog phobia when only Emily was actually bitten by a dog?1
Ahmet and Hasan were both male Turkish citizens in their 30s
who were arrested, imprisoned, and tortured. Ahmet was impris-
oned for several years and experienced a great deal of torture;
Hasan was imprisoned for several weeks and experienced far less
torture (M. Bas ¸og ˇlu, personal communication, 1997). Why did
Hasan but not Ahmet develop posttraumatic stress disorder? What
accounts for such individual differences?
Although behavioral/learning approaches were the
dominant empirical perspective on anxiety disorders
from the 1920s until the 1970s, these approaches have
been widely criticized since the early 1970s (for reviews,
see Mineka, 1985; Rachman, 1978, 1990). Many of these
criticisms stemmed from the inability of early learning
approaches to account for the diverse factors involved in
the origins of people’s anxieties, such as the individual
differences illustrated in the cases cited above. More recently,
however, a resurgence of interest in learning approaches has
occurred, as these approaches have incorporated some of the
complexity predicted by contemporary learning theory and
research (e.g., Barlow, 1988, 2002; Craske, 1999).
In this article we discuss how prior criticisms of
earlier learning approaches are addressed by contemporary
approaches grounded in modern research on learning (e.g.,
see Rescorla, 1988, for one classic review).2Unfortunately,
however, these advances have not been adequately com-
Susan Mineka, Department of Psychology, Northwestern University;
Richard Zinbarg, Department of Psychology, Northwestern University
and The Family Institute at Northwestern University, Evanston, Illinois.
Correspondence concerning this article should be addressed to Susan
Mineka, Department of Psychology, Northwestern University, 2029 Sheridan
Road, Evanston, IL 60208-2710: E-mail: email@example.com
1Unless otherwise noted, all cases presented in this article are slight
adaptations of cases that the authors either have been acquainted with
personally or have become familiar with during supervised treatment of
these individuals by student therapists.
2The reader may note that the subtitle of our article (“It’s Not What
You Thought It Was”) is similar to the subtitle used in Rescorla’s (1988)
American Psychologist article.
10 January 2006 ● American Psychologist
Copyright 2006 by the American Psychological Association 0003-066X/06/$12.00
Vol. 61, No. 1, 10–26DOI: 10.1037/0003-066X.61.1.10
municated to clinical scientists who advance the theories of
etiology and maintenance. Therefore, we attempt to ad-
vance such communication by describing how certain in-
sights from this area of research overcome the shortcom-
ings of earlier, overly simplistic behavioral approaches. We
show how early learning histories serve as vulnerability (or
invulnerability) factors that can dramatically affect the
emotional consequences of traumatic and stressful life
events often implicated in the origins of anxiety disorders.
Such early learning histories, when considered together
with temperamental vulnerabilities, can serve as diatheses
that make certain individuals more susceptible to adverse
and stressful experiences that sometimes lead to the devel-
opment of anxiety disorders. Consideration of these diathe-
ses, as well as of various contextual variables during and
following traumatic learning events, can lead to a rich and
nuanced understanding of the etiology and course of anx-
iety disorders (Mineka & Zinbarg, 1996).
Overall, then, we believe that a contemporary learning
theory approach can account for such complexity at least as
well as the other two prominent contemporary psycholog-
ical approaches to anxiety disorders—that is, the cognitive
(e.g., Beck & Emery, 1985) and the psychodynamic (e.g.,
Michels, Frances, & Shear, 1985) approaches. Moreover,
we believe that these new contemporary learning ap-
proaches have several advantages over the other two ap-
proaches: The learning approaches are better grounded in
the theories and methods of experimental psychology, they
provide more comprehensive formulations of the etiology
of anxiety disorders, and they provide a more explicit
analysis of factors promoting or inhibiting the development
of different anxiety disorders. In this article we summarize
some of the available research illustrating these points for
each of the six primary anxiety disorders.3,4
Individuals with specific phobias show intense and
irrational fears of certain objects or situations that they
usually go to great lengths to avoid. Watson and Rayner
(1920) originally argued that phobias are simply intense
classically conditioned fears that develop when a neutral
stimulus is paired with a traumatic event, such as occurred
when their Little Albert acquired an intense fear of rats
after hearing a frightening gong paired with the presence of
a rat several times. Indeed, a large number of studies using
retrospective recall, although fraught with interpretive pit-
falls (Mineka & O¨hman, 2002a; Mineka & Sutton, in
press), have confirmed that many people with phobias can
recall a traumatic conditioning event when their phobia
began (e.g., O¨st & Hugdahl, 1981; see Muris & Merckle-
bach, 2001, for a review). This straightforward view of
phobia acquisition was later criticized for several reasons
we will discuss. However, each of the cited problems with
such views generally reflects some ignorance of a wide
variety of vulnerability and invulnerability factors that are
predicted by contemporary learning theory to strongly in-
fluence who develops phobias among all those who have
had the seemingly requisite conditioning experiences. For
example, consider Marian and Emily, the two teenage girls
described at the outset who were both attacked by dogs, yet
only Marian developed dog phobia. Marian was a bit shy
and timid and had had minimal experience with dogs
before having been chased and cornered by three dogs.
Emily, by contrast, was not shy and had had a lot of
experience with dogs prior to having been badly bitten. It
thus makes more intuitive sense that only Marian devel-
oped dog phobia, but why exactly?
Vicarious Conditioning of Fears and Phobias
One criticism of early conditioning approaches centered on
the observation that many people with phobias do not
appear to have had any relevant history of classical condi-
tioning (Rachman, 1978, 1990). How can one account for
the origins of fears and phobias in these individuals? Cli-
nicians have long speculated that simply observing others
experiencing a trauma or behaving fearfully could be suf-
ficient for some phobias to develop. The results of some
retrospective studies are consistent with this idea (e.g.,
Muris & Mercklebach, 2001; O¨st & Hugdahl, 1981). One
such case involved a boy who had witnessed his grand-
3Because of space constraints, our reviews are less thorough for
generalized anxiety disorder and obsessive–compulsive disorder.
4A significant amount of the research that we review is based on
animal studies. We recognize the long-standing controversies about the
relevance of animal research to understanding human emotional disorders.
We and many others have reviewed these issues extensively elsewhere
(e.g., Abramson & Seligman, 1977; McKinney, 1974; Mineka, 1985,
2002; Mineka & Zinbarg, 1991; Pavlov, 1927). We and many (but not all)
others have concluded that there are far more advantages than disadvan-
tages in relying on animal research to advance knowledge about etiology,
treatment, and prevention, as long as careful attention is paid to docu-
menting compelling similarities as well as differences between the animal
and the human observations and findings.
11January 2006 ● American Psychologist
father vomit while dying; shortly thereafter the boy devel-
oped a strong and persistent vomiting phobia. Indeed, in
middle age he even contemplated suicide one time when he
was nauseous and feared vomiting.
The strongest experimental evidence for the role of
vicarious conditioning in phobia acquisition stems from a
primate model showing that strong and persistent phobic-
like fears can indeed be learned rapidly through observa-
tion alone. In Mineka and Cook’s series of experiments on
observational conditioning of snake fear (e.g., Cook,
Mineka, Wolkenstein, & Laitsch, 1985), laboratory-reared
young adult rhesus monkeys who initially were not afraid
of snakes served as observers who watched unrelated wild-
reared model monkeys reacting very fearfully in the pres-
ence of live and toy snakes. These lab-reared observer
monkeys showed rapid acquisition of an intense phobic-
like fear of snakes that did not diminish over a three-month
follow-up period. This vicarious conditioning also occurred
simply through watching videotapes of models behaving
fearfully (Cook & Mineka, 1990), suggesting that humans
are also susceptible to acquiring fears vicariously simply
through watching movies and TV (as expected from nu-
merous anecdotal observations). Thus, traditional condi-
tioning models erred in not attending to the important role
that vicarious conditioning can play in the origins of phobic
Sources of Individual Differences in the
Acquisition of Fears and Phobias
Another problem with traditional views is how to explain
why many individuals who do undergo traumatic experi-
ences do not develop phobias (e.g., Mineka & Sutton, in
press; Mineka & Zinbarg, 1996; Rachman, 1990). For
example, in studies using retrospective recall, many non-
phobic individuals report having had traumatic experiences
in the presence of some potentially phobic object without
having acquired a fear or phobia (e.g., Poulton & Menzies,
2002). From our diathesis-stress perspective, such findings
are actually expected. First, there seems to be a modest
genetically based vulnerability for phobias (e.g., Kendler,
Neale, Kessler, Heath, & Eaves, 1992; Kendler et al.,
1995). This genetic vulnerability may well be mediated
through genetic contributions to fear conditioning (Het-
tema, Annas, Neale, Kendler, & Fredrikson, 2003), which
may in turn be mediated through personality variables such
as high trait anxiety that also seem to serve as vulnerability
factors, affecting the speed and strength of conditioning
(e.g., Levey & Martin, 1981). Moreover, children catego-
rized as behaviorally inhibited (excessively timid, shy, etc.)
at 21 months of age have been found to be at higher risk for
the development of multiple specific phobias (an average of
three to four per child) by seven to eight years of age than
were uninhibited children (32% vs. 5%, respectively) (e.g.,
Biederman et al., 1990). In our example at the outset,
Marian was at higher risk than Emily for developing pho-
bias partially because of her timidity.
Differences in life experiences among individuals can
also strongly affect the outcome of conditioning experi-
ences. Such experiential factors may serve as vulnerability
(or invulnerability) factors for the development of phobias.
The relevant differences in life experiences may occur
before, during, or following a fear-conditioning experience,
and they can act singly or in combination to affect how
much fear is experienced, acquired, or maintained over
time (e.g., Forsyth & Eifert, 1996a, p. 90; Mineka, 1985;
Mineka & Zinbarg, 1996). This point was not appreciated
by proponents of earlier conditioning models because they
extrapolated directly from the results of experiments on
naive animals studied in isolated conditioning chambers to
how conditioning experiences might affect humans in their
everyday lives. But humans are neither “naive” blank slates
nor isolated from the world during conditioning events.
Impact of prior experiences.
known as latent inhibition demonstrates that simple prior
exposure to a conditioned stimulus (CS) before the condi-
tioned and unconditioned stimulus (US) are ever paired
together reduces the amount of subsequent conditioning to
the CS when paired with the US (e.g., Lubow, 1998).
Consistent with this, several studies have shown that chil-
dren who have had more previous nontraumatic encounters
with a dentist are less likely to develop dental anxiety if
subsequently traumatized at the dentist’s office than are
those with fewer prior encounters when they are trauma-
tized (e.g., Kent, 1997). Moreover, Mineka and Cook
(1986) also showed that most monkeys who had initially
simply watched a nonfearful model monkey behaving non-
fearfully with snakes were completely immunized against
acquiring a fear of snakes when subsequently exposed to
fearful monkeys behaving fearfully with snakes. By anal-
ogy, if a child has extensive preexposure to a nonfearful
parent behaving nonfearfully with the phobic object or
situation (e.g., heights) of the other parent, this may im-
munize the child against the later effects of seeing the
12 January 2006 ● American Psychologist
phobic parent behaving fearfully with heights. Using Mar-
ian and Emily as examples, we note that Emily’s own
extensive prior exposure to dogs probably immunized her
from developing dog phobia.
A person’s history of control over important aspects of
his or her environment is another important experiential
variable strongly affecting reactions to frightening situa-
tions. For example, Mineka, Gunnar, and Champoux
(1986) found that infant monkeys reared with a sense of
mastery and control over their environments for 7–10
months later adapted more readily to frightening events and
novel anxiety-provoking situations than did monkeys
reared in identical environments except without the expe-
riences with control. This experimental evidence supports
claims long made by developmentalists using correlational
data that infants and children raised in environments in
which they gain a sense of control over their environment
are less frightened by (and better able to cope with) novel
and frightening events. Such research suggests that chil-
dren reared with a stronger sense of mastery over their
environments should be more invulnerable to developing
phobias following traumatic experiences (see also Chorpita
& Barlow, 1998).
Impact of contextual variables during con-
Several different features of conditioning
events themselves also have a strong impact on how much
fear is acquired. For example, having control over a trau-
matic event (such as being able to escape it) has a major
impact on how much fear is conditioned to CSs paired with
that trauma. Far less fear is conditioned when the aversive
event is escapable than when it is inescapable (e.g.,
Mineka, Cook, & Miller, 1984). Thus, consider our exam-
ple of Marian, who was not able to end the dog attack
herself. Instead, the dog’s owner pulled the dog off her.
Our perspective suggests that if she had instead been able
to escape on her own at the same point during the attack
that the dog owner pulled the dog off her, she would have
been less likely to have developed dog phobia, even though
the dog attack would have been of equal duration and
Impact of postevent variables.
kinds of experiences that people can have following con-
ditioning also affect the strength of the conditioned fear
that is maintained over time. For example, animal and
human research suggests that a person who is exposed to a
more intense traumatic experience (not paired with the CS)
after conditioning of a mild fear is likely to show an
increase in fear of the CS (e.g., Rescorla, 1974; White &
Davey, 1989). This so-called inflation effect suggests that a
person who acquired a mild fear of automobiles following
a minor crash might develop full-blown driving phobia if
he or she were later physically assaulted even though no
automobile was present during the assault (Mineka, 1985;
Mineka & Zinbarg, 1996). Davey (1997) has also described
a US reevaluation process that can occur when a person
receives verbally or socially transmitted information about
the US being more dangerous than when she or he origi-
nally experienced it paired with the US. This can result in
an inflated level of fear to the CS. Finally, Davey and
Matchett (1994) also found that simple mental rehearsal of
CS–US relationships can lead to enhanced strength of the
conditioned fear response.
In sum, a number of factors occurring before, during,
and following potentially traumatic conditioning experi-
ences (direct or vicarious) affect the strength of fear that is
experienced, conditioned, and maintained over time. These
examples of many individual differences in vulnerability
(and invulnerability), as well as various contextual factors
during and after noxious events, go a long way toward
addressing the limitations of earlier conditioning models.
Selective Associations in the Conditioning of
Fears and Phobias
Early conditioning models predicted that fears and phobias
would occur to any random group of objects associated
with trauma. However, clinical observations show that
people are much more likely to have phobias of snakes,
water, heights, and enclosed spaces than of bicycles, guns,
or cars, even though today the latter objects (not present in
our early evolutionary history) may be at least as likely to
be associated with trauma. Seligman (1971) and O¨hman
(e.g., O¨hman, Dimberg, & O¨st, 1985) argued that primates
may be evolutionarily prepared to rapidly associate certain
kinds of objects (such as snakes, spiders, water, heights)
with aversive events. This is because there may have been
a selective advantage in the course of evolution for pri-
mates who rapidly acquired fears of certain objects or
situations that posed threats to humans’ early ancestors (see
also Cook & Mineka, 1990). Consistent with this, several
studies have shown that the content of most phobias is rated
by independent raters as “prepared” in the sense that it was
probably dangerous to pretechnological humans (e.g., see
O¨hman & Mineka, 2001).
Thus, prepared fears are not seen as inborn or innate
but rather as very easily acquired and/or especially resistant
to extinction. O¨hman and his colleagues conducted an
important series of human conditioning experiments sup-
porting this theory. Using slides of snakes and spiders as
fear-relevant CSs and mild shock as the US, they have
consistently found superior conditioning compared with
what is found when more fear-irrelevant CSs, such as slides
of flowers, mushrooms, or electric outlets, are used (e.g.,
O¨hman et al., 1985; O¨hman & Mineka, 2001). More recent
studies have also shown that with fear-relevant CSs (but
not with fear-irrelevant CSs), conditioning can even occur
when subliminal CSs are used (i.e., CSs that cannot be
consciously identified; e.g., O¨hman & Soares, 1998). Other
research shows that fear learning with fear-relevant stimuli
is more impenetrable to conscious cognitive control than is
fear learning with fear-irrelevant stimuli (see Mineka &
O¨hman, 2002b). Such results may help explain the irratio-
nality of phobias. If the fears of individuals with phobias
arise automatically from cognitive structures not under the
control of conscious intentions, the fears will remain unaf-
fected by rational analysis (O¨hman & Soares, 1993).
In addition, Cook and Mineka’s experiments (e.g.,
1989, 1990) using videotaped model monkeys showed that
naive monkeys can easily acquire fears of such fear-rele-
13January 2006 ● American Psychologist
vant stimuli as a toy snake or a toy crocodile but not of such
fear-irrelevant stimuli as flowers or a toy rabbit. It is
important to note that these lab-reared monkeys had no
prior exposure to any of the fear-relevant or fear-irrelevant
stimuli before participating in these experiments. Thus, the
monkey experiments more strongly support the role of
evolutionary factors in the biased associations seen with
preparedness effects because there is no possibility that the
monkeys had preexisting negative associations to snakes,
as may be the case in human experiments on this topic (see
Mineka & O¨hman, 2002b).
In both monkeys and humans, therefore, evolutionary
fear-relevant stimuli more readily enter into selective as-
sociations with aversive events, and these same stimuli
seem more likely than others to become the objects of
human phobias. Moreover, the special characteristics of
fear learning seen with fear-relevant (but not fear-irrele-
vant) stimuli (e.g., its automaticity and its resistance to
higher cognitive control) suggest that the acquisition of
phobias involves a primitive basic emotional level of learn-
ing that humans share with many other mammalian species
(O¨hman & Mineka, 2001).
In summary, the origins and maintenance of fears and
phobias are considerably more complex than was assumed
by traditional conditioning models. Nevertheless, these
complexities are expected from the standpoint of contem-
porary research on conditioning, which reveals a large
variety of vulnerability, invulnerability, and contextual
variables that have a strong impact on the outcome of direct
and observational conditioning experiences. Moreover,
evolutionary pressures seem to have affected the kinds of
objects and situations that humans are most prone to learn
Individuals with social phobias show excessive fears of
situations in which they might be evaluated or judged by
others, and they either avoid such situations or endure them
with marked distress. Several theorists have suggested that
social phobia can arise as a result of direct traumatic
conditioning, and several retrospective studies (with all
their inherent limitations) have produced evidence consis-
tent with this idea. For example, McCabe, Antony, Sum-
merfeldt, Liss, and Swinson (2003) reported that 92% of
their adult sample with social phobia reported a history of
severe teasing in childhood, compared with only 50% and
35% in their panic disorder and obsessive–compulsive
disorder groups, respectively. Another study found that
56% of those with specific social phobias recalled direct
traumatic conditioning experiences as having played a role
in the origins of their social phobia (Stemberger, Turner,
Beidel, & Calhoun, 1995). One relevant case involved a
young man who came home one day to find his fiance ´e in
bed with his best friend (a humiliating and traumatic ex-
perience for him). Over the next few months he developed
severe social phobia.
Social Learning and Social Phobia
Another potentially important pathway to the development
of social phobia is through social learning. Vicarious con-
ditioning—that is, simply observing another being ridi-
culed or humiliated or behaving in a very anxious way in
some social situation—is one potent form of social learning
that may be sufficient to make the observer develop social
phobia of similar situations. For example, a vulnerable
student might begin to develop social phobia after watching
a high school friend give a speech in front of class and then
be sharply criticized by the teacher and laughed at by the
class. O¨st and Hughdahl (1981) found that 13% of individ-
uals with social phobia recalled vicarious learning experi-
ences as having played a role in the origins of their social
phobia. Additionally, there is evidence of modeling of
social anxiety in families of those with social phobia.
Retrospective studies show that adults with social phobia
report that their parents were more likely than the parents
of those without social phobia to have been avoidant of
social interactions (e.g., Bruch & Heimberg, 1994). It is
noteworthy that Rapee and Melville (1997) extended these
findings by including reports from the mothers of individ-
uals with social phobia and found that the reports from both
the offspring and the mothers indicated more social avoid-
ance among the families of patients with social phobia than
among those of the nonclinical controls.
Direct social reinforcement and verbal instruction are
also likely to play a role. For example, Dadds, Barrett,
Rapee, and Ryan (1996) found that parents of anxious
children were more likely than other parents to reciprocate
their children’s proposals of avoidant solutions. It is im-
portant to note that discussion of the potentially threatening
situations with the parents appears to strengthen the anx-
ious children’s avoidant tendencies (Barrett, Rapee, Dadds,
& Ryan, 1996).
Finally, culturally transmitted display rules and norms
are another form of social learning that is likely to exert
powerful influences on the expression of social anxiety, as
is illustrated by the Japanese disorder taijin kyofusho,
which is related to the Western diagnosis of social phobia.
People with this disorder fear that they will do something
to embarrass or offend others. For example, they may fear
they will offend others by blushing, emitting an offensive
odor, or staring inappropriately into someone else’s eyes
(Barlow, 2002; Kirmayer, 1991). By contrast, Westerners
with a diagnosis of social phobia are afraid of other people
because they believe they will be the object of scrutiny by
others and will act in ways that will be personally humil-
iating or embarrassing. Kirmayer (1991) and Tanaka-Mat-
sumi, Seiden, and Lam (1996) have argued that the pattern
of symptom expression in taijin kyofusho and social phobia
has clearly been shaped by cultural factors. In Japan, a
collectivistic society that reinforces interdependence and
social affiliation, there is a great deal of emphasis on
implicit communication—being able to guess another’s
thoughts and feelings and being sensitive to them. People
who make too much eye contact are likely to be considered
aggressive and insensitive; children are taught to look at the
14 January 2006 ● American Psychologist
throats of people with whom they are conversing rather
than into their eyes.
Preparedness and Social Phobia
Another fruitful perspective comes from O¨hman et al.’s
(1985) extension of preparedness theory to social anxiety.
O¨hman et al. proposed that social anxiety is a by-product of
the evolution of dominance hierarchies and therefore pre-
dicted that social stimuli signaling dominance and intra-
specific threat should be fear-relevant or prepared CSs for
social anxiety. Because angry facial expressions are prom-
inent signals during dominance conflicts among primates,
O¨hman et al. tested whether angry facial expressions would
function as fear-relevant CSs for social fears in humans.
O¨hman and Dimberg (1978) indeed found superior condi-
tioning when slides of angry faces were used as CSs,
compared with slides of happy faces or neutral faces (all
paired with mild shocks). Subsequent studies have reported
other intriguing findings, parallel to those found for specific
phobias. For example, fear CRs can even be conditioned to
subliminal presentations of angry faces—that is, to angry
faces that cannot be consciously perceived (e.g., Esteves,
Parra, Dimberg, & O¨hman, 1994). Thus, a person can claim
to “know” rationally that a social situation is safe and yet
still experience anxiety that is automatically activated in
response to subtle cues that are not consciously processed.
Behavioral Inhibition as a Temperamental
Diathesis for Social Phobia
Obviously not everyone who has experienced a socially
traumatic episode or observed a socially anxious model
will develop social phobia. For example, Stemberger et al.
(1995) reported that 20% of their control sample without
social phobia nevertheless recalled socially traumatic ex-
periences. As for specific phobias, one source of individual
differences involves the temperamental diathesis known as
behavioral inhibition, which may influence the outcome of
exposure to socially traumatic experiences and/or socially
anxious models (e.g., Gray, 1982; Kagan, 1994). Indeed,
early behavioral inhibition not only predicts the onset of
multiple specific phobias in childhood (when they are most
likely to emerge) but also social phobias in adolescence
(when they are most likely to develop; e.g., Hayward,
Killen, Kraemer, & Taylor, 1998; Kagan, 1997).
Uncontrollability and Social Phobia
Perceptions of uncontrollability are another likely source of
individual differences in vulnerability to the effects of
socially traumatic episodes or exposure to socially anxious
models. For example, animal research has shown that un-
controllable (but not controllable) electric shock increases
submissiveness (e.g., Williams & Lierle, 1986). Second,
animal studies (e.g., Uhrich, 1938) of repeated social defeat
(another uncontrollable stressor) show that it also leads to
increased submissiveness to any other conspecific behaving
in an aggressive manner. Moreover, repeated social defeat
in animals also produces many of the classic learned help-
lessness effects usually associated with uncontrollable
shock, including escape deficits (Hebert, Evenson, Lumley,
& Meyerhoff, 1998) and exaggerated fear CRs (Williams
& Scott, 1989). We concur with Williams and colleagues,
who concluded from such findings that the deleterious
effects of social defeat are probably mediated by percep-
tions of uncontrollability. Moreover, such research sug-
gests that perceptions of uncontrollability are likely to play
a role in the origins of social anxiety. Indeed, cross-sec-
tional evidence documents a moderate to strong association
between perceptions of uncontrollability and social phobia
in people (e.g., Leung & Heimberg, 1996).
In summary, a model of social phobias based on
contemporary learning principles can account for many of
the complexities involved in their etiology because it ac-
knowledges both temperamental diatheses such as behav-
ioral inhibition, and experiential diatheses such as learning
histories, including those that lead to perceptions of uncon-
trollability. Our model of social phobia also takes into
account the preparedness of certain cues to become CSs for
social anxiety, as well as social learning experiences as
alternative pathways to the acquisition of social phobia.
Panic Disorder With and Without
People with panic disorder (PD) experience recurrent un-
expected panic attacks that occur without their being aware
of any cues or triggers, and they must also experience
worry, anxiety, or behavioral change related to having
another attack. Many, but not all, people with PD also go
on to develop some degree of agoraphobic avoidance of
situations in which they perceive that escape might be
either difficult or embarrassing if they were to have a panic
attack. An early conditioning approach to panic disorder
with agoraphobia (PDA) was proposed by Goldstein and
Chambless (1978), who argued, as had others, that extero-
ceptive conditioning (where CSs impinge on the external
sensory receptors like eyes and ears) is central to the
development of agoraphobia. But Goldstein and Chambless
also described a process they termed “fear of fear” devel-
oping from interoceptive conditioning. In interoceptive
conditioning, the CSs are the body’s own internal sensa-
tions (e.g., Razran, 1961). Thus, Goldstein and Chambless
proposed that when low-level somatic sensations of anxiety
precede and are paired with higher levels of anxiety/panic,
the low-level somatic sensations of anxiety come to elicit
high levels of anxiety and panic.5
Such early conditioning approaches to the etiology of
panic disorder were criticized for several reasons not dis-
cussed here (e.g., Clark, 1988; McNally, 1994). However,
Bouton, Mineka, and Barlow (2001) reviewed these criti-
cisms and presented a contemporary learning theory per-
spective that overcomes these limitations. Bouton et al.’s
perspective also builds on growing evidence from two
5The distinction between panic disorder and agoraphobia was not
clearly stated until the DSM–III in 1980 (American Psychiatric Association,
1980)—two years after the article by Goldstein and Chambless (1978) ap-
peared. Moreover, knowledge of the psychometric and neurobiological dif-
ferences between panic and anxiety was not developed until even later.
15 January 2006 ● American Psychologist
sources that anxiety and panic are at least partially unique
emotional experiences. First, psychometric analyses of
panic- and anxiety-related symptomatology reveal two sep-
arable factors: Panic is accompanied by strong autonomic
arousal, extreme fear, and fight-or-flight action tendencies,
and anxiety is accompanied by apprehension, worry, and
tension (Barlow, 1988, 2002; Craske, 1999). Second, neu-
robiological research supports the existence of these two
partially distinct emotional states of panic and anxiety (e.g.,
Fanselow, 1994). In what follows, it is important to keep
this distinction between anxiety and panic in mind because
both are central to PD and PDA.
Exteroceptive and Interoceptive Conditioning
in Panic Disorder
Bouton et al. (2001) argued that the conditioning that may
occur in vulnerable individuals (see the Vulnerability Fac-
tors for Panic Disorder section) during initial panic attacks
sets the stage for the development of PD and PDA. The
initial attacks become associated with initially neutral in-
teroceptive and exteroceptive CSs through conditioning.
Specifically, because of their status as terrifying emotional
events, replete with strong interoceptive stimuli, panic at-
tacks are capable of supporting conditioning, according to
contemporary research (e.g., Bouton et al., 2001; Forsyth &
Eiffert, 1996b). Prototypic CSs during initial (as well as
later) panic attacks include heart palpitations and dizziness
(interoceptive CSs), as well as escalators and malls (ex-
teroceptive CSs). Bouton et al. proposed that the primary
effect of this conditioning is that anxiety becomes condi-
tioned to these CSs, but another effect is that panic attacks
themselves are also likely to be conditioned to certain
internal cues. For example, one young man with PD, who
was particularly sensitive to signs of his heart racing,
surprisingly experienced a panic attack one night when he
got excited while watching an announcement that his fa-
vorite presidential candidate had won a hotly contested state.
Thus, the panic attack occurred at a time when he was happy
and excited; however, this was also a time when his heart
was racing, which probably served as an interoceptive CS.
On the basis of the premise that the conditioned re-
sponse (CR) must resemble the unconditioned response
(UR), some earlier critics objected that panic (the UR)
cannot be involved in conditioning of anxiety (the CR)
(e.g., McNally, 1994), but this criticism is based on out-
dated views of conditioning (Bouton et al., 2001). Today it
is known that although some CRs are very similar to certain
components of the UR, other CRs are actually opposite in
nature to the UR, although in all cases they serve to
“prepare” for an upcoming US and UR. For example,
aversive conditioning in rats usually involves a CR like
freezing (anxiety) rather than the activity burst that consti-
tutes the UR to a shock US (analogue of a panic attack)
(e.g., Fanselow, 1994).
Bouton et al. (2001) also noted that different CSs
paired with the very same US often result in qualitatively
different CRs to the different CSs. For example, in fear
conditioning, Cook, Hodes, and Lang (1986) showed that
when fear-relevant CSs (e.g., snake slides) are paired with
mild electric shock (US), one component of the CR is heart
rate acceleration, in contrast to heart rate deceleration seen
when fear-irrelevant CSs (e.g., flower slides) are paired
with the shock US (see O¨hman & Mineka, 2001). Thus,
some internal and external cues present before and during
panic attacks may become conditioned to elicit anxiety
(different from the panic UR), and others may become
conditioned to elicit panic itself. Interoceptive cues in close
proximity to panic (as opposed to more distal external cues)
may serve as “prepared” or fear-relevant CSs and may be
especially likely to elicit panic, just as other fear-relevant
CSs seem especially likely to condition CRs that strongly
resemble URs (see O¨hman & Mineka, 2001).
Interoceptive conditioning is robust and stable (Raz-
ran, 1961). In one example, Razran described a dog that
received exposure to distention of the intestine (interocep-
tive CS) paired with a mixture of 10% carbon dioxide
delivered directly to the trachea (interoceptive US). Fairly
quickly, the CS of intestinal distention came to elicit hy-
percapnic respiratory changes (very labored breathing) as
the CR. This example is particularly relevant to PD and
PDA because various internal sensations like intestinal
contractions (interoceptive CSs) might be paired with sen-
sations of hyperventilation (interoceptive USs—often oc-
curring during panic); later the contractions might come to
elicit hyperventilation or even panic by themselves (Bouton
et al., 2001).
Interoceptive conditioning has also been observed
when small doses of a drug like morphine serve as CSs for
larger doses of morphine USs (e.g., Kim, Siegel, & Pat-
enall, 1999). Moreover, small increases in blood pressure
can become associated with larger increases in blood pres-
sure (e.g., Dworkin, 1993). So weak versions of some event
can become conditioned by pairing it with a strong version
of the same event—much as when early internal signs of
panic become conditioned when paired with full-blown
panic attacks (Bouton et al., 2001).
The Development of Agoraphobia
As noted, many people who develop panic attacks and
panic disorder go on to develop some degree of agorapho-
bic avoidance of situations in which they fear or expect
they may have a panic attack in the future (e.g., Barlow,
2002; Craske, 1999). Many of these situations are the
commonly observed ones such as shopping malls, driving,
standing in line, sitting in a theater, and so forth. Anxiety
and avoidance of such situations have long been thought to
develop as a result of exteroceptive conditioning of anxiety
to these situations when panic attacks have occurred there
in the past, followed by learned avoidance of these situa-
tions to minimize anxiety. Moreover, as the disorder de-
velops, these agoraphobic fears and avoidance often gen-
eralize to other similar situations, as would be expected on
the basis of the principles of classical conditioning. In
addition, for many individuals, numerous everyday activi-
ties also come to be avoided, such as aerobic exercise,
sauna baths, scary movies, caffeine intake, and even sexual
activity because the interoceptive cues created by these
16 January 2006 ● American Psychologist
activities resemble those experienced at the beginning of a
panic attack (and often trigger panic).
Who is most vulnerable to developing agoraphobia?
Two major risk factors are gender and employment:
Women are far more likely to than men to develop agora-
phobia, and people who must leave the house to work are
less likely than those who do not work or who work from
home to develop agoraphobia (Barlow, 2002; Craske,
1999). The learning theory approach can explain both of
these factors because in each case the person is allowed to
avoid his or her feared situations rather than be exposed to
them, which would extinguish his or her anxiety. For
example, in American culture it is more acceptable for
women than men to stay home and adopt a traditionally
feminine role (Chambless & Mason, 1986). We might also
expect that panic attacks perceived as particularly unpre-
dictable and uncontrollable might well lead to the devel-
opment of greater agoraphobic avoidance because these
factors are known to be associated with greater condition-
ing of anxiety (e.g., Bouton et al., 2001; Mineka & Zinbarg,
1996; see also Craske, Glover, & DeCola, 1995).
Further Complexities of the Role of
Conditioning in Panic Disorder and
Conditioned stimuli not only come to elicit conditioned
responses but also to facilitate or inhibit responses con-
trolled by other events (Bouton et al., 2001). For example,
for individuals with PD or PDA, inhibitory CSs for safety
(such as the presence of a trusted companion) can attenuate
panic symptoms ordinarily elicited by carbon dioxide in-
halation (a US which is widely used as a panic provocation
agent; Carter, Hollon, Carson, & Shelton, 1995). By con-
trast, CSs for anxiety (e.g., a warning about facing a stress-
ful day at work) can strengthen ongoing avoidance behav-
ior (such as agoraphobic avoidance like calling in sick) or
engagingin “safety behaviors”
Conditioned anxiety also serves to increase startle
responses (e.g., Lang, 1995), and Bouton et al. (2001) have
argued that is quite likely that CSs for anxiety similarly
lower the threshold for, or exaggerate, panic reactions (see
also Zinbarg, Barlow, Brown, & Hertz, 1992). Indeed,
baseline levels of anxiety strongly predict who will expe-
rience panic in response to various panic provocation
agents as USs (e.g., Bouton et al., 2001). Moreover, several
prospective monitoring studies have shown that anxiety is
very often a precursor to (and may potentiate) panic at-
tacks. For example, Bas ¸og ˇlu, Marks, and Sengu ¨n (1992)
had 39 patients with PDA carefully monitor their panic and
anxiety over three 24-hour periods. Over two thirds of their
patients reported that elevated anxiety preceded all their
panic attacks, and 87% reported that this happened at least
some of the time. Finally, because the same CR (anxiety)
elicited from two different sources can summate (e.g.,
baseline anxiety and anxiety elicited by a shopping mall),
one would also expect that baseline anxiety could potenti-
ate agoraphobic anxiety and avoidance.
(such as carrying
Vulnerability Factors for Panic Disorder
Far more people experience panic attacks than go on to
develop panic disorder. Among those individuals who ex-
perience panic attacks, what determines who will develop
PD or PDA? Bouton et al. (2001) considered three types of
vulnerability factors that seem likely to predispose some
people who experience panic attacks to develop PD or
PDA. First, they discussed a moderate nonspecific genetic
vulnerability for PD and PDA (that overlaps especially
with the genetic vulnerability for phobias) that may well be
mediated by temperamental or personality vulnerability
factors like neuroticism or trait anxiety (e.g., Kendler et al.,
1995). As discussed earlier, these personality variables
serve to potentiate conditioning of anxiety responses as
well as aversive expectations (e.g., Zinbarg & Mohlman,
1998). Thus, genetic and temperamental variables seem to
serve as nonspecific vulnerability factors that increase the
likelihood of developing panic (e.g., Hayward, Killen,
Kraemer, & Taylor, 2000) as well as various other anxiety
A second set of nonspecific vulnerability factors dis-
cussed by Bouton et al. (2001) are prior learning experi-
ences that lead to perceptions of lack of control and help-
lessness and that may serve as psychological vulnerability
factors influencing the development of panic, agoraphobia,
and other emotional disorders. Early experience with un-
controllable stressful life events such as death and divorce
can enhance vulnerability to PD and depression (see Bou-
ton et al., 2001). Conversely, as noted earlier, early expe-
riences with control and mastery in infancy and childhood
are important for developing the ability to cope with stress
and anxiety-provoking situations (invulnerability factors)
(e.g., Chorpita & Barlow, 1998; Mineka et al., 1986).
Finally, Bouton et al. (2001) also reviewed some more
specific learning experiences that may play a more unique
role in creating risk for developing PD and PDA per se. For
example, Ehlers (1993) found that people who experienced
panic attacks as adults (relative to controls) were likely to
have stronger learning histories of having been encouraged
to engage in sick role behavior when experiencing panic
symptoms (but not cold symptoms). Adults who had expe-
rienced some panic attacks were also more likely than
controls to report others having had chronic illnesses in
their households while growing up, suggesting that observ-
ing a lot of physical suffering may contribute to the eval-
uation of somatic symptoms as dangerous.
In summary, Bouton et al.’s (2001) perspective on PD
and PDA argues that initial panic attacks set the stage for
conditioning of anxiety to external and internal cues asso-
ciated with panic, thus explaining the origins of anticipa-
tory and agoraphobic anxiety. Moreover, the process of
interoceptive conditioning also allows certain bodily sen-
sations to become conditioned to elicit panic itself such
that, for example, heart palpitations occurring early in an
attack become predictors of the rest of the attack and
acquire the capacity to provoke panic. Finally, people with
certain genetic/temperamental and/or experiential vulnera-
bilities will show stronger conditioning of anxiety and
17 January 2006 ● American Psychologist
panic, explaining why only a subset of those who have
panic attacks develop PD or PDA.
Posttraumatic Stress Disorder
The symptoms of posttraumatic stress disorder (PTSD)
include reexperiencing the trauma, passively avoiding re-
minders of the trauma, numbing of affect, and heightened
general arousal. Although a traumatic event is necessary
for the diagnosis of PTSD, there are many puzzles in the
development of PTSD that require explanation. Consider
the cases of Ahmet and Hasan mentioned at the outset—
both were torture victims in Turkey (M. Bas ¸og ˇlu, personal
communication, 1997). Ahmet had been a political activist
who was imprisoned by the government for his political
activities for 48 months, during which he was tortured
approximately 300 times. By contrast, Hasan was an ordi-
nary citizen who was not politically active. One day he had
called the police to report what he thought was a burglar
lurking in his neighborhood. Unable to find the alleged
burglar, the police arrested Hasan instead, charging him
with having intentionally made a crank call. Despite the
fact that Hasan was imprisoned and tortured for only a few
weeks and experienced far less torture than Ahmet, Hasan
developed PTSD, whereas Ahmet did not. Why? We be-
lieve this seemingly puzzling fact and many other com-
plexities associated with PTSD are entirely consistent with
contemporary learning theory research on the effects of
uncontrollable and unpredictable stress in animals (some-
times known as learned helplessness effects).
Before exploring the implications of this perspective,
we examine some of the compelling similarities between
the symptoms of PTSD and the effects of uncontrollable
and unpredictable stress, as well as the stressors involved in
producing each. Regarding symptoms, animals exposed to
uncontrollable and/or unpredictable stress, like individuals
with PTSD, show heightened generalized anxiety and
arousal. They also show enhanced passive avoidance be-
havior (e.g., Rush, Mineka, & Suomi, 1982) resembling the
avoidance symptoms of PTSD (Foa, Zinbarg, & Rothbaum,
1992). In addition, the opioid-mediated analgesia (relative
insensitivity to pain) produced by cues associated with
uncontrollable stress in animals may resemble certain
numbing symptoms seen in PTSD. For example, Vietnam
veterans with PTSD who watched a video of a combat
scene (a reminder cue for their uncontrollable stressful
experiences) showed analgesia (Pitman, Van der Kolk, Orr,
& Greenberg, 1990). Finally, as discussed by Foa et al., the
reexperiencing symptoms of PTSD include fear and dis-
tress at exposure to reminders of trauma (as well as night-
mares and flashbacks), and these emotional reactions can
be seen as conditioned emotional responses elicited by
reminder cues (CSs previously paired with trauma).
In addition, the intense physical stressors typically
used in animal studies of the effects of uncontrollable and
unpredictable stress (e.g., electric shocks, defeats in phys-
ical fighting) are similar to several forms of human trauma
often associated with PTSD—torture (e.g., electric shocks,
beatings), child abuse, and assault. Considered together, the
symptom similarities and resemblance of physical stressors
used to produce them lead us to believe that the uncontrol-
lable, unpredictable stress animal model is highly relevant
for understanding PTSD. Specifically, to aid in the under-
standing of the onset, severity, and course of PTSD symp-
toms, Bas ¸og ˇlu and Mineka (1992) argued for the relevance
of animal research on variables occurring before, during, or
following exposure to unpredictable, uncontrollable stress,
the topic to which we now turn.
The most obvious prediction from the animal model is that
traumas that are perceived to be uncontrollable and unpre-
dictable are more likely to result in PTSD. Results from
retrospective studies consistent with this prediction have
been reported (e.g., Bolstad & Zinbarg, 1997; Regehr,
Cadell, & Jansen, 1999). In addition, organisms differ in
how they respond to the experience of uncontrollable
stress, and such differences can strongly affect the out-
come. For example, animal research on social defeat shows
that the analgesia induced by extensive attacks from a
conspecific (analgesia is a characteristic facet of learned
helplessness) is more highly correlated with the extent to
which the animal assumed the characteristic posture of
defeat than with the number of bites received from the
conspecific (e.g., Rodgers & Hendrie, 1983). Conversely,
Weiss, Glazer, and Pohorecky (1976) found that allowing
animals to express aggression during uncontrollable stress
attenuated the effects of the stress.
Such findings led Bas ¸og ˇlu and Mineka (1992) to hy-
pothesize that the amount of trauma inflicted during torture
(or assault) may be less predictive of the long-term emo-
tional effects of the torture than is the victim’s psycholog-
ical state of resistance and fighting back versus giving up
and conceding defeat. Consistent with this, studies of po-
litical prisoners and assault survivors have shown that
survivors with PTSD were more likely than those without
PTSD to retrospectively report having experienced mental
defeat during their traumatization (e.g., Ehlers, Maercker,
& Boos, 2000). It is important to note that Dunmore, Clark,
and Ehlers (2001) also found that perceived mental defeat
measured within four months of an assault predicted PTSD
symptom severity at follow-up assessments six and nine
months after the assault.
The effects of prior experiences with control or lack of
control on reactions to subsequent uncontrollable trauma
are complex. Sometimes prior uncontrollable stress sensi-
tizes an organism to the harmful effects of subsequent
exposure to such trauma (e.g., Moye, Hyson, Grau, &
Maier, 1983). This means that as repeated exposures to
trauma occur, the adverse effects get larger rather than
smaller (as in habituation). Thus, a history of prior trauma
(especially repeated trauma, which is more likely to be
perceived as uncontrollable) should be associated with
increased risk of developing PTSD to a recent trauma. For
example, one retrospective study showed that a history of
repeated childhood abuse had an even greater sensitizing
effect than a single incident of child abuse (Bolstad &
18 January 2006 ● American Psychologist
Zinbarg, 1997). Moreover, the results of a recent meta-
analysis support this prediction, although prior trauma was
more strongly related to PTSD when the prior trauma
involved interpersonal violence such as child abuse rather
than combat exposure or an accident (Ozer, Best, Lipsey, &
Weiss, 2003). We should also note that there are a few
studies supporting a moderate genetic contribution to lia-
bility for PTSD (e.g., True et al., 1993), which might
perhaps be mediated by sensitivity to uncontrollable and
unpredictable stress (an interesting question for future
By contrast, a prior history of control over stressful
events can immunize against the harmful effects of subse-
quent uncontrollable stress (e.g., Williams & Maier, 1977).
Hence, we also predict that a prior history of control over
stress can immunize against the development of PTSD to a
recent trauma. Consistent with this prediction (and the case
studies at the outset), there is suggestive evidence from a
study of torture survivors that psychological readiness prior
to being tortured (such as the psychological readiness Ah-
met had because of his background as a political activist) is
associated with a decreased likelihood of subsequently
developing PTSD despite very high levels of torture (Ba-
s ¸og ˇlu, Mineka, et al., 1997). Psychological readiness was
measured by a questionnaire asking about the participants’
knowledge of torture methods, their expectations of being
imprisoned and tortured, and their training in stoicism
techniques, among other things. By contrast, torture vic-
tims like Hasan, who were ordinary citizens arrested “out
of the blue,” were much less prepared for their torture
experiences. Therefore, despite the fact that they experi-
enced far less torture, they were more likely to develop
PTSD and other forms of psychopathology than were the
torture victims who had been political activists. Given that
the psychological readiness measure used was multifac-
eted, the active ingredients responsible for this protective
effect are unclear. However, the protective effects of readi-
ness were probably mediated at least in part by rendering
the subsequent torture more predictable. In addition, the
stoicism training may have led the torture to be perceived
as partially controllable and therefore less likely to produce
mental defeat; indeed the nonactivists reported less per-
ceived control but more distress than the activists.
One potential discrepancy between the effects of uncon-
trollable, unpredictable stress and the symptoms of human
PTSD is that at least some of the effects of uncontrollable,
unpredictable stress in animals are short-lived (usually
dissipating over several days), in contrast to the sometimes
chronic nature of PTSD, which often lasts many years. How-
ever, this discrepancy may be more apparent than real. First,
only a small number of individuals who develop initial signs
of PTSD go on to develop chronic PTSD (e.g., in one study,
0% of male assault victims and approximately 30% of female
assault victims developed PTSD; Riggs, Rothbaum, & Foa,
1995). Second, if rats are simply exposed at two- to three-day
intervals to the context in which they had previously experi-
enced uncontrollable stress (without actually experiencing the
stress), the usual time course for at least some of the effects of
the uncontrollable stress is prolonged “indefinitely” (Maier,
2001, p. 763). Such contexts associated with stress become
CSs for anxiety, and we consider such exposure to anxiety
CSs as eliciting conditioned emotional responses that are
analogous to the anxiety-provoking aspects of the reexperi-
encing symptoms in PTSD.
On the basis of earlier reports of Maier’s (2001)
findings with rats, Mineka and Zinbarg (1996) predicted
that the severity of reexperiencing symptoms after a trauma
should influence the course of the symptoms of PTSD;
greater reexperiencing would be associated with a more
persistent course. At least six prospective studies of trauma
survivors now support this prediction (e.g., Ehlers, Mayou,
& Bryant, 1998). Moreover, in several studies, early reex-
periencing symptoms significantly predicted later PTSD,
but early avoidance (e.g., McFarlane, 1992) or numbing
symptoms (e.g., Rothbaum, Foa, Riggs, Murdock, &
Walsh, 1992) did not, suggesting some degree of specificity
between early reexperiencing symptoms and later PTSD.
What other kinds of posttrauma events can affect the
course and intensity of PTSD symptoms? The inflation and
US reevaluation effects, described earlier, can also affect
the course of PTSD. Thus, mild PTSD symptoms could
become full-blown PTSD when there is reason to reevalu-
ate the danger posed by the original trauma (e.g., finding
out weeks after an assault that one’s assailant was in fact a
convicted murderer). In addition, reinstatement of fear is
the term for the phenomenon in which after a CS has
extinguished, the CS can regain its ability to elicit a CR by
simply exposing the animal to the US (not paired with the
CS) (cf. Rescorla & Heth, 1975). Thus, a posttherapy
trauma not linked with any of the cues associated with the
original trauma could lead to a reinstatement of the original
In summary, there are compelling parallels between
the literature on uncontrollable, unpredictable stress in an-
imals and human PTSD. These include striking similarities
in the symptoms of each as well as significant resemblance
between the kinds of stressors. These parallels suggest that
perceptions of uncontrollability and unpredictability play a
role in the development and course of PTSD. Moreover,
striking parallels have been observed between the trauma
and pre- and posttrauma variables affecting the outcome of
exposure to unpredictable, uncontrollable stress in animals
and several forms of human trauma associated with PTSD.
Generalized Anxiety Disorder
People with generalized anxiety disorder (GAD) are pri-
marily characterized by chronic excessive worry about a
number of events or activities for at least six months, and
the worry must be experienced as difficult to control (see
Footnote 3). Several theorists have suggested that uncon-
trollable and unpredictable aversive events may play an
important role in the development of GAD (e.g., Barlow,
2002; Mineka, 1985; Mineka & Zinbarg, 1996). As re-
viewed earlier, such events are much more stressful and
create more fear and anxiety than do controllable and
predictable aversive events. Although the unpredictable
19 January 2006 ● American Psychologist
and uncontrollable events involved in GAD are generally
not as severe and traumatic as those involved in the origins
of PTSD, there is some evidence (albeit limited) that peo-
ple with GAD may be more likely to have a history of
childhood trauma than are people with several other anxi-
ety disorders (Borkovec, Alcaine, & Behar, 2004). More-
over, people with GAD have far less tolerance for uncer-
tainty than do nonanxious controls (Dugas, Buhr, &
Ladoucuer, 2004), which suggests that they are especially
disturbed by not being able to predict the future (Roemer,
Orsillo, & Barlow, 2002). People with GAD may also have
a relative lack of safety signals in their lives (indicators
telling them when bad things are very unlikely to happen);
this may help explain why they feel constantly tense and
vigilant for possible future threats (Mineka & Zinbarg,
1996; Rapee, 2001).
Conversely, there is evidence relevant to individual
differences suggesting that people with extensive experi-
ence controlling important aspects of their lives may be
immunized against developing GAD (Barlow, 2002; Chor-
pita & Barlow, 1998; Mineka & Zinbarg, 1996). For ex-
ample, as noted earlier, infant monkeys reared in control-
lable as opposed to uncontrollable environments were
much better able to cope with frightening and novel situ-
ations later in their first year of life (Mineka et al., 1986).
Moreover, because neuroticism and trait anxiety are known
risk factors for anxiety disorders, it is possible that people
high on neuroticism or trait anxiety may be especially
susceptible to the effects of uncontrollable and unpredict-
able aversive events.
In recent years worry has increasingly come to be seen
as the central feature of GAD. Borkovec (1994) and col-
leagues (e.g., Borkovec et al., 2004) have incorporated
ideas from learning theory to help explain why worry is
such a persistent process. They have investigated both the
perceived benefits of worry as well as what actual functions
worry serves. Most of the perceived benefits of worry
center around people’s beliefs that worry helps avoid ca-
tastrophe (either superstitiously or in reality) and deeper
emotional topics that they do not want to think about.
Investigations of how worry actually functions help reveal
why the worry process can become so self-sustaining. First,
when people with GAD worry, their emotional and phys-
iological responses to aversive imagery are actually sup-
pressed. Borkovec et al. argued that this suppression of
emotional and physiological responses serves to reinforce
the process of worry (i.e., increases its probability). Be-
cause worry suppresses physiological responding, this also
prevents the person from fully experiencing or processing
the topic that is being worried about; such processing is
necessary if extinction of anxiety is to occur. Thus, the
threatening meaning of the topic will be maintained. Fi-
nally, Borkovec et al. noted that worry may also be rein-
forced superstitiously because the vast majority of things
people worry about never happen. The theory thus pro-
poses that worry serves as a cognitive avoidance response.
Because avoidance responses in animals are notoriously
resistant to extinction (e.g., see Mineka, 1979, and Mineka,
Yovel, & Pineles, 2002, for reviews), Borkovec argued by
analogy that this helps to explain why worry is so
But why do people with GAD come to perceive worry
as uncontrollable—a defining feature in the DSM–IV
(APA, 1994)? Worry is not an enjoyable activity and can
actually lead to a greater sense of danger and anxiety
because of all the catastrophic outcomes the worrier envi-
sions. Moreover, it can actually lead to more negative
intrusive thoughts (e.g., Wells & Papageorgiou, 1995), and
for these and other reasons people sometimes try to sup-
press or control their worrying (e.g., Craske, Rapee, Jackel,
& Barlow, 1989). But research has also shown that attempt-
ing to control thoughts and worry may paradoxically lead
to increased experience of intrusive thoughts and enhanced
perception of being unable to control them (e.g., Abramow-
itz, Tolin, & Street, 2001; Wells, 1999). As we have noted,
perceptions of uncontrollability are known to be associated
with increased anxiety, and further intrusive thoughts can
serve as trigger topics for more worry. Thus a vicious circle
of anxiety, worry, and intrusive thoughts may develop,
leading to this sense of being unable to control worry in
GAD (Mineka, 2004; Mineka et al., 2002; Zinbarg, Craske,
& Barlow, 1994).
In summary, people who have a history of uncontrol-
lable and unpredictable life stress may be especially prone
to developing GAD. Worry about possible bad outcomes or
dangerous events, the central characteristic of GAD, seems
to serve as a cognitive avoidance response that is reinforced
because it suppresses emotional and physiological respond-
ing. Because attempts to suppress or control worry may
lead to more negative intrusive thoughts, perceptions of
uncontrollability over worry may develop, which is in turn
associated with greater anxiety, leading to a vicious cycle.
The central features of obsessive–compulsive disorder
(OCD) are unwanted and intrusive thoughts, impulses, or
images that cause marked anxiety or distress; these are
usually accompanied by compulsive behaviors or mental
rituals that are performed to neutralize or prevent the dis-
tressing thoughts or images. Even “normal” people expe-
rience occasional cognitive intrusions that do not differ in
content from those seen in OCD (e.g., Rachman & de Silva,
1978). What distinguishes people with OCD is that clinical
intrusions/obsessions are (a) associated with greater dis-
tress, (b) more frequent, and (c) more strongly resisted.
Thus, one critical question regarding the development of
OCD is why intrusions are more distressing, frequent, and
strongly resisted in certain people.
Verbal Conditioning and Social Learning
Although research thus far has failed to find good evidence
that a traumatic conditioning history is involved in the
origins of many cases of OCD (Mineka & Zinbarg, 1996;
Rachman & Hodgson, 1980), verbal transmission of dan-
gerous thoughts (as discussed for specific phobias) may
occur (e.g., “my mother told me that dirt and contamination
may be on doorknobs or toilet seats”). In addition, direct
20 January 2006 ● American Psychologist
verbal conditioning may also occur in which a neutral idea
is paired with some scary idea that a person may have (e.g.,
“I saw that filthy person use the toilet and wash her hands
before me so I should be very careful in all bathrooms”). In
addition, there is some evidence to suggest that social
learning factors may play a role in the greater levels of
distress and resistance associated with clinical intrusions.
One idea is that parents (or teachers) who encourage a
broad sense of responsibility and rigid rules for duty and
conduct may create a vulnerability for the development of
OCD in the child (e.g., Salkovskis, Shafran, Rachman, &
Freeston, 1999). Similar factors may lead to beliefs referred
to as thought–action fusion in which people are taught the
idea that thoughts, desires, and impulses are morally equiv-
alent to actions (e.g., “it is just as immoral to think about
harming someone as it is to actually cause harm”) (Shafran,
Thordarson, & Rachman, 1996). A second type of thought–
action fusion involves the idea that thoughts of a specific
catastrophe increase the probability that it will actually
occur (e.g., “thinking about my mother getting cancer
increases the risk that she actually will”). Evidence sug-
gests that each of these factors may lead to greater distress
and attempts to resist intrusions (see Salkovskis et al.,
1999, for several relevant cases).
Social learning is also involved in culturally transmit-
ted beliefs and norms from the teachings of various reli-
gions that may exert influences on the expression of some
obsessions and compulsions. One hypothesis is that highly
religious individuals may be at slightly higher risk for
developing OCD. Although only correlational data are
available, Steketee, Quay, and White (1991) found that
religiosity was significantly correlated with measures of
OCD symptom severity (see also Rassin & Koster, 2003).
Moreover, Rassin and Koster found that religiosity is
strongly (r ? .71) correlated among Protestants with be-
liefs about thought–action fusion (e.g., that thoughts and
desires are morally equivalent to actions) (r ? .42 among
The Role of Avoidance in the Maintenance of
Most patients with OCD engage in either behavioral or
mental rituals intended to neutralize or prevent their obses-
sions, and it is therefore not surprising that rituals have
long been construed as avoidance responses (e.g., Rachman
& Hodgson, 1980). As noted in the section on GAD,
learned avoidance responses are remarkably resistant to
extinction (for reviews, see Mineka, 1979, 1985) and thus
provide a useful analogue to compulsive rituals in humans
with OCD. Moreover, the best way to extinguish avoidance
responses is by preventing them from occurring during
prolonged exposure to the conditioned aversive stimulus;
this procedure is highly analogous to the exposure and
response-prevention procedures long considered to be the
most effective treatment for OCD (e.g., Rachman & Hodg-
son, 1980; Steketee & Barlow, 2002). Mineka and Zinbarg
(1996) also summarized other findings from the avoidance
learning literature that are relevant for understanding the
fluctuations in frequency of occurrence of compulsive rit-
uals as a person goes through varying levels of stress.
Preparedness and OCD
Finally, the preparedness concept discussed earlier in the
context of specific and social phobias is also relevant to
understanding OCD in an evolutionary context. As dem-
onstrated years ago by De Silva, Rachman, and Seligman
(1977), the contents of most obsessions and the forms of
most compulsive rituals are nonrandom (thoughts about
dirt, contamination, and danger are the most common ob-
sessions; cleaning and checking for danger are the most
common rituals). Indeed, humans’ obsessions with dirt,
contamination, and danger have deep evolutionary roots
(Mineka, 1985; Mineka & Zinbarg, 1996). Moreover, the
displacement activities that many animal species engage in
under situations of conflict or high arousal (such as groom-
ing, nesting, or tidying) bear a significant resemblance to
the compulsive rituals seen in OCD. Whether these fears
are based on nontraumatic conditioning (verbal or vicarious
verbal conditioning) or on social learning, the idea is that
people with OCD do not typically obsess about random
scary thoughts and do not show arbitrary ritualistic behav-
iors; rather, their thoughts and behaviors have deep evolu-
In summary, we have argued that verbal learning and
social learning factors are involved in the acquisition of
beliefs that lead to heightened levels of distress and resis-
tance to intrusions; these beliefs tend to center on a few
evolutionarily relevant themes. Moreover, once compul-
sive rituals have developed as methods to neutralize or
prevent obsessive thoughts, they show the same high re-
sistance to extinction seen in learned avoidance responses
The influence of early, rather simplistic learning ap-
proaches to understanding the etiology and course of anx-
iety disorders had begun to wane around the mid-1970s
following some forceful criticisms of such models for
phobias. However, contemporary learning theory and re-
search provide a foundation for the development of models
of the anxiety disorders that can capture the richness and
complexity associated with the development and course of
these disorders (see Figure 1). Not only do contemporary
learning models of the etiology of anxiety disorders have a
great deal of explanatory power but they are also testable.
We have outlined several falsifiable predictions derived
from these models here and elsewhere (e.g., Bas ¸og ˇlu &
Mineka, 1992; Bouton et al., 2001; Foa et al., 1992;
Mineka, 1985; Mineka & Zinbarg, 1996). Some of these
predictions have already been tested and supported. Only
time and additional research will tell if the remaining
predictions are also supported or refuted.
Whereas some of the predictions derived from learn-
ing theory models of the etiology of anxiety disorders
overlap with predictions made by models emanating from
other theoretical perspectives, we believe that the learning
21January 2006 ● American Psychologist
theory models often have the advantage of being more
comprehensive. Consider as an example two psychological
models of PDA. Both the leading cognitive perspective in
which catastrophic cognitions about the consequences of
feared bodily sensations are thought to play a causal role in
panic attacks (e.g., Clark, 1988, 1996) and the learning
theory perspective on etiology outlined here predict the
occurrence of panic attacks in many of the circumstances in
which they tend to occur (see Bouton et al., 2001). How-
ever, the cognitive perspective does not distinguish clearly
between panic and anxiety as partially distinct emotional
states as does the learning theory approach. Therefore the
cognitive model of etiology is silent about the ways in
which anxiety can potentiate panic as outlined previously
and does not do as thorough a job outlining the exact
processes by which occasional panic attacks only some-
times develop into panic disorder. Moreover, the cognitive
model is silent about the variety of different vulnerability
factors that the learning theory approach explicitly ad-
dresses affecting which individuals with panic attacks are
most likely to develop PD or PDA.
Finally, although our primary focus here is on etiol-
ogy, it is also necessary to note that contemporary learning
models of the anxiety disorders have important implica-
tions for treatment and prevention. We start with the old
premise that what can be learned can also be unlearned and
perhaps also prevented. Thus it is not surprising that treat-
ment approaches based on learning principles have been
shown to be effective for each of the anxiety disorders
(e.g., Barlow, 2002; Barlow, Allen, & Coate, 2004; Craske,
1999; Craske & Mystkowski, in press). More specifically,
behavioral/learning research on habituation, extinction,
counterconditioning, and, most recently, safety signals
sowed the seeds for the development of exposure ther-
apy—a central component of many empirically supported
treatment packages for the anxiety disorders as they are
practiced today. In addition, more recent studies on the
important role that contextual shifts have on enhancing the
return of fear following extinction in animals (e.g., Bouton
& Swartzentruber, 1991) have provided new insights into
the similar role that contextual shifts play in the return of
fear in individuals who have undergone exposure treatment
for specific phobias and probably other anxiety disorders as
well, although there is not yet much research from this
perspective (e.g., Craske & Mystkowski, in press; Myst-
kowski & Mineka, in press).
Regarding prevention, by incorporating the role of
genetic and temperamental variables, contemporary learn-
ing models can help identify which individuals are at
highest risk for the development of anxiety disorders. How-
ever, contemporary learning models can also provide a
theoretical and empirical basis for targeting individuals at
risk on the basis of their prior learning histories as well.
Thus, perhaps as a result of both genetic transmission and
vicarious learning of specific fears, as well as avoidant
coping styles and patterns of attentiveness to somatic cues,
children of parents with anxiety disorders seem to be at
elevated risk for the development of anxiety disorders.
Overview of Major Elements Incorporated in Contemporary Learning Models of the Etiology of the Anxiety
Note.US ? unconditioned stimulus.
22 January 2006 ● American Psychologist
Contemporary learning models predict that rearing aimed
at facilitating the development of a strong sense of mastery,
as well as extensive exposure to nonanxious models, from
an early age should provide powerful immunization in such
Detailing how to implement the immunization princi-
ples suggested by contemporary learning models of anxiety
disorders is beyond the scope of the current article (see
Feldner, Zvolensky, & Schmidt, 2004, for a review of the
status of prevention research for anxiety disorders). How-
ever, one example would be how to address the not-un-
common tendency of parents in treatment for anxiety dis-
orders today to voice concerns that their children will
develop similar problems. In discussing such concerns, the
therapist might offer several suggestions that these parents
could implement to lower their children’s risk, in addition
to providing the usual reassurance that many such children
do not go on to develop anxiety disorders. For example, the
therapist might discuss various parenting behaviors and
strategies that should facilitate the development of mastery
and a nonavoidant coping style. In addition, the therapists
might suggest to parents with phobias that they ensure that
their child has extensive direct exposure to nonfearful
models interacting with the phobic object.
It is certainly the case that much work remains to be
done on the etiology, course, treatment, and prevention of
anxiety disorders. We have focused here on the implica-
tions for etiology and course of anxiety disorders because
of space constraints and have only briefly noted implica-
tions for treatment and prevention. It should be clear,
however, that we believe contemporary learning theory has
much to offer in guiding the efforts to solve the remaining
mysteries associated with anxiety disorders.
American Psychiatric Association. (1980). Diagnostic and statistical
manual of mental disorders (3rd ed.). Washington, DC: Author.
American Psychiatric Association. (1994). Diagnostic and statistical
manual of mental disorders (4th ed.). Washington, DC: Author.
Abramowitz, J. S., Tolin, D. F., & Street, G. P. (2001). Paradoxical effects
of thought suppression: A meta-analysis of controlled studies. Clinical
Psychology Review, 21, 683–703.
Abramson, L. Y., & Seligman, M. E. P. (1977). Modeling psychopathol-
ogy in the laboratory: History and rationale. In J. P. Maser & M. E. P.
Seligman (Eds.), Psychopathology: Experimental models (pp. 1–26).
San Francisco: Freeman.
Barlow, D. H. (1988). Anxiety and its disorders: The nature and treatment
of anxiety and panic. New York: Guilford Press.
Barlow, D. H. (2002). Anxiety and its disorders: The nature and treatment
of anxiety and panic (2nd ed.). New York: Guilford Press.
Barlow, D. H., Allen, L. B., & Coate, M. L. (2004). Toward a unified
treatment of emotional disorders. Behavior Therapy, 35, 205–230.
Barrett, P., Rapee, R., Dadds, M., & Ryan, S. (1996). Family enhancement
of cognitive style in anxious and aggressive children. Journal of Ab-
normal Child Psychology, 24, 187–203.
Bas ¸og ˇlu, M., Marks, I., & Sengu ¨n, S. (1992). A prospective study of panic
and anxiety in agoraphobia with panic disorder. British Journal of
Psychiatry, 160, 57–64.
Bas ¸og ˇlu, M., & Mineka, S. (1992). The role of uncontrollability and
unpredictability of stress in the development of post-torture stress
symptoms. In M. Bas ¸og ˇlu (Ed.), Torture and its consequences: Current
treatment approaches (pp. 182–225). Cambridge, England: Cambridge
Bas ¸og ˇlu, M., Mineka, S., Paker, M., Aker, T., Livanou, M., & Go ¨k, S.
(1997). Psychological preparedness for trauma as a protective factor in
survivors of torture. Psychological Medicine, 27, 1421–1433.
Beck, A. T., & Emery, G. (1985). Anxiety disorders and phobias: A
cognitive perspective. New York: Basic Books.
Biederman, J., Rosenbaum, J. F., Hirshfeld, D. R., Farone, S. V., Bolduc,
E. A., Gersten, M., et al. (1990). Psychiatric correlates of behavioral
inhibition in young children of parents with and without psychiatric
disorders. Archives of General Psychiatry, 47, 21–26.
Bolstad, B., & Zinbarg, R. (1997). Sexual victimization, generalized
perception of control, and posttraumatic stress disorder symptom se-
verity. Journal of Anxiety Disorders, 11, 523–540.
Borkovec, T. D. (1994). The nature, functions and origins of worry. In
G. C. Davey & F. Tallis (Eds.), Worrying: Perspectives on theory,
assessment and treatment (pp. 5–33). New York: Wiley.
Borkovec, T. D., Alcaine, O., & Behar, E. (2004). Avoidance theory of
worry and generalized anxiety disorder. In R. G. Heimberg, C. L. Turk,
& D. S. Mennin (Eds.), Generalized anxiety disorder: Advances in
research and practice (pp. 77–108). New York: Guilford Press.
Bouton, M., Mineka, S., & Barlow, D. H. (2001). A contemporary
learning theory perspective on the etiology of panic disorder. Psycho-
logical Review, 108, 4–32.
Bouton, M. E., & Swartzentruber, D. (1991). Sources of relapse after
extinction in Pavlovian and instrumental learning. Clinical Psychology
Review, 11, 123–140.
Bruch, M., & Heimberg, R. (1994). Differences in perceptions of parental
and personal characteristics between generalized and nongeneralized
social phobics. Journal of Anxiety Disorders, 8, 155–168.
Carter, M. M., Hollon, S. D., Carson, R., & Shelton, R. C. (1995). Effects
of a safe person on induced distress following a biological challenge in
panic disorder with agoraphobia. Journal of Abnormal Psychology,
Chambless, D. L., & Mason, J. (1986). Sex, sex role stereotyping, and
agoraphobia. Behaviour Research and Therapy, 24, 231–235.
Chorpita, B. F., & Barlow, D. H. (1998). The development of anxiety: The
role of control in the early environment. Psychological Bulletin, 124,
Clark, D. M. (1988). A cognitive model of panic attacks. In S. Rachman
& J. D. Maser (Eds.), Panic: Psychological perspectives (pp. 71–89).
Hillsdale, NJ: Erlbaum.
Clark, D. M. (1996). Panic disorder: From theory to therapy. In P. M.
Salkovskis (Ed.), Frontiers of cognitive therapy (pp. 318–344). New
York: Guilford Press.
Cook, E., Hodes, R., & Lang, P. (1986). Preparedness and phobia: Effects
of stimulus content on human visceral conditioning. Journal of Abnor-
mal Psychology, 95, 195–207.
Cook, M., & Mineka, S. (1989). Observational conditioning of fear to
fear-relevant versus fear-irrelevant stimuli in rhesus monkeys. Journal
of Abnormal Psychology, 98, 448–459.
Cook, M., & Mineka, S. (1990). Selective associations in the observa-
tional conditioning of fear in monkeys. Journal of Experimental Psy-
chology: Animal Behavior Processes, 16, 372–389.
Cook, M., Mineka, S., Wolkenstein, B., & Laitsch, K. (1985). Observa-
tional conditioning of snake fear in unrelated rhesus monkeys. Journal
of Abnormal Psychology, 94, 591–610.
Craske, M. G. (1999). Anxiety disorders: Psychological approaches to
theory and treatment. Boulder, CO: Westview Press.
Craske, M. G., Glover, D., & DeCola, J. (1995). Predicted versus unpre-
dicted panic attacks: Acute versus general distress. Journal of Abnor-
mal Psychology, 104, 708–718.
Craske, M. G., & Mystkowski, J. (in press). Exposure therapy and ex-
tinction: Clinical studies. In M. Craske & D. Hermans (Eds.), Fear and
learning: Contemporary perspectives. Washington, DC: American Psy-
Craske, M. G., Rapee, R. M., Jackel, L., & Barlow, D. H. (1989).
Qualitative dimensions of worry in DSM–III–R generalized anxiety
disorder subjects and nonanxious controls. Behaviour Research and
Therapy, 27, 397–402.
Dadds, M., Barrett, P., Rapee, R., & Ryan, S. (1996). Family process and
child anxiety and aggression: An observational analysis. Journal of
Abnormal Child Psychology, 24, 715–734.
Davey, G. C. L. (1997). A conditioning model of phobias. In G. C. L.
23January 2006 ● American Psychologist
Davey (Ed.), Phobias: A handbook of theory, research, and treatment
(pp. 301–322). Chichester, England: Wiley.
Davey, G. C. L., & Matchett, G. (1994). Unconditioned stimulus rehearsal
and the retention and enhancement of differential “fear” conditioning:
Effects of trait and state anxiety. Journal of Abnormal Psychology, 103,
De Silva, P., Rachman, S., & Seligman, M. E. P. (1977). Prepared phobias
and obsessions: Therapeutic outcome. Behaviour Research and Ther-
apy, 22, 87–91.
Dugas, M. J., Buhr, K., & Ladouceur, R. (2004). The role of intolerance
of uncertainty in etiology and maintenance. In R. G. Heimberg, C. L.
Turk, & D. S. Mennin (Eds.), Generalized anxiety disorder: Advances
in research and practice. New York: Guilford Press.
Dunmore, E., Clark, D. M., & Ehlers, A. (2001). A prospective investi-
gation of the role of cognitive factors in persistent posttraumatic stress
disorder (PTSD) after physical or sexual assault. Behaviour Research
and Therapy, 39, 1063–1084.
Dworkin, B. R. (1993). Learning and physiological regulation. Chicago:
University of Chicago Press.
Ehlers, A. (1993). Somatic symptoms and panic attacks: A retrospective
study of learning experiences. Behaviour Research and Therapy, 31,
Ehlers, A., Maercker, A., & Boos, A. (2000). Posttraumatic stress disorder
following political imprisonment: The role of mental defeat, alienation,
and perceived permanent change. Journal of Abnormal Psychology,
Ehlers, A., Mayou, R., & Bryant, B. (1998). Psychological predictors of
chronic posttraumatic stress disorder after motor vehicle accidents.
Journal of Abnormal Psychology, 107, 508–519.
Esteves, F., Parra, C., Dimberg, U., & O¨hman, A. (1994). Nonconscious
associative learning: Pavlovian conditioning of skin conductance re-
sponses to masked fear relevant facial stimuli. Psychophysiology, 31,
Fanselow, M. S. (1994). Neural organization of the defensive behavior
system responsible for fear. Psychonomic Bulletin & Review, 1, 429–
Feldner, M. T., Zvolensky, M. J., & Schmidt, N. B. (2004). Prevention of
anxiety psychopathology: A critical review of the empirical literature.
Clinical Psychology: Science and Practice, 11, 405–424.
Foa, E. B., Zinbarg, R., & Rothbaum, B. O. (1992). Uncontrollability and
unpredictability in posttraumatic stress disorder: An animal model.
Psychological Bulletin, 112, 218–238.
Forsyth, J. P., & Eifert, G. H. (1996a). Cleaning-up cognition in triple-
response fear assessment through individualized functional behavior
analysis. Journal of Behavior Therapy and Experimental Psychiatry,
Forsyth, J. P., & Eifert, G. H. (1996b). Systemic alarms in fear condi-
tioning I: A reappraisal of what is being conditioned. Behavior Ther-
apy, 27, 441–462.
Goldstein, A. J., & Chambless, D. L. (1978). A reanalysis of agoraphobia.
Behavior Therapy, 9, 47–59.
Gray, J. A. (1982). The neuropsychology of anxiety: An enquiry into the
functioning of the septo-hippocampal system. Oxford, England: Oxford
Hayward, C., Killen, J., Kraemer, H., & Taylor, C. B. (1998). Linking
self-reported childhood behavioral inhibition to adolescent social pho-
bia. Journal of the American Academy of Child & Adolescent Psychi-
atry, 37, 1308–1316.
Hayward, C., Killen, J. D., Kraemer, H. C., & Taylor, C. B. (2000).
Predictors of panic attacks in adolescence. Journal of the American
Academy of Child & Adolescent Psychiatry, 39, 207–214.
Hebert, M., Evenson, A., Lumley, L., & Meyerhoff, J. (1998). Effects of
acute social defeat on activity in the forced swim test: Parametric
studies in DBA/2 mice using a novel measurement device. Aggressive
Behavior, 24, 257–269.
Hettema, J. M., Annas, P., Neale, M. C., Kendler, K. S., & Fredrikson, M.
(2003). A twin study of the genetics of fear conditioning. Archives of
General Psychiatry, 60, 702–707.
Kagan, J. (1994). Galen’s prophecy. New York: Basic Books.
Kagan, J. (1997). Temperament and the reactions to unfamiliarity. Child
Development, 68, 139–143.
Kendler, K. S., Neale, M. C., Kessler, R. C., Heath, A. C., & Eaves, L. J.
(1992). The genetic epidemiology of phobias in women: The interre-
lationship of agoraphobia, social phobia, situational phobia, and simple
phobia. Archives of General Psychiatry, 49, 273–281.
Kendler, K. S., Walters, E. E., Neale, M. C., Kessler, R. C., Heath, A., &
Eaves, L. J. (1995). The structure of the genetic and environmental risk
factors for six major psychiatric disorders in women: Phobia, general-
ized anxiety disorder, panic disorder, bulimia, major depression, and
alcoholism. Archives of General Psychiatry, 52, 374–382.
Kent, G. (1997). Dental phobias. In G. C. Davey (Ed.), Phobias: A
handbook of theory, research and treatment (pp. 107–127). Chichester,
Kessler, R. C., Berglund, P., Demler, O., Jin, R., & Walters, E. E. (2005).
Lifetime prevalence and age-of-onset distributions of DSM–IV disor-
ders in the National Comorbidity Survey Replication. Archives of
General Psychiatry, 62, 593–602.
Kim, J. A., Siegel, S., & Patenall, V. R. A. (1999). Drug-onset cues as
signals: Intraadministration associations and tolerance. Journal of Ex-
perimental Psychology: Animal Behavior Processes, 25, 491–504.
Kirmayer, L. J. (1991). The place of culture in psychiatric nosology:
Taijin kyofusho and DSM–III–R. Journal of Nervous and Mental Dis-
ease, 179, 19–28.
Lang, P. J. (1995). The emotion probe: Studies of motivation and atten-
tion. American Psychologist, 50, 372–385.
Leung, A. W., & Heimberg, R. G. (1996). Homework compliance, per-
ceptions of control, and outcome of cognitive–behavioral treatment of
social phobia. Behaviour Research and Therapy, 34, 423–432.
Levey, A., & Martin, I. (1981). Personality and conditioning. In H.
Eysenck (Ed.), A model for personality (pp. 123–168). Berlin, West
Lubow, R. E. (1998). Latent inhibition and behavior pathology: Prophy-
lactic and other possible effects of stimulus preexposure. In W.
O’Donohue (Ed.), Learning and behavior therapy (pp. 107–121).
Needham Heights, MA: Allyn & Bacon.
Maier, S. F. (2001). Exposure to the stressor environment prevents the
temporal dissipation of behavioral depression/learned helplessness. Bi-
ological Psychiatry, 49, 763–773.
McCabe, R., Antony, M., Summerfeldt, L., Liss, A., & Swinson, R.
(2003). Preliminary examination of the relationship between anxiety
disorders in adults and self-reported history of teasing or bullying
experiences. Cognitive Behaviour Therapy, 32, 187–193.
McFarlane, A. (1992). Posttraumatic stress disorder among injured sur-
vivors of a terrorist attack: Predictive value of early intrusion and
avoidance symptoms. Journal of Nervous and Mental Disease, 180,
McKinney, W. T. (1974). Animal models in psychiatry. Perspectives in
biology and Medicine, 17, 529–541.
McNally, R. J. (1994). Panic disorder: A critical analysis. New York:
Michels, R., Frances, A., & Shear, K. (1985). Psychodynamic models of
anxiety. In J. Maser & A. Tuma (Eds.), Anxiety and the anxiety
disorders (pp. 595–618). Hillsdale, NJ: Erlbaum.
Mineka, S. (1979). The role of fear in theories of avoidance learning,
flooding, and extinction. Psychological Bulletin, 86, 986–1010.
Mineka, S. (1985). Animal models of anxiety-based disorders: Their
usefulness and limitations. In J. Maser & A. Tuma (Eds.), Anxiety and
the anxiety disorders (pp. 199–244). Hillsdale, NJ: Erlbaum.
Mineka, S. (2002). Animal models of clinical psychology. In N. Smelser
& P. Baltes (Eds.), International encylopedia of the social and behav-
ioral sciences (pp. 2020–2025). Oxford, England: Elsevier Science.
Mineka, S. (2004). The positive and negative consequences of worry in
the etiology of generalized anxiety disorder: A vicious circle? In J.
Yiend (Ed.), Cognition, emotion and psychopathology: Theoretical,
empirical and clinical directions (pp. 29–48). Cambridge, England:
Cambridge University Press.
Mineka, S., & Cook, M. (1986). Immunization against the observational
conditioning of snake fear in rhesus monkeys. Journal of Abnormal
Psychology, 95, 307–318.
Mineka, S., Cook, M., & Miller, S. (1984). Fear conditioned with escap-
able and inescapable shock: The effects of a feedback stimulus. Journal
of Experimental Psychology: Animal Behavior Processes, 10, 307–323.
Mineka, S., Gunnar, M., & Champoux, M. (1986). Control and early
socioemotional development: Infant rhesus monkeys reared in control-
24January 2006 ● American Psychologist
lable versus uncontrollable environments. Child Development, 57,
Mineka, S., & O¨hman, A. (2002a). Born to fear: Non-associative versus
associative factors in the etiology of phobias. Behaviour Research and
Therapy, 40, 173–184.
Mineka, S., & O¨hman, A. (2002b). Phobias and preparedness: The selec-
tive, automatic and encapsulated nature of fear. Biological Psychiatry,
Mineka, S., & Sutton, J. (in press). Contemporary learning theory per-
spectives on the etiology of fears and phobias. In M. Craske & D.
Hermans (Eds.), Fear and learning: Contemporary perspectives. Wash-
ington, DC: American Psychological Association.
Mineka, S., Yovel, I., & Pineles, S. (2002). Toward a psychological model
of the etiology of generalized anxiety disorder. In D. J. Nutt, K. Rickels,
& D. J. Stein (Eds.), Generalized anxiety disorder: Symptomatology,
pathogenesis and management. London: Martin Dunitz.
Mineka, S., & Zinbarg, R. (1991). Animal models of psychopathology. In
C. E. Walker (Ed.), Clinical psychology: Historical and research
foundations (pp. 51–86). New York: Plenum Press.
Mineka, S., & Zinbarg, R. (1996). Conditioning and ethological models of
anxiety disorders. In D.A. Hope (Ed.), Nebraska Symposium on Moti-
vation: Vol.43. Perspectives on anxiety, panic, and fear: Current theory
and research in motivation (pp. 135–210). Lincoln: University of
Moye, T., Hyson, R., Grau, J., & Maier. S. (1983). Immunization of
opioid analgesia: Effects of prior escapable shock on subsequent shock-
induced and morphine-induced antinociception. Learning and Motiva-
tion, 4, 238–251.
Muris, P., & Mercklebach, H. (2001). The etiology of childhood specific
phobia: A multifactorial model. In M. Vasey & Dadds, M. (Eds.), The
developmental psychopathology of anxiety (pp. 355–385). New York:
Oxford University Press.
Mystkowski, J., & Mineka, S. (in press). Behavior therapy for fears and
phobias: Context specificity of fear extinction. In T. B. Baker, R.
Bootzin, & T. Treat (Eds.), Psychological clinical science: Recent
advances in theory and practice: Integrative perspectives in honor of
Richard M. McFall. Hillsdale, NJ: Erlbaum.
O¨hman, A., & Dimberg, U. (1978). Facial expressions as conditioned
stimuli for electrodermal responses: A case of “preparedness”? Journal
of Personality and Social Psychology, 36, 1251–1258.
O¨hman, A., Dimberg, U., & O¨st, L-G. (1985). Animal and social phobias:
Biological constraints on the learned fear response. In S. Reiss & R.
Bootzin (Eds.), Theoretical issues in behavior therapy (pp. 123–175).
New York: Academic Press.
O¨hman, A., & Mineka, S. (2001). Fears, phobias, and preparedness:
Toward an evolved module of fear learning. Psychological Review,
O¨hman, A., & Soares, J. (1993). On the automatic nature of phobic fear:
Conditioned electrodermal responses to masked fear-relevant stimuli.
Journal of Abnormal Psychology, 102, 121–132.
O¨hman, A., & Soares, J. (1998). Emotional conditioning to masked
stimuli: Expectancies for aversive outcomes following nonconsciously
recognized fear-relevant stimuli. Journal of Experimental Psychology:
General, 127, 69–82.
O¨st, L-G., & Hugdahl, K. (1981). Acquisition of phobias and anxiety
response patterns in clinical patients. Behaviour Research and Therapy,
Ozer, E., Best, S., Lipsey, T., & Weiss, D. (2003). Predictors of posttrau-
matic stress disorder and symptoms in adults: A meta-analysis. Psy-
chological Bulletin, 129, 52–73.
Pavlov, I. P. (1927). Conditioned reflexes. London: Oxford University
Pitman, R., Van der Kolk, B., Orr, S., & Greenberg, M. (1990). Naloxone-
reversible analgesic response to combat-related stimuli in posttraumatic
stress disorder: A pilot study. Archives of General Psychiatry, 47,
Poulton, R., & Menzies, R. (2002). Non-associative fear acquisition: A
review of the evidence from retrospective and longitudinal research.
Behaviour Research and Therapy, 40, 127–149.
Rachman, S. (1978). Fear and courage. San Francisco: Freeman.
Rachman, S. (1990). Fear and courage (2nd ed.). New York: Freeman.
Rachman, S., & de Silva, P. (1978). Abnormal and normal obsessions.
Behaviour Research and Therapy, 16, 411–423.
Rachman, S., & Hodgson, R. S. (1980). Obsessions and compulsions.
Englewood Cliffs, NJ: Prentice Hall.
Rapee, R. (2001). The development of generalized anxiety. In M. Vasey
& M. Dadds (Eds.), The developmental psychopathology of anxiety (pp.
481–503). Oxford, England: Oxford University Press.
Rapee, R. & Melville, L. (1997). Recall of family factors on social phobia
and panic disorder: Comparison of mother and offspring reports. De-
pression and Anxiety, 5, 7–11.
Rassin, E., & Koster, E. (2003). The correlation of thought–action fusion
and religiosity in a normal sample. Behaviour Research and Therapy,
Razran, G. (1961). The observable unconscious and the inferable con-
scious in current Soviet psychophysiology: Interoceptive conditioning,
semantic conditioning, and the orienting reflex. Psychological Review,
Regehr, C., Cadell, S., & Jansen, K. (1999). Perceptions of control and
long-term recovery from rape. American Journal of Orthopsychiatry,
Rescorla, R. (1974). Effect of inflation of the unconditioned stimulus
value following conditioning. Journal of Comparative and Physiolog-
ical Psychology, 86, 101–106.
Rescorla, R. A. (1988). Pavlovian conditioning: It’s not what you think it
is. American Psychologist, 43, 151–160.
Rescorla, R., & Heth, C. (1975). Reinstatement of fear to an extinguished
conditioned stimulus. Journal of Experimental Psychology: Animal
Behavior Processes, 1, 88–96.
Riggs, D., Rothbaum, B. O., & Foa, E. (1995). A prospective examination
of symptoms of posttraumatic stress disorder in vicitms of nonsexual
assault. Journal of Interpersonal Violence, 10, 201–214.
Rodgers, R. J., & Hendrie, C. A. (1983). Social conflict activates status-
dependent endogenous analgesic and hyperalgesic mechanisms in male
mice. Physiology & Behavior, 30, 775–780.
Roemer, L., Orsillo, S., & Barlow, D. H. (2002). Generalized anxiety
disorder. In D. H. Barlow (Ed.), Anxiety and its disorders (pp. 477–
515). New York: Guilford Press.
Rothbaum, B. O., Foa, E., Riggs, D., Murdock, T., & Walsh, W. (1992).
A prospective examination of post-traumatic stress disorder in rape
victims. Journal of Traumatic Stress, 5, 455–475.
Rush, D. K., Mineka, S., & Suomi, S. J. (1982). The effects of control and
lack of control on active and passive avoidance in rhesus monkeys.
Behaviour Research and Therapy, 20, 135–152.
Salkovskis, P., Shafran, R., Rachman, S., & Freeston, M. (1999). Multiple
pathways to inflated responsibility beliefs in obsessional problems:
Possible origins and implications for therapy and research. Behaviour
Research and Therapy, 37, 1055–1072.
Seligman, M. (1971). Phobias and preparedness. Behavior Therapy, 2,
Shafran, R., Thordarson, D., & Rachman, S. (1996). Thought–action
fusion in obsessive–compulsive disorder. Journal of Anxiety Disorders,
Steketee, G., & Barlow, D. H. (2002). Obsessive-compulsive disorder. In
D. H. Barlow (Ed), Anxiety and its disorders (pp. 516–550). New York:
Steketee, G., Quay, S., & White, K. (1991). Religion and guilt in OCD
patients. Journal of Anxiety Disorders, 5, 359–367.
Stemberger, R., Turner, S., Beidel, D., & Calhoun, K. (1995). Social
phobia: An analysis of possible developmental factors. Journal of
Abnormal Psychology, 104, 526–531.
Tanaka-Matsumi, J., Seiden, D. Y., & Lam, K. N. (1996). The culturally
informed functional assessment (CIFA) interview: A strategy for cross-
cultural behavioral practice. Cognitive and Behavioral Practice, 3,
True, W. R., Rice, J., Eisen, S. A., Heath, A. C., Goldberg, J., Lyons,
M. J., et al. (1993). A twin study of genetic and environmental contri-
butions to liability for posttraumatic stress symptoms. Archives of
General Psychiatry, 50, 257–264.
Uhrich, J. (1938). The social hierarchy in albino mice. Journal of Com-
parative Psychology, 25, 373–413.
Watson, J. B., & Rayner, R. (1920). Conditioned emotional reactions.
Journal of Experimental Psychology, 3, 1–14.
25 January 2006 ● American Psychologist
Weiss, J. M., Glazer, H. I., & Pohorecky, L. A. (1976). Coping behavior Download full-text
and neurochemical changes: An alternative explanation for the original
“learned helplessness” experiments. In A. Serban & A. Kling (Eds.),
Animal models in human psychobiology (pp. 141–173). New York:
Wells, A. (1999). A cognitive model of generalized anxiety disorder.
Behavior Modification, 23, 526–555.
Wells, A., & Papageorgiou, C. (1995). Worry and the incubation of
intrusive images following stress. Behaviour Research and Therapy,
White, K., & Davey, G. (1989). Sensory preconditioning and US inflation
in human “fear” conditioning. Behaviour Research and Therapy, 27,
Williams, J. L., & Lierle, D. M. (1986). Effects of stress controllability,
immunization, and therapy on the subsequent defeat of colony intrud-
ers. Animal Learning & Behavior, 14, 305–314.
Williams, J., & Maier, S. (1977). Transsituational immunization and
therapy of learned helplessness in the rat. Journal of Experimental
Psychology: Animal Behavior Processes, 3, 240–252.
Williams, J. L., & Scott, D. K. (1989). Influence of conspecific and
predatory stressors and the associated odors on defensive burying and
freezing. Animal Learning & Behavior, 17, 383–393.
Zinbarg, R. E., Barlow, D. H., Brown, T. A., & Hertz, R. M. (1992).
Cognitive–behavioral approaches to the nature and treatment of anxiety
disorders. Annual Review of Psychology, 43, 235–267.
Zinbarg, R., Craske, M., & Barlow, D. H. (1994). Therapist’s guide for
the mastery of your anxiety and worry program. Albany, NY: Gray-
Zinbarg, R., & Mohlman, J. (1998). Individual differences in the acqui-
sition of affectively-valenced associations. Journal of Personality and
Social Psychology, 74, 1024–1040.
26January 2006 ● American Psychologist