Journal of Personality and Social Psychology

1979, Vol. 37, No. 10, 1798-1809

Characterological Versus Behavioral Self-Blame: Inquiries Into

Depression and Rape
Ronnie Janoff-Bulman
University of MassachusettsAmherst
Two types of self-blamebehavioral and characterologicalare distinguished.
Behavioral self-blame is control related, involves attributions to a modifiable
source (one's behavior), and is associated with a belief in the future avoidability of a negative outcome. Characterological self-blame is esteem related,
involves attributions to a relatively nonmodifiable source (one's character), and
is associated with a belief in personal deservingness for past negative outcomes.
Two studies are reported that bear on this self-blame distinction. In the first
study, it was found that depressed female college students engaged in more
characterological self-blame than nondepressed female college students, whereas
behavioral self-blame did not differ between the two groups; the depressed
population was also characterized by greater attributions to chance and decreased beliefs in personal control. Characterological self-blame is proposed as
a possible solution to the "paradox in depression." In a second study, rape
crisis centers were surveyed. Behavioral self-blame, and not characterological
self-blame, emerged as the most common response of rape victims to their
victimization, suggesting the victim's desire to maintain a belief in control,
particularly the belief in the future avoidability of rape. Implications of this
self-blame distinction and potential directions for future research are discussed.

In a study by Bulman and Wortman

(1977) on the relationship between blame
attributions and coping, self-blame emerged
as a predictor of good coping among paralyzed victims of freak accidents. A conclusion
that is consistent with these resultsthat
self-blame is a positive psychological mechanismderives primarily from the implications of this attribution for a belief in personal control over one's outcomes. The advantages of perceived control have been repeatedly demonstrated in social psychological
experiments (see, e.g., Bowers, 1968; Glass &
Singer, 1972; Langer & Rodin, 1976; Schulz,
1976); Walster's (1966) formulation of observers' reactions to victims and Kelley's
(1971) view of attributional processes "as a
The author thanks Philip Brickman, Irene Frieze,
and Camille Wortman for their valuable comments on an earlier draft of this article.
Requests for reprints should be sent to Ronnie
Janoff-Bulman, University of Massachusetts, Department of Psychology, Amherst, Massachusetts
01003.

means of encouraging and maintaining [his]

effective exercise of control in the world"
(p. 22) are also based upon a recognition of
the significance of perceived control. The
tenuous link between control and self-blame
becomes comprehensible as one realizes that
in order to maximize a belief in control when
attributing blame to particular factors, one's
choice is influenced by the perceived modifiability of the potential factor(s). As Medea
and Thompson (1974) write in the case of
rape, "If the woman can believe that somehow she got herself into the situation, if she
can make herself responsible for it, then she's
established some sort of control over rape.
It wasn't someone arbitrarily smashing into
her life and wreaking havoc" (p. 105).
Unfortunately, this adaptive, control-oriented view of self-blame too easily ignores
the more popular conception of the phenomenon, by which self-blame is regarded as maladaptive, a correlate of depression, and a
reflection of psychological problems. Thus
Beck (1967), writing about depressed pa-

Copyright 1979 by the American Psychological Association, Inc. 0022-3514/79/3710-1798$00.7S

1798

SELF-BLAME: DEPRESSION AND RAPE

tients, states, "Another symptom, self-blame,

expresses the patient's notion of causality.
He is prone to hold himself responsible for
any difficulties or problems that he encounters" (p. 21). Self-blame as a maladaptive
psychological mechanism is generally related
to harsh self-criticism and low evaluations of
one's worth.
Two Types oj Self-Blame
Recognizing that self-blame may be both
adaptive and maladaptive is a first step towards the conclusion that there are two
different types of self-blame, one representing
an adaptive, control-oriented response, the
other a maladaptive, self-deprecating response. The primary distinction between
these two self-attributions is the nature of
the focus of blame, for it is proposed that
the control related self-blame focuses on one's
own behavior, whereas the esteem related
self-blame focuses on one's character, an
overall view of the kind of people individuals
perceive themselves to be. In other words,
individuals can blame themselves for having
engaged in (or having failed to engage in) a
particular activity, thereby attributing blame
to past behaviors; or individuals can blame
themselves for the kind of people they are,
thereby faulting their character. To facilitate
discussion of these two self-attributional
strategies, the esteem related blame will be
labeled "characterological" self-blame and
the control related type, "behavioral" selfblame. In the case of rape, for example, a
woman can blame herself for having walked
down a street alone at night or for having
let a particular man into her apartment (behavioral blame), or she can blame herself for
being "too trusting and unable to say no"
or a "careless person who is unable to stay
out of trouble." This behavioral-characterological distinction parallels findings in the
area of the just world theory. In their recent
review, Lerner and Miller (1978) state
that innocent victims who cannot be characterologically blamed (i.e., derogated) by
virtue of their reputedly good character are
instead blamed for some behavior in which
they engaged (i.e., behavioral blame).
While this distinction between charactero-

1799

logical and behavioral self-blame appears

related to the state-trait distinction in clinical psychology (see, e.g., Spielberger, 1972),
it more specifically corresponds to the distinctions drawn by Weiner and his colleagues
(Weiner et al., 1971) in their scheme of
attributions in the area of achievement. In
attributing failure to oneself (internal attribution), one can point to his/her own lack
of ability or effort, attributions that have
very different implications for perceived control. Individuals who make an attribution
to poor ability believe that there is little
they can do to control the situation and succeed, for ability is stable and relatively unchangeable. Effort attributions, on the other
hand, will lead one to believe that as long
as he/she tries harder, he/she will be able to
control outcomes in a positive manner (see
Dweck, 1975). Similarly, characterological
self-blame corresponds to an ability attribution, and behavioral self-blame corresponds
to an effort attribution, having very different
implications for perceived personal control.
While the dimension used by Weiner and his
colleagues to distinguish between ability and
effort is that of stability (stable-unstable),
the differences between the attributions may
also be captured through the use of a controllability dimension (cf. Elig & Frieze,
1975; Weiner, 1974). The primary distinction to be drawn between behavioral and
characterological self-blame is the perceived
controllability (i.e., modifiability through
one's own efforts) of the factor(s) blamed.
In a recent reformulation of learned helplessness, Abramson, Seligman, and Teasdale
(1978) have posited a third dimension of attributionsglobal-specificthat is important
to specify in determining subsequent perceived helplessness. While this global-specific
dimension characterizes one of the differences
to be noted between characterological and
behavioral self-blame, it is proposed that the
dimension of significance distinguishing these
two types of self-blame is perceived controllability, and the globalspecific and
stable-unstable dimensions are important because of their contribution to perceived control. Abramson, Seligman, and Teasdale
(1978), however, note that the dimension of
"controllability is logically orthogonal to the

1800

RONNIE JANOFF-BULMAN

Internal X Global X Stable dimensions . . ."

(p. 62). The position presented here is consistent with a comment by Wortman and
Dintzer (1978) in their recent reaction to
the learned helplessness reformulation. They
state, "We feel that assessments of the controllability of the causal factor may be of
the utmost importance in predicting the nature and magnitude of subsequent deficits"
(p. 82).
In a discussion of self-blame by Abramson,
Seligman, and Teasdale (1978) these authors
state that self-blame in helplessness and depression follows from the "attribution of
failure to factors that are controllable" (p.
62). The self-blame they are dealing with is
that which is consistent with "self-esteem
deficits" and "self-criticism" and thus parallels characterological self-blame. The authors
do not recognize a second type of self-blame,
behavioral self-blame, which according to
the present analysis is the type of self-blame
following from attributions to controllable
factors. Contrary to the assertions of Abramson, Seligman, and Teasdale, it is proposed
that characterological self-blame follows from
attributions to uncontrollable factors.
A further distinction between behavioral
and characterological self-blame lies in the
time orientation of the attributor. It is proposed that in blaming one's behavior, an
individual is concerned with the future, particularly the future avoidability of the negative outcome. This concern for future avoidability is consistent with the control-motivated basis for behavioral self-blame. The
future-oriented concerns of behavioral selfblamers need not focus exclusively on the
future avoidability of the negative outcome
for which the attributor is blaming him/herself; rather, behavioral self-blame may promote a general belief in one's ability to avoid
negative outcomes and to effect positive outcomes in the future. Thus, the paralyzed
victims in the Bulman and Wortman (1977)
study were apt to be better copers if they
blamed themselves, but self-blame was more
likely to be in the service of a general belief
in future control (e.g., I'll be able to improve my physical condition through physical therapy), rather than a more specific
belief in the future avoidability of their own

paralysis, which was medically regarded as

irreversible in all cases.
In blaming himself or herself characterologically, the individual is not concerned
with control in the future, but rather with
the past, particularly deservingness for past
outcomes. Individuals who engage in behavioral self-blame are apt to have an eye
towards the future and what they can do
to avoid a recurrence of the negative outcome (or the occurrence of negative outcomes in general). Individuals who engage
in characterological self-blame are apt to
focus more on the past and what it was about
them that rendered them deserving of the
negative outcome for which they are blaming
themselves.1 Perceived avoidability and behavioral self-blame are thus assumed to be
part of the same blame cluster, whereas characterological self-blame and feelings of deservingness are representative of another
blame cluster.
Selj-Blame and Depression;
Toward the Resolution of a Paradox
Distinguishing between characterological
and behavioral self-blame may be a first step
toward resolving the "paradox in depression"
recently recognized and discussed by Abramson and Sackeim (1977). According to these
authors, there are two symptom clusters of
depression, one represented by hopelessness,
powerlessness, and futility, the other by selfblame, self-deprecation, and guilt. Abramson
and Sackeim discuss two prominent theories
of depression that are based on cognitions
of hopelessness and self-blame. Seligman's
(1975) theory of learned helplessness suggests that depression results from a belief
in the uncontrollability of outcomes. According to Beck's (1967) theory of depression,
the depressed individual blames him/herself
for negative outcomes, particularly personal
failures. It is the conjunction of these two
models that accounts for the paradox. That
1
These distinctions are consistent with a recent
analysis of responsibility by Harvey and Rule
(1978), in which causal responsibility and deservingness are regarded as conceptually distinct aspects
of responsibility.

SELF-BLAME: DEPRESSION AND RAPE

is, how can individuals blame themselves

for outcomes over which they feel they have
had no control? How can an individual feel
both helpless and self-blaming? Abramson
and Sackeim discuss several possible resolutions to this paradox but remain dissatisfied
with the alternatives presented to date. However, a recognition of self-blame, not as a
unitary phenomenon, but rather as a label
for two very different self-attributional strategies, may inform and resolve the apparent
paradox in depression.
One reason why a resolution to the depression paradox has not been forthcoming
is suggested by Abramson, Seligman, and
Teasdale's
(1978) assertion (presented
above) that self-blame follows from attributions to controllable factors. In assuming that
self-blame naturally involves blaming controllable factors, the possibility that individuals can simultaenously feel they do not have
control and blame themselves is foreclosed.
Instead, if we recognize the distinction between behavioral and characterological selfblame, then the paradox begins to disappear.
Essential to an understanding of this assertion is the proposition that in blaming himself or herself for the kind of person he/she is,
the individual is not necessarily placing
blame for an event regarded as personally
controllable. A person can believe that he/she
deserves what happened and is therefore "responsible" for it (see Harvey & Rule, 1978),
without believing that he/she is capable of
altering the outcome in the past, present, or
future.
In the case of personal failures, the charterological blamers will point to deficits in
themselves that are believed to account for
these failures. The deficits are likely to lie
in the realm of characteristics that generally
define them, characteristics that are relatively nonmodifiable, stable, and global.
Thus, in achievement tasks, an ability attribution would represent a characterological
self-blaming strategy. In the case of selfblame for failures that are further removed
from the individual, represented by the "delusions of depressives" who blame themselves
for the "violence and suffering in the world"
(see Beck, 1967), the individuals appear to
regard themselves as being punished for who

1801

or what they are. In this case, rather than

perceive themselves as responding, active organisms, depressed individuals seem to perceive themselves as passive stimuli. They do
not believe they actively bring about outcomes that remain under their control.
Rather, negative outcomes occur in reaction
to them by other people and the world at
large. In sum, self-blame by depressives is
proposed as characterological in nature. Since
characterological self-blame and feelings of
helplessness are not logically inconsistent,
their conjunction in depressed individuals
should not be regarded as paradoxical.
Self-Blame Among Rape Victims
The association between self-blame and
depression is probably well recognized and
accepted within this culture. While the association between self-blame and rape is probably not as strong, the more or less popular
image of the self-blaming rape victim may
be more accurate than many would like to
believe. The pervasiveness of self-blame has
been well documented in literature on rape
(see, e.g., Burgess & Holmstrom, 1974a,
1974b, 1976; Griffin, 1971; Hursch, 1977;
Weis & Weis, 197S; Bryant & Cirel, Note 1).
Although fear (of injury, death, and the
rapist) is the primary reaction to rape, selfblame may be second only to fear in frequency of occurrence; perhaps surprisingly,
it is far more common than anger.
In considering the few existing facts on
victim precipitation in the crime of rape,
however, it becomes obvious that the victims' self-attributional strategies (i.e., selfblame) do not reflect an accurate appraisal
of the woman's causal role in the assault.
The National Commission on the Causes and
Prevention of Violence (1969) concluded that
only 4.4% of all rapes are precipitated by
the victim. Although a higher figure, 19%,
has been proposed by Amir (1971), he used a
considerably broader definition in establishing his criteria for victim precipitation. Thus,
criteria such as "risky situations marred with
sexually" were used, affording the interpreter
of data considerable discretion. In light of
these percentages, the pervasiveness of selfblame becomes a puzzling phenomenon.

1802

RONNIE JANOFF-BULMAN

An attempt to account for the pervasiveness of such feelings has involved the proposition that women have been socialized to
accept blame for their own victimization. As
Brownmiller (197S) suggests, women are
conditioned to a female victim mentality.
Brownmiller discusses the psychologies of
Deutsch and Horney and concludes that
masochism is a female trait, one that has
been socialized by men. Similarly, Burgess
and Holmstrom (1974a) contend that women
are socialized to the attitude of "blaming
the victim," a perspective shared by Bryant
and Cirel (Note 1). While there is no doubt
much truth to this socialization hypothesis,
it may paint a very incomplete picture of
the factor(s) responsible for self-blame in
women and the rape victim in particular.
It fits nicely with a portrait of women as
helpless and masochistic and may unwittingly perpetuate a view of women too consistent with the role of rape victim. In particular, this view entirely overlooks the possibility that self-blame by victims of rape may
represent an adaptive response, an attempt
to reestablish control following the trauma
of rape.
A common reaction to rape is the feeling
of a loss of control over one's life (Bard &
Ellison, 1974; Bryant & Cirel, Note 1). The
woman does not feel sure of herself and questions her self-determination. She needs to
feel a sense of control (Hilberman, 1976),
for she feels extremely vulnerable and particularly fears the rapist and a recurrence of
rape. In blaming herself, perhaps the rape
victim is engaging in a type of self-blame
that maximizes a belief in control; that is,
perhaps rape victims engage in behavioral
self-blame rather than characterological selfblame. Whereas the latter type of blame
would provide some support for a view of
women as helpless and masochistic, the former would foster a different image of the
rape victim and her reactions, that of an individual reacting in an adaptive manner
to her recent loss of control.
If the rape victim engages in behavioral
self-blame and attributes her victimization
to a modifiable behavior (e.g., I should not
have walked alone, I should have locked the
windows), she is likely to maintain a belief

in the future avoidability of a similar misfortune, while simultaneously maintaining a

belief in personal control over important life
outcomes. If, on the other hand, the rape
victim blames herself characterologically, attributing the victimization to more or less
unchangeable factors (e.g., I'm a weak person
and can't say so, I'm the type of person who
attracts rapists), she will presumably be considerably less likely to believe that she is
capable of alleviating her vulnerability in the
future and may begin to perceive herself as
a chronic victim.
Two studies were conducted in order to
test the usefulness of the distinction between behavioral and characterological selfblame in the areas of depression and rape.
Study 1 was designed to determine whether
characterological self-blame is a distinguishing characteristic of depressed individuals
and whether it co-occurs with decreased beliefs in personal control among female college students. Study 2 involved surveying
rape crisis centers across the country in order
to determine which type of self-blamebehavioral or characterologicalmore accurately characterizes the reactions of rape victims served by these centers.
Study 1: Depression
Method
Subjects. Subjects were 129 undergraduate women at a large state university who were volunteers drawn from a number of undergraduate psychology courses. Each received one experimental
credit for her participation. Responses from 9 of
the subjects lacked much data, and these were eliminated from the analyses, leaving the responses of
120 subjects.
Procedure, Data - were collected during group
sessions that generally ranged from 10 to IS students. Subjects were told that we were interested in
the relationship between personality variables and
artistic taste, and that there would be three parts
2
These data were collected by Laurie Gunsolley
for her senior honors thesis, which was designed
and completed under the direction of the author.
While Gunsolley was particularly interested in selfesteem, the data have been reanalyzed for this
presentation, using the responses to the Zung SelfRating Depression Scale (1965) as the basis for
distinguishing between the two groups of interest.

SELF-BLAME: DEPRESSION AND RAPE

to the study: completing personality scales, viewing and rating a series of art slides, and reacting
to several "real-life" types of situations. The subjects were asked to complete three personality scales.
The Zung Self-Rating Depression Scale (Zung, 1965)
was used to measure depression.3 In addition, subjects completed the Janis-Field Feelings of Inadequacy Scale (Eagly, 1967), a self-esteem measure,
and the Rotter Internal-External Locus of Control
Scale (Rotter, 1966). Having completed these, the
subjects were then asked to rate seven art slides on
"aesthetic appeal." An overhead projector was used
to show the slides, and ratings were made on 5point scales. These artistic ratings not only provided a "cover" for the study but, more important,
served as a distraction between the first (personality scales) and third (self-blame measures) parts
of the experiment.
In the third phase of the study, subjects were
asked to read four scenarios and to imagine that
the various situations described had actually been
experienced by them; that is, they were told to
react to the scenarios given that they were the
target people presented. In each situation the outcome was negative and the role of the target person was intentionally ambiguous. The scenarios involved the following situations: (a) a car driven
by the target person is in an accident on a snowy
winter day; (b) a social invitation by the target
person is rejected (on the basis of false excuses)
by an individual she recently met and regarded as
a friend: (c) an urgent call for a roommate results
in the target person's taking down the wrong number; the roommate is subsequently unable to return
the call successfully; (d) an intense love relationship is ended when the target person's boyfriend
leaves her and immediately gets involved with another woman.
Subjects were asked to respond to five questions
following each scenario; responses were made on
6-point scales with endpoints not at all and completely. Subjects were asked to indicate how much
they blamed themselves, other people, the environment (i.e., impersonal world), and chance, for the
situations described. The question that tapped
characterological self-blame asked, "Given what
happened, how much do you blame yourself for
the kind of person you are (e.g., the kind of person
who is in an accident [Scenario A], the kind of
person who has invitations turned down [Scenario
B], the kind of person who causes inconveniences
for others [Scenario C], the kind of person who is
rejected in relationships [Scenario D])?" The appropriate "kind of person" was included separately
for each scenario, so that for scenario D only "the
kind of person who is rejected in relationships"
was included. Question 3 sought to tap behavioral
self-blame and asked, "Given what happened, how
much do you blame yourself for what you did
(e.g., your driving behavior [Scenario A], how you
acted when you first met the person [Scenario B],
how you acted when taking down the telephone
number [Scenario C], how you acted with your

1803

boyfriend [Scenario D])?" Question 4 asked, "How

much do you think you deserved what happened?"
and Question 5 following each scenario was, "If the
same situation arose in the future, to what extent
do you believe that you could avoid what happened in this case?" All subjects were thoroughly
debriefed upon completion of the session.

Results
Using a median split, subjects were divided
into nondepressed (responses ranged from
6 to 21 on the Zung scale) and depressed
(22 to 45 on the Zung scale) groups.4 On the
Janis-Field Feelings of Inadequacy Scale,
the depressed group scored lower (i.e., had
lower self-esteem) than the nondepressed
group (64.05 vs. 74.97), F(l, 118) = 36.72,
p < .001, and the depressed group was found
to be more external than the nondepressed
group on Rotter's Internal-External Locus
of Control Scale (13.36 vs. 10.57), /?(!, 118)
= 12.32, p< .001.
Parallel attributional and self-blame measures were summed across the four scenarios;
for example, a score for characterological
blame was derived by adding the individual
responses to each of the four questions (one
following each scenario) that asked about
characterological self-blame. In order to
justify adding the four scales, alpha reliability coefficients were calculated for each of
the eight summed scores. Despite the fact
that each was composed of four scores, only
the perceived avoidability measure failed to
reach a reliability of .50. The general self
and other people attributions were less than
.60, and the other five measures had alpha
reliabilities between .62 and .74.5 Each
3

In accordance with work on depression by Bonnie Strickland, a clinical psychologist in the Department of Psychology, University of Massachusetts, Amherst, a response category labeled "none
of the time" was added to the Zung Self-Rating
Depression Scale (1965). According to Strickland
(personal communication), this renders the scale
particularly sensitive to depression in a college
population.
4
Male and female college students completed the
same depression scale in a study by Haley and
Strickland (Note 2 ) ; their data had a median of 23.
5
Nunnally (1967) writes that in early work on
"hypothesized measures of a construct", reliabilities
of .50 or .60 are sufficient standards (see p. 226).

1804

RONNIE JANOFF-BULMAN

summed score could range from a total of

0 to 24.
The depressed and nondepressed groups
did not differ in the amount of blame they
attributed to themselves in general, nor did
they differ in the amount of behavioral selfblame reported, F ( l , 118) = 2.47, ns. However, the groups did differ significantly in
the amount of characterological self-blame
reported, with more characterological selfblame reported by the depressed than the
nondepressed group (11.59 vs. 10.03), F(l,
118) =4.33, p < .05. Other differences on
the total scores were found for attributions
to chance; consistent with their greater externality on Rotter's scale, the depressed
group blamed chance more than the nondepressed group (11.38 vs. 9.73), F(l, 118) =
4.54, p < .05. Further, there was a marginally significant difference between the two
groups on the question of how much they
felt they deserved what happened, with the
depressed group reporting greater deservingness than the nondepressed group (8.31 vs.
7.24), F ( l , 118) = 3.68, p = .057.6
A stepwise discriminant analysis was conducted in order to determine the variables
that differentiated best between the two
groups. All blame attribution measures were
entered, with the exception of general selfblame, since characterological and behavioral
self-blame were assumed to be finer distinctions of the general measure. Attributions to
chance emerged as the best discriminator,
F(5, 114) = 4.54, Wilks A. = .963, and characterological self-blame emerged as the next
strongest differentiator, F(5, 114) = 3.29,
Wilks A = .937. These were followed, respectively, by attributions to other people, environment, and behavioral self-blame.
The correlations between deservingness,
perceived avoidability, and the two types of
self-blame were all strong. As an exploratory
tool, an analysis of variance was conducted in
order to further investigate the relationship
between the variables. It should be noted that
the low reliability of the avoidability measure
calls for caution in interpreting the results of
this analysis. Median splits were performed
on the behavioral self-blame and characterological self-blame totals, and deservingness

and avoidability totals were each analyzed

by behavioral (high-low) and characterological (high-low) self-blame. A main effect for
characterological self-blame emerged from the
analysis of deservingness, with less deservingness reported by those who engaged in low
characterological self-blame as compared with
those who engaged in high characterological
self-blame (6.20 vs. 9.26), F(l, 119) = 29.78,
p < .001. On the other hand, a main effect
for behavioral self-blame emerged from the
analysis of perceived future avoidability, with
less perceived future avoidability reported by
those who engaged in low behavioral selfblame as compared with those who engaged in
high behavioral self-blame (11.89 vs. 13.88),
F(l, 118) = 6.85, p < .01.
Discussion
When self-blame was treated as a single
entity (i.e., "self" as one of several possible
factors tapped for blame attributions), no
differences were found between the depressed
and nondepressed students on this variable.
However, when self-blame was divided into
two types of self-attributions, behavioral and
characterological, differences between the
groups emerged. While the depressed and
nondepressed students did not differ in terms
of behavioral self-blame, they did differ significantly in terms of characterological selfblame; that is, the depressed students blamed
themselves more characterologically than did
the nondepressed students.
The suggestion that characterological selfblame follows from attributions to uncontrollable factors received strong support.
Those who were depressed were more likely to
attribute negative outcomes to chance, a
variable that differentiated best between the
depressed and nondepressed groups. Further,
the depressed subjects were more external in
locus of control orientation. Low self-esteem
and somewhat increased feelings of deservingWhen analyses were rerun using the conservative Scheffe procedure to correct for error rate inflation, significant differences between the depressed
and nondepressed groups were again found on selfesteem, internal-external control, characterological
self-blame, and chance.

SELF-BLAME: DEPRESSION AND RAPE

ness characterized the depressed population,

suggesting that characterological self-blame is
esteem related, not control related.
The results would have been considerably
more compelling if it were found that those
students who were not depressed engaged in
more behavioral self-blame than depressed
students, yet this was not the case. However,
it can perhaps be argued that the behavioral
self-blame reported by the depressed and
nondepressed populations differed in an important way; for the depressed group the behavioral self-blame co-occurred with characterological self-blame, and blaming one's
behavior was thus an extension of blaming
one's character. It may be difficult to blame
one's character without blaming one's behavior, yet it may be very possible to blame
one's behavior without blaming one's character. In the former instance the behavior
may be regarded as uncontrollable in that it
is a direct and unalterable extension of one's
character (i.e., controlled by one's character).
In the latter case the behavioral self-blame
does not reflect decreased self-esteem, but
rather the belief that one's behavior is modifiable. Perhaps behavioral self-blame, when
displayed in conjunction with characterological self-blame, is simply a further reflection
of characterological self-blame. However,
when it occurs alone it is likely to represent
an adaptive response, stemming from a desire
to maintain a belief in personal control following a negative outcome.
Study 2: Rape 7
Method
Respondents. Respondents were rape crisis centers
located throughout the United States. Center names
were derived primarily from a list located in a federal report on rape and its victims (Brodyaga et al.,
Note 3 ) ; this list was supplemented by names of
rape crisis centers found in an informal directory at
a local women's center. Services that were hotlines
only or were task forces without counseling services
were excluded from the final list. Questionnaires were
mailed to 120 centers representing 37 states and the
District of Columbia. Thirty of the questionnaires
were returned "addressee unknown." Of the remaining 90 crisis centers, 48 responded (53% return
rate; including those returned "addressee unknown,"
the return rate was 40%).
Questionnaire. In a cover letter I identified myself
as a social psychologist interested in the nature of

1805

self-blame among victims of rape; letter recipients

were asked to base their questionnaire responses on
their experiences as counselors of rape victims. The
questionnaire items dealt primarily with the issue of
self-blame. Crisis centers were asked to indicate
approximately how many rape victims they see yearly
and of those they see, the percentage who blame
themselves, at least in part, for the rape. The behavioral self-blame question asked, "Of the rape
victims you see, what percentage blame themselves
for the rape because of some behavior (act or omission) they engaged in at the time of or immediately
prior to the rape (e.g., 'I should not have walked
alone,' 'I should not have hitchhiked,' 'I should have
locked my windows') ?" The rape crisis centers were
then asked to provide specific examples of behavioral
self-blame related by the women they have counseled. The characterological self-blame question asked,
"Of the rape victims you see, what percentage blame
themselves for the rape because of some character
trait or personality flaw they believe they have (e.g.,
'I am so stupid, I deserved to be raped,' 'I'm the
kind of woman who attracts rapists,' 'I am a weak
person and can't say no') ?" Specific examples of this
type of blame were then requested as well. The
centers were also asked to indicate on two 7-point
scales, with endpoints almost not at all and completely, how much self-blaming characterized the
women who engaged in behavioral and characterological self-blame, respectively; this was included in
order to ascertain whether behavioral and characterological self-blamers differ in terms of the amount of
self-blame they attribute to themselves for the rape.

Results
Of the 48 rape crisis centers that responded,
38 completed the questionnaire, 6 wrote letters
providing general comments, and 4 wrote that
they did not provide direct counseling services
and were therefore unable to complete the
items. Results were therefore based on the
completed questionnaires of 38 centers. The
rape crisis centers differed markedly in the
scope of their operation, with the 3 smallest
serving 12, 30, and 40 rape victims yearly,
and the 3 largest serving 1,200, 1,250, and
1,500; the mean number of rape victims seen
across the centers was 335.
In general, self-blame was reported as quite
common; the reported mean percentage of
7

The results of this study were reported by the

author at the symposium "New Directions in Control Research" at the convention of the American
Psychological Association, Toronto, 1978. The author
thanks Chris Eagan for her invaluable help on the
project.

1806

RONNIE JANOFF-BULMAN

women who blamed themselves at least in part

for the rape was 14%. Of those who blamed
themselves, behavioral self-blame was reported
as considerably more common than characterological self-blame, and the differences between the reported incidence of the two blaming strategies was significant, F(\, 32) =
140.90, p < .001; an average of 69% of the
women were reported as blaming themselves
behaviorally, whereas an average of 19% were
reported as blaming themselves characterologically. Further, examples of the two types of
self-blame provided by the rape crisis centers
confirmed the fact that they were readily able
to distinguish between the two. Frequently
mentioned examples of behavioral self-blame
included the following: I shouldn't have let
someone I didn't know into the house, I
shouldn't have been out that late, I should not
have walked alone (in that neighborhood),
I should not have hitchhiked, I should not
have gone to his apartment, I shouldn't have
left my window open, I should have locked
my car. Examples of characterological selfblame that were frequently reported included:
I'm too trusting, I'm a weak person, I'm too
naive and gullible, I'm the kind of person who
attracts trouble, I'm not a very aware person,
I'm not at all assertiveI can't say no, I'm
immature and can't take care of myself, I'm
not a good judge of character, I'm basically a
bad person. It is perhaps worth noting that
examples of behavioral blame were, without
exception, reported in the past tense (i.e., I
should have/should not have), whereas examples of characterological self-blame were
presented in the present tense (I am/am not),
perhaps implicitly indicating the presumed
modinability/nonmodinability of factors associated with behavioral and characterological
self-blame respectively (cf. Elig & Freize,
1975). However, the examples of the two
types of self-blame provided in the questionnaire were consistent with the different tenses
reported in the examples of the crisis centers,
and this alone could have accounted for the
findings. The author did not realize that she
had made these distinctions on the questionnaire until the results clearly differentiated
between the tenses used for the two types of
self-blame. Finally, in responding to how

much women blamed themselves for the rape,

the centers reported that characterological
self-blamers blamed themselves significantly
more for the rape than did behavioral selfblamers (3.92 vs. 3.23), F(l, 25) = 11.29,
p < .002.
Discussion
The rape crisis center counselors reported
that the majority of rape victims do blame
themselves, at least in part, following the
rape. However, the focus of this self-attribution is a behavioral act or omission engaged in
at the time of (or immediately preceding) the
rape. Fewer than one-fifth of the women
served by the centers blamed themselves in a
characterological way, evidence that the "popular" view of the masochistic rape victim who
perceives herself as worthless is largely unfounded. Rather, the self-blame in which most
rape victims engage may represent a control
maintenance strategy, a functional response
to a traumatic event. Given the large discrepancies between those who blame themselves
behaviorally and characterologically, it follows that most women clearly blame themselves in a behavioral manner only and do not
combine this response with characterological
self-blame, as may be the case with depressives (see above). In suggesting that behavioral self-blame reflects a positive impulse in
rape victims, there is no intention of implying that the rape was the woman's fault. It is
even likely that the woman who engages in
behavioral self-blame does not do so to the
exclusion of blaming the rapist, society, or
other factors. These blame attributions, instead, would stem from different motivations,
control maintenance being the motivation
behind self-blame.
Two potentially serious objections to this
study require a response. First, many of the
women who volunteer or work in rape crisis
centers may be ardent feminists who would
be more likely to indicate that women blame
themselves behaviorally rather than characterologically, for the latter suggests that
women see themselves as worthless. In response, if the crisis center workers truly
wanted to present women in a positive light,

SELF-BLAME: DEPRESSION AND RAPE

they would have indicated quite simply that

women infrequently blame themselves. Further, there was nothing in the questionnaire
or cover letter to indicate that one type of
self-blame was "healthier" than another, and
several counselors commented that they had
never before distinguished between types of
self-blame but that it appeared interesting to
them. Comments by the counselors indicated
that these women were concerned about the
health of the women they served and that preserving a positive image of womanhood in
general was clearly not central to their activities as rape counselors. The second criticism
that could be raised is potentially more serious. It is that women who go to rape crisis
centers are most likely to be individuals who
do not blame themselves characterologically
and do not feel they deserved to be raped.
Thus, there is a self-selected population of
behavioral self-blamers served by rape crisis
centers. It is difficult to counter this claim, for
there is probably much truth to it. One must
realize, however, that the literature written
on rape is almost entirely derived from
women who seek help after rape and not from
women who quietly keep the trauma to themselves, ashamed to talk about it or admit it.
The pervasiveness of self-blame documented
in the rape literature is drawn primarily from
observations of women at rape crisis centers,
from women's centers, or from women who
agree to be interviewed by researchers, also a
population likely to be self-selected. Thus, the
negative image of the rape victim engaging in
masochistic, maladaptive self-blame derives
from a rape victim population likely to be
very similar to that served by the rape crisis
centers surveyed. It might also be mentioned
that those women who have least difficulty
coping with the rape and who are apt to be
behavioral self-blamers are probably also
missing from the rape center population, for
they may not require help (outside their own
circle of family and friends) following the
rape. Perhaps it is sufficient to point out that
within the population of women served by
rape crisis centers, self-blame has been improperly understood as self-derogating, reflecting the woman's belief in her own worthlessness, rather than as a response that

1807

reflects a positive attempt to reestablish personal control.

General Conclusions and Implications
Self-blame appears to be a label for two
very different self-attributions, characterological self-blame being esteem related, and behavioral self-blame being control related. Selfblame as a predictor of good coping and selfblame as a concomitant of depression are no
longer inconsistent in light of the two types
of self-blame. Further, the paradox in depressionthat individuals are simultaneously
helpless and self-blamingcan be resolved if
characterological self-blame characterizes depressives and differentiates them from nondepressed individuals. The division of selfblame into two different phenomena even has
political or cultural implications, for selfblame by a victimized group such as rape victims can now be understood in such a manner
as to preclude the perpetuation of a negative
image of the group in question. It is perhaps
unfortunate that one term has been used as
a label for these two different self-attributions,
for the singular term self-blame blurs important distinctions between adaptive and
maladaptive responses to failures and victimizations. Since popularly the term has
negative connotations, it would perhaps be
desirable to provide a more neutral label for
behavioral self-blame. Particularly in the case
of rape, this would render more politically
palatable the proposition that behavioral
self-blame is of functional value for victims
of rape.
The recognition of two types of self-blame
may have therapeutic implications. Seligman's
(197S) control-oriented strategies continue to
seem appropriate for depressives, whose selfblaming does not imply high perceived control, but rather lack of control. Further, a
cognitive therapy that entails reattributing
the focus of one's attributions (e.g., from
character to behavior) might be of value in
treating depressives. In general, leading people to focus on behaviors that are alterable,
rather than on their relatively nonmodifiable,
more global character, may increase perceived
future avoidability of negative events and

1808

RONNIE JANOFF-BULMAN

perceived control in general, outcomes that

would presumably be of positive value.
Dweck's (1975) successful reattribution training with helpless students, involving reattributing their ability attributions for failure to
effort attributions, suggests the potential of
such cognitive strategies using self-blame.
In the case of rape, the control concerns
that may be implicit in the rape victim's selfblame often seem to be ignored in counseling,
not because they are regarded as unimportant,
but because they may go unrecognized. One
counseling technique for rape victims includes
repeatedly telling a woman that there is nothing she could have done to avoid the rape,
that it was entirely the rapist's doing and outside of her control. Although meant to be
reassuring, these statements could conceivably be not at all helpful, in light of the
proposition that the women are seeking to reestablish a sense of control. Rather, counselors
should perhaps recognize the functional value
of behavioral self-blame and concentrate on
enabling the victim to reestablish a belief in
her relative control over life outcomes (e.g.,
discussing possible ways of minimizing the
likelihood of a future rape). Too often, behavioral self-blame is regarded as detrimental
to mental health. Rather, it may serve as an
indicator of the victim's psychological needs
at the time.
Behavioral and characterological self-blame
appear to be distinct reactions yet are far
from fully understood. Ideas raised in this
paper have been tested only with female subjects and thus may not generalize to other
populations; this issue of generalizability particularly calls for research with male subjects.
Further, the relationship between the two
types of self-blame would appear to be a
fruitful area for future study. Does behavioral
self-blame that occurs with characterological
self-blame, for example, lose its adaptive
value, or is it similar to behavioral blame that
occurs without characterological self-blame?
Is characterological self-blame that occurs
without behavioral self-blame more or less
maladaptive than characterological self-blame
that occurs with behavioral self-blame? In
addition, longitudinal studies designed to tap
the coping implications of these two types of

self-blame would be important contributions

to our understanding of the relationship between coping and attributional strategies.
Another possible direction lies in the area of
blaming strategies by help-givers. Brickman
and his colleagues (Brickman et al., Note 4)
have presented a compelling case for the psychological tensions that exist between conditions that render helping appropriate (i.e.,
regarding the recipient of help as not responsible) and conditions that render helping
effective (i.e., attributing responsibility to
the recipient of help). That is, one is apt to
help an individual who is not to blame for a
misfortune, yet this attribution minimizes the
belief that one's help will be effective. Perhaps training both help-givers and recipients
of help to hold behavioral blame orientations
(as opposed to characterological blame orientations) would help resolve the existing tensions. Last, the therapeutic implications of
the two types of self-blamebehavioral and
characterologicalremain an area ripe for
future study.
Reference Notes
1. Bryant, G., & Cirel, P. A community response to
rape: An exemplary project (Polk County Rape/
Sexual Assault Care Center). Washington, D.C.:
National Institute of Law Enforcement and Criminal Justice, 1977.
2. Haley, W. E., & Strickland, B. R. Locus of control
and depression. Paper presented at the meeting
of the Eastern Psychological Association, Boston,
1977.
3. Brodyaga, L., Gates, M., Singer, S., Tucker, M.,
& White, R. Rape and its victims: A report for
citizens, health facilities, and criminal justice
agencies. Washington, D.C.: National Institute of
Law Enforcement and Criminal Justice, 1975.
4. Brickman, P., Rabinowitz, V. C., Coates, D., Cohn,
E., Kidder, L., & Karuza, J. Helping. Unpublished
manuscript, University of Michigan, 1979.

References
Abramson, L. Y., & Sackeim, H. A. A paradox in depression: Uncontrollability and self-blame. Psychological Bulletin, 1977, 84, 838-8S1.
Abramson, L. Y., Seligman, M. E. P., & Teasdale,
J. D. Learned helplessness in humans: Critique
and reformulation. Journal of Abnormal Psychology, 1978, 87, 49-74.
Amir, M. Patterns in forcible rape. Chicago, 111.: University of Chicago Press, 1971.

SELF-BLAME: DEPRESSION AND RAPE

Bard, N., & Ellison, K. Crisis intervention and investigation of forcible rape. The Police Chief, 1974,
41, 68-73.
Beck, A. T. Depression: Clinical, experimental, and
theoretical aspects. New York: Harper & Row,
1967.
Bowers, K. Pain, anxiety, and perceived control.
Journal of Consulting and Clinical Psychology,
1968, 32, 596-602.
Brownmiller, S. Against our will: Men, women, and
rape. New York: Simon & Schuster, 1975.
Bulman, R. J., & Wortman, . B. Attributions of
blame and coping in the "real world": Severe accident victims react to their lot. Journal of Personality and Social Psychology, 1977, 35, 351-363.
Burgess, A. W., & Holmstrom, L. L. Rape trauma
syndrome. American Journal of Psychiatry, 1974,
131, 981-985. (a)
Burgess, A. W., & Holmstrom, L. L. Rape: Victims
of crisis. Bowie, Md.: Robert J. Brady, 1974. (b)
Burgess, A. W., & Holmstrom, L. L. Coping behavior
of the rape victim. American Journal of Psychiatry, 1976, 133, 413-318.
Dweck, C. S. The role of expectations and attributions in the alleviation of learned helplessness.
Journal of Personality and Social Psychology, 1975,
31, 674-685.
Eagly, A. H. Involvement as a determinant of response to favorable and unfavorable information.
Journal of Personality and Social Psychology
Monograph, 1967, 7(3, Pt. 2).
Elig, T., & Frieze, I. A multidimensional scheme for
coding and interjecting perceived causality for
success and failure events: The coding scheme of
perceived causality (CSPC). JSAS: Catalog of
Selected Documents in Psychology, 1975, 5, 313.
(Ms. No. 1069)
Glass, D., & Singer, J. Urban stress. New York:
Academic Press, 1972.
Griffin, S. Rape: The ail-American crime. Ramparts
Magazine, 1971, 10, 26-35.
Harvey, M. D., & Rule, B. G. Moral evaluations and
judgments of responsibility. Personality and Social
Psychology Bulletin, 1978, 4, 583-588.
Hilberman, E. Rape: The ultimate violation of the
self. American Journal of Psychiatry, 1976, 133,
436-437.
Hursch, C. H. The trouble with rape. Chicago:
Nelson-Hall, 1977.
Kelley, H. H. Attribution in social interaction. Morristown, N.J.: General Learning Press, 1971.
Langer, E. J., & Rodin, J. The effects of choice and
enhanced responsibility for the aged: A field ex-

1809

periment in an institutional setting. Journal of

Personality and Social Psychology, 1976, 33, 951955.
Lerner, M. J., & Miller, D. T. Just world research
and the attribution process: Looking back and
ahead. Psychological Bulletin, 1978, 85, 1030-1051.
Medea, A., & Thompson, K. Against rape. New York:
Farrar, Strauss & Giroux, 1974.
National Commission on the Causes and Prevention
of Violence. Crimes of violence (Vol. 2). Washington, D.C.: U.S. Government Printing Office, 1969.
Nunnally, J. C. Psychometric theory. New York:
McGraw-Hill, 1967.
Rotter, J. B. Generalized expectancies for internal
versus external control of reinforcement. Psychological Monographs, 1966, 80(1, Whole No. 609).
Schulz, R. Effects of control and predictability on the
physical and psychological well-being of the institutionalized aged. Journal of Personality and Social
Psychology, 1976, 33, 563-573.
Seligman, M. E. P. Helplessness: On depression, development and death. San Francisco: Freeman,
1975.
Spielberger, C. D. Anxiety as an emotional state. In
C. D. Spielberger (Ed.), Anxiety: Current trends
in theory and research (Vol. 1). New York:
Academic Press, 1972.
Walster, E. Assignment of responsibility for an accident. Journal of Personality and Social Psychology,
1966, 3, 73-79.
Weiner, B. Achievement motivation as conceptualized
by an attribution theorist. In B. Weiner (Ed.),
Achievement Motivation and attribution theory.
Morristown, N.J.: General Learning Press, 1974.
Weiner, B., Frieze, L, Kukla, A., Reed, L., Rest, S.,
& Rosenbaum, R. M. Perceiving the causes of success and failure. New York: General Learning
Press, 1971.
Weis, K., & Weis, S. Victimology and the justification of rape. In I. Drapkin & E. Viano (Eds.),
Exploiters and exploited: Vol. /. Victimology: A
new focus. Lexington, Mass.: Lexington Books,
1975.
Wortman, C. B., & Dintzer, L. Is an attributional
analysis of the learned helplessness phenomenon
viable? A critique of the Abramson-SeligmanTeasdale reformulation. Journal of Abnormal Psychology, 1978, 87, 75-90.
Zung, W. K. A self-rating depression scale. Archives
of General Psychiatry, 1965, 12, 63-70.
Received November 21, 1978