Weiner

Download as pdf or txt
Download as pdf or txt
You are on page 1of 26

Psychological Review

1985, Vol. 92, No. 4, 548-573

Copyright 1985 by the American Psychological Association, Inc.


0033-295X/85/S00.75

An Attributional Theory of Achievement Motivation and Emotion


Bernard Weiner
University of California, Los Angeles
A theory of motivation and emotion is proposed in which causal ascriptions play
a key role. It is first documented that in achievement-related contexts there are a
few dominant causal perceptions. The perceived causes of success and failure share
three common properties: locus, stability, and controllability, with intentionality
and globality as other possible causal structures. The perceived stability of causes
influences changes in expectancy of success; all three dimensions of causality affect
a variety of common emotional experiences, including anger, gratitude, guilt, hopelessness, pity, pride, and shame. Expectancy and affect, in turn, are presumed to
guide motivated behavior. The theory therefore relates the structure of thinking to
the dynamics of feeling and action. Analysis of a created motivational episode involving achievement strivings is offered, and numerous empirical observations are
examined from this theoretical position. The strength of the empirical evidence,
the capability of this theory to address prevalent human emotions, and the potential
generality of the conception are stressed.

In 1645, Miyomota Musashi was contem- of the Redwood National Park for their sorry state. Tim
plating the causes of his past success as a war- Skaggs, the union business agent, shrugged. 'You could
spend a lifetime fixing blame,' he said. (Martinez, 1982,
rior. In A Book of Five Rings he mused,
Pt. 5, p. 1)
When I reached thirty I looked back on my past. The previous victories were not due to my having mastered strategy.
Perhaps it was natural ability, or the order of heaven, or
that other schools' strategy was inferior. (1645/1974, p. 35)

And even the former coach of my favorite


football team found it necessary to soul search
about causality following a series of losses.
Again from the Los Angeles Times:

About 275 years later, and approximately


11,000 miles away, the editors of Scientific Here it is Thanksgiving week, and the Los Angeles Rams
American were wondering why America was are looking like the biggest turkeys, in town. Coach Ray
flourishing. They reasoned, "The wealth and Malavasi has eliminated bad luck, biorhythms, and sunas the reasons why his football team has lost 9 of its
general prosperity of the country are largely spots
last 10 games. Now he's considering the unthinkable posdue to the intelligence and energy of its people, sibilities that: (a) he has lousy players or (b) they aren't
but it can hardly be disputed that it is equally really trying. (Robert, 1982, Pt. 3, p. 3)
due to the natural wealth of the country" (Staff,
Why this constant pursuit of "why"? A
1926, p. 228). Unfortunately, battles are lost
as often as they are won, and countries undergo number of explanations come to mind (see
economic decline as well as enrichment. Dur- Forsyth, 1980; Weiner, 1985). We might just
ing our recent financial recession the Los An- want to know, that is, to understand the environment, to penetrate ourselves and our surgeles Times reported,
roundings. This familiar motivational interTimber industry experts blame high interest rates, the pretation is known as the principle of mastery
housing slump, tough logging regulation, and expansion
(White, 1959). In addition, it clearly is functional to know why an event has occurred. As
Kelley (1971) stated, "The attributor is not
This article was written while the author was supported simply an attributor, a seeker after knowledge;
by Grant MH38014 from the Public Health Service, his latent goal in attaining knowledge is that
National Institute of Mental Health.
of effective management of himself and his enThe author wishes to thank Sandra Graham for her vironment" (p. 22). Once a cause, or causes,
helpful comments on an earlier draft of this article.
Requests for reprints should be sent to Bernard Weiner, are assigned, effective management may be
Department of Psychology, University of California, Los possible and a prescription or guide for future
action can be suggested. If the prior outcome
Angeles, California 90024.
548

ACHIEVEMENT MOTIVATION AND EMOTION

was a success, then there is likely to be an attempt to reinstate the prior causal network.
On the other hand, if the prior outcome or
event was undesiredsuch as exam failure,
social rejection, political loss, or economic declinethen there is a strong possibility that
there will be an attempt to alter the causes to
produce a different (more positive) effect.
Desire for mastery and functional search,
two of the generators of causal exploration, do
not seem to specifically characterize one geographical area or one period of human history.
The Japanese warrior in the Middle Ages and
today's union representative are engaged in the
same endeavor: attempting to assign causality.
Indeed, one might argue that adaptation is not
possible without causal analysis. The warrior
needs to know why he is winning battles so he
can survive the next one, just as the union representative needs to explain why the industry
is doing poorly in order to urge wiser actions
in the future. Because of the apparent pancultural, timeless aspect of causal search and
exploration, and because of the evident adaptive significance of this activity, causal ascriptions are proposed to provide the building
blocks for the construction of a theory of motivation and emotion.
This article advances an attributional theory
of motivation and emotion, with achievement
strivings as the theoretical focus. Initially, the
most salient causes of success and failure in
achievement-related contexts are identified.
The basic properties of these causes, or the
structure of causal thinking, is then determined from both a dialectic and an empirical
perspective. Three causal dimensions are discovered: locus, stability, and controllability.
The structure of causal thinking is next related
to emotion and motivation. Thus, this article
progresses from a description of causal perceptions to causal structure, and then from
causal structure to an examination of the dynamics of action. It is documented that causal
stability influences changes in goal anticipations, while the three causal dimensions determine the emotional experiences of anger,
gratitude, guilt, hopelessness, pity, pride, and
shame. Guided by Expectancy X Value theory,
I presume that expectancy and affect direct
motivated behavior. Examples of research on
the disparate topics of parole decisions, smoking cessation, and helping behavior suggest the

549

generalizability of the theory beyond the


achievement-related theoretical focus.
Perceived Causes of Success and Failure
In the opening paragraphs of this article,
far-ranging examples of causal search are provided. Many investigations have been conducted that more systematically examine
causal perceptions, particularly the perceived
causes of success and failure in achievementrelated situations. Two appropriate research
procedures have been followed. In one, subjects
are provided only with outcome information,
namely, that success or failure has taken place.
The outcome might be imagined, induced, or
have occurred in a real setting, and might pertain to the subject or to another who is being
judged. The subjects are then asked to explain
the outcome, using a free-response procedure
where the possibilities that come to mind are
listed. In a related methodology, participants
are provided with a large list of causes and rate
the contribution of each cause to the outcome.
These causes often were ascertained in pilot
research using a free-response methodology,
and represent the dominant perceptions or
embracing categories.
A summary of 10 pertinent studies is presented in Table 1. Table 1 reveals the source
of the data, the characteristics of the sample
and task, and the four most dominant causal
ascriptions for success. The data for failure reveal an identical story and are not presented.
The message of Table 1 is clear, particularly
inasmuch as the research investigations made
use of a variety of types of subjects judging a
variety of achievement situations, and involving the self or another. A virtually infinite
number of causal ascriptions are available in
memory. However, within the achievement
domain, a relatively small number from the
vast array tend to be salient. The most dominant of these causes are ability and effort. That
is, success is ascribed to high ability and hard
work, and failure is attributed to low ability
and the absence of trying. This holds true for
the majority of cultures that have been examined (see Triandis, 1972). The economy or
simplicity in causal thinking evident in the
achievement domain appears in the explanation of other outcomes, such as wealth and
poverty (Feather & Davenport, 1981; Furn-

550

BERNARD WEINER

ham, 1982a, 1982b) and affiliative acceptance


and rejection (Anderson, 1983a; Sobol & Earn,
in press). However, relatively few studies have
been conducted outside of achievement-related
contexts.
The Structure of Perceived Causality
I now turn from causal description and
identify the underlying structure of perceived
causality. A reasonable initial question to raise
is why does one want to determine causal
structure? What purpose or role does this play
in the goal of theory construction? In response
to this query, consider that, within any particular activity, a myriad of distinct causal explanations are possible. Furthermore, for example, the causes of success and failure at

achievement-related activities, such as ability


and effort, may be quite unlike the perceived
causes of social acceptance and rejection, such
as personality or physical attractiveness. One
puzzle that arises is the relation or the comparability between the various causal explanationsin what way(s) are ability and effort,
or ability and physical beauty, alike and in
what way(s) do they differ? A taxonomic structure enables this question to be answered, for
by finding the underlying properties of causes,
or their common denominators, previous incomparable qualitative distinctions can be replaced with quantitative causal comparisons.
For example, rather than merely being different, both ability and physical beauty may be
considered properties of the actor and thus are
similar, whereas they both differ from a cause

Table 1
Investigations of the Perceived Causes of Success and Failure
Experiment

Subjects

Perspective

Frieze (1976)

College students

Self and
other

Hypothetical school
and game
performance

Effort, ability, luck, and other


persons

Elig & Frieze (1979)

College students

Self

Anagrams

Task, ability, stable effort, and


mood

Frieze & Snyder (1980)

1 st-5th graders

Other

Hypothetical
academic test, art
project, sports, and
game

Unstable effort, ability, interest,


and task

Cooper & Burger (1980)

Teachers

Other

School performance of
students

Typical effort, academic ability,


immediate effort, and
attention

Burger, Cooper, & Good


(1982)

Teachers

Other

School performance of
students

Ability, immediate effort, stable


effort, and attention

Anderson (1983a)

College students

Other

Variety of hypothetical
situations

Behavioral preparation,
experience and skill, effort
level, and general knowledge

Self

School exam

Self

School exam

Effort, luck/chance, task


characteristics, and interest
Effort, ability, task
characteristics, and interest

Advantaged
students

Self

Academic test

Disadvantaged
students

Self

Academic test

Willson& Palmer (1983)


Study 1
College students
Study 2
Bar-Tal, Goldberg, &
Knaani(1984)
Study 1
Study 2

College students

Task

Dominant attributions

7th graders
Test preparation, effort for
study, concentration during
study, and teacher's ability
Test preparation, concentration
during study, effort for study,
and self-confidence

ACHIEVEMENT MOTIVATION AND EMOTION

that is not a property of the actor, such as the


objective ease or difficulty of a task. This type
of analysis facilitates empirical study so that
other associations may be discovered that contribute to the meaning and significance of a
cause.
Logical Analysis of Causal Structure
The first systematic analysis of causal structure was proposed by Heider (1958). Rightly
called the originator of the attributional approach in psychology, Fritz Heider has been
in the background of much of the present theory. The most fundamental causal distinction
made by Heider (1958) was stated as follows:
"In common-sense psychology (as in scientific
psychology) the result of an action is felt to
depend on two sets of conditions, namely, factors within the person and factors within the
environment" (p. 82).
Since the early 1950s, psychologists have
embraced an internal-external distinction (see
Collins, Martin, Ashmore, & Ross, 1974). But
the domination of internal-external comparisons in psychology arrived with the work of
Rotter (1966), for his classification of individuals into internals and externals became a focus for research. Thus, the analysis of the
structure of causality logically began with an
internal-external (locus) dimension.
The argument was then made by Weiner et
al. (1971) that a second dimension of causality
was required. The reasoning was that, among
the internal causes, some fluctuate, whereas
others remain relatively constant. For example,
ability (or, more appropriately, aptitude) is
perceived as a constant capacity; in contrast,
other causal factors including effort and mood
are perceived as more variable, changing from
moment to moment or from period to period.
Among the external causes the same reasoning
applies. For example, success in rowing across
a lake may be perceived as due to the unchanging narrowness of the lake or because of
the variable presence of wind. Weiner et al.
(1971) thus characterized the causes they
thought were most dominant in achievementrelated contexts, namely, ability, effort, task
difficulty, and luck, within a 2 X 2 categorization scheme. Ability was classified as internal
and stable, effort as internal and unstable, task
difficulty was thought to be external and stable,
and luck was considered external and unstable.

551

It is now realized that there are many shortcomings of this classification (see Weiner,
1983). Ability may be perceived as unstable if
learning is possible; effort often is perceived as
a stable trait, captured with the labels of lazy
and industrious; tasks can be changed to be
more or less difficult; and luck may be thought
of as a property of a person (lucky or unlucky).
Thus, the causes within the four cells did not
truly represent the classification system (i.e.,
they did not conform to the phenomenology
of the naive attributor). Less ambiguous entries
might have been aptitude, temporary exertion,
objective task characteristics, and chance (see
Weiner, 1983). Hindsight, however, is better
than foresight, and the problems so evident
now were not fully recognized in 1971.
A third dimension of causality was then established with the same deductive reasoning
that led to the identification of the stability
dimension. Rosenbaum (1972) recognized that
mood, fatigue, and temporary effort, for example, all are internal and unstable causes.
Yet they are distinguishable in that effort is
subject to volitional controlan individual
can increase or decrease effort expenditure.
This is not typically true of mood or the onset
of fatigue, which under most circumstances
cannot be willed to change. The same distinction is found among the internal and stable
causes. Some so-called traits such as laziness,
slovenliness, or tolerance often are perceived
as under volitional or optional control, whereas
this is not characteristic of other internal and
stable causes such as math or artistic aptitude
and physical coordination.
The identification of this property, now
called controllability (Weiner, 1979), enlightened and solved some issues while creating
other difficulties. Among the illuminated topics
was the distinction by Rotter (1966) between
internal versus external perceptions of control
of reinforcement. Within the three-dimensional taxonomy, two of the proposed causal
properties are labeled locus and control. A
cause therefore might be internal yet uncontrollable, such as math aptitude. If failure is
ascribed to poor aptitude, then the performance is perceived as determined by skill and
ability. According to Rotter, this indicates that
the outcome is perceived as subject to internal
control. Yet a genetically determined aptitude
will not be perceived as controllable by a failing

552

BERNARD WEINER

math pupil. Thus, confusion is evident in the


Rotter one-dimensional taxonomy. Locus and
control, not locus of control, describe causal
perceptions. To avoid confusion, the locus dimension should be labeled locus of causality.
Empirical Analysis of Causal Structure
The logical analysis of causal structure has
an inherent limitation: Causal dimensions are
derived from attribution theorists, rather than
from their subjects. It is conceivable that each
theorist might have his or her own rational
scheme of causal organization and that these
postulated structures will not be identical between theorists nor the same as those of the
layperson. Empirical evidence therefore is
needed concerning the organization or the interrelations in causal structure. Three mathematical techniques have been used to analyze
the responses of research participants for underlying causal structure: factor or cluster
analysis (Foersterling, 1980; J. Meyer, 1980;J.
Meyer &Koelbl, 1982;Wimer&Kelley, 1982),
multidimensional scaling (Falbo & Beck, 1979;
Lee, 1976; Michela, Peplau, & Weeks, 1982;
Passer, 1977; Passer, Kelley, & Michela, 1978;
Stern, 1983), and correlations with a priori
schemes (Stern, 1983).
A brief summary of the empirical research
concerning causal structure is shown in Table
2. Table 2 includes 7 of the 10 pertinent investigations already cited. The research of
Foersterling (1980) and Lee (1976) is omitted
because they only examined the ratings of four
causes (they did find the locus and stability
dimensions). In addition, the study by Falbo
and Beck (1979) is excluded because of methodological flaws (see Michela et al., 1982; Weiner, 1983).
It is evident from Table 2 that all studies,
with the possible exception of Passer et al.
(1978), identify a locus dimension of causality.
Given the prominence of this psychological
property, perhaps this finding increases one's
belief in the entire set of data. Turning to the
stability dimension, investigators with the exception of Passer (1977), and perhaps Passer
et al. (1978) and Wimer and Kelley (1982),
find a temporary-enduring property of causality. Finally, all of the investigations save
those by Michela et al. (1982) and Wimer and
Kelley (1982) describe a dimension called ei-

ther control or intent (a possible distinction between these labels will be examined). In three
investigations other dimensions have emerged,
but they are not manifest in more than one
study.
The data therefore strongly support the
contention that there are three dimensions or
properties of perceived causality, which is consistent with the causal properties derived from
the logical analysis (although it must be recognized that even in the empirical studies the
dimensions require a subjective or experimenter labeling, and at times the empirical
methodologies have imposed constraints on
the causal perceptions of the subjects). The
empirical dimensions that have emerged are
reliable, general across situations, and meaningful. Other suggested dimensions are either
unreliable (perhaps intimating that they are
specific to a particular context) and/or are not
clearly meaningful, as the unnamed factor isolated by J. Meyer and Koelbl (1982).
It also seems to be the case that the structure
of causality is not merely a convenient classification system imposed by attribution theorists (see Schiitz, 1967, p. 59). The scaling and
the correlational procedures, as well as those
of factor analyses, yielded comparable dimensions corresponding to those that evolved from
the logical thinking of attribution theorists.
The dimensions, therefore, may be considered
part of lay psychology. Finally, there is a relative simplicity in the organization of causal
thinking, just as there is in the selection of specific causes.
Causal Comparisons
Recall that one of the purposes of creating
a causal taxonomy is to enable the investigator
to compare and contrast causes. Consider, for
example, low math aptitude and physical unattractiveness as respective causes of achievement failure and social rejection. Both appear
to be internal, stable, and uncontrollable
causes. Similar correspondence can be found
between other causes of achievement and social
failure such as "the school has hard requirements" versus "religious restrictions on dating" (both being external, stable, and uncontrollable). This demonstrates that the structural analysis is not limited to achievement

553

ACHIEVEMENT MOTIVATION AND EMOTION

contexts, which is an important consideration


in the search for theoretical generality.
Issues Concerning Causal Structure
Although the rational (deductive) and empirical (inductive) approaches converged and

identified the same three causal properties, a


number of pertinent questions nevertheless
remain to be addressed. Definitive answers can
be provided only to some of these questions.
Might there be less than three dimensions?
This question is guided by a reliable finding
that the dimensional ratings of causes are cor-

Table 2
Empirical Studies of Causal Dimensions
Dimensions
Domain

Locus

Stability

Control
(intent)

Factor analysis

Achievement
(hypothetical
exam of others)

Meyer, }. & Koelbl


(1982)

Factor analysis

Achievement
(examination
performance)

Wimer & Kelley


(1982)

Factor analysis

All

X'

Passer (1977)

Multidimensional Achievement
scaling
(hypothetical
exam
performance)
Failure
Success

Experiment

Procedure

Meyer, J.( 1980)

Passer, Kelley, &


Michela(1978)

Multidimensional
scaling

X?

X
X

Michela, Peplau,
& Weeks (1982)

Multidimensional
scaling

Loneliness
(hypothetical
other)

Stern (1983)

Correlation with
a priori
scheme
using
concept
formation
tasks
Free-sort
Sort-resort
Sequential sort
Graph
building
M ultidimensional
scaling

Achievement
(academic &
sports)

Good-bad; complexsimple; motivation

X
X?

Unnamed

X
X"

Marital conflict
(hypothetical
other)
Actor
Partner

* Unipolar.
b
Only internal causes.

'

Other

or X?

X
X
X
X

X
X
X
X

X
X
X
X

Attitude toward
partner
Attitude toward
partner

554

BERNARD WEINER

related. Consider, for example, a representative


study by Anderson (1983a). Anderson had
subjects generate causes for success and failure
in both achievement and social contexts. The
63 most dominant causes were then rated by
other subjects on the three causal dimensions.
These ratings were highly intercorrelated, suggesting that the dimensions are not independent and that there may be less than three underlying causal properties.
A number of arguments can be marshaled
against this position. Many causal perceptions,
particularly in social contexts, implicate traits.
Traits tend to be perceived as both internal
and stable. Inasmuch as a preponderance of
causal ascriptions then fall within an internalstable quadrant, the locus and stability dimensions will be correlated in the causal ratings.
However, as also noted by Anderson
(1983a), a failure of orthogonality at the empirical level does not invalidate separation at
the conceptual level. For example, height and
weight are positively correlated but nonetheless
are distinct characteristics; certainly tall, light
individuals as well as those who are short and
heavy can be identified. As Passer et al. (1978)
state, "There is no necessity that the elements
used in multidimensional scaling be distributed evenly over the space identified by the
analysis. In fact, there may be psychological
reasons . . . for certain regions of the space
not to contain any elements" (p. 961).
Might there be more than three dimensions?
Two other properties of causes have been suggested, intentionality (Weiner, 1979) and globality (Abramson, Seligman, &Teasdale, 1978).
These are discussed in turn.
The logical analysis of causality strongly
hints that intentionality is a causal property,
and this label also was suggested in some of
the empirical investigations. Consider, for example, a logical examination of effort versus
strategy as perceived causes of success and
failure. One might succeed because of hard
work or because of proper strategy while
studying, or fail because of insufficient effort
or poor strategy. Failure due to lack of effort
meets the criteria to infer personal responsibility, inasmuch as not trying is carried out
"purposively, knowingly, recklessly, and/or
negligently" (see Fincham & Jaspers, 1980).
But these criteria are not met given poor strategy as a cause of failure. One does not pur-

posively or knowingly use bad strategy. The


property that perhaps best describes the contrast between effort and strategy has been labeled intentionality (Werner, 1979).
Intent and control generally covary highly,
with reported correlations near r = .90 (see
Anderson, 1983a). Individuals intend to do
what is controllable, and can control what is
intended. But there are important instances
where intent and control are distinguishable.
For example, an overachiever might state that
he or she intends to take some time off from
work, but cannot control his or her working
habits. Or, one might not have intended to kill
a pedestrian, but should have controlled his or
her speeding. The differentiation between intent and control lies at the heart of the distinction between murder and manslaughter.
It seems reasonable, then, to separate control from intent and consider them both dimensions of causality. A difficult conceptual
problem, however, is created. A cause is not
intentionalintent describes an action, or a
motivational state of an organism. One might
refer to aptitude as internal, or stable, but can
it be described as unintentional? It seems not;
intent does not appear to be a characteristic
of a cause. But solving this difficult philosophical problem is beyond the scope of this article
and, even further, beyond the capability of this
writer. Thus, the possibility that intentionality
is a dimension of causality is put aside for now.
The contention of Abramson et al. (1978)
is that some causes are specific to a situation,
whereas others generalize across settings. For
example, an individual may perceive failure at
math as due to low math aptitude (specific) or
to low intelligence (general). Intelligence is
perceived as influencing performance in a
greater variety of situations than is math aptitude.
The argument in favor of a distinction between general and specific causes certainly
cannot be faulted on grounds of face validity.
To elevate this distinction to a dimension,
however, does pose some problems. A generalspecific property has not emerged in a single
empirical investigation. Thus, it is not known
whether this distinction held by some attribution theorists also is perceived or unknowingly made by the layperson.
When personality psychologists discuss
traits, both temporal aspects (consistency over

ACHIEVEMENT MOTIVATION AND EMOTION

time) and generalizability (consistency across


situations) are considered. In a similar manner,
causes can logically be construed in terms of
those two characteristics. Globality therefore
might be a basic property of causes, but more
evidence is needed before this possibility is accepted.
Is the dimensional location of a cause constant? Attributional decisions represent phenomenal causalitythe causal world as perceived by the viewer. Perceived causality certainly will differ from person to person and
within an individual over occasions. This is
true not only for a specific causal inference,
but also for the meaning or dimensional location of the cause. For one individual, luck
may be perceived as an external, unstable cause
of success; for another, luck is conceived as an
enduring personal property. Indeed, a cause
might convey different meanings in disparate
contexts (e.g., effort ascriptions connote greater
stability given success than given failure; see
Dalai, Weiner, & Brown, 1985). But although
the interpretation of specific causal inferences
might vary over time and between people and
situations, the underlying dimensions on which
causes are "understood" or given meaning remain constant. That is, dimensions are conceived as invariant, whereas the location of any
specific cause on a dimension is variable (see
Weiner, 1983).
Motivational Dynamics of Perceived
Causality: Expectancy Change
Thus far it has been suggested that individuals search for causality and that a relatively
small number of causes are particularly salient.
In addition, causes share three properties (locus, stability, and controllability) and perhaps
can be characterized according to intentionality and globality. I now turn from causal description and causal structure to the dynamics
of behavior. Two topics are of special importance in the understanding of action tendencies: expectancy and value.
Goal expectancies is a concern that keeps
reappearing in the study of motivation. Every
major cognitive motivational theorist includes
the expectancy of goal attainment among the
determinants of action. If one hopes to construct an attributional theory of motivation, it
would therefore seem necessary to search for

555

some connection, some linkage, between attributional thinking and goal expectancy.
Two possibilities come to mind. On the one
hand, the influence of causal variables on the
absolute expectancy of goal attainment could
be ascertained. Heider (1958), for example,
reasoned that goal expectancies in achievement-related contexts are determined by perceived ability and planned effort expenditure,
relative to the perceived difficulty of the task.
This is an enticing analysis to follow, inasmuch
as attributional concepts already are introduced.
But other theorists have had completely different notions about the antecedents of goal
expectancy. Tolman (1925), for example, stipulated that expectancy is a function of the frequency, primacy, and recency of reinforcement. According to Rotter (1966), expectancies
are determined by the percentage of reinforcements of a particular response in a particular
setting, the percentage of reinforcements of this
response in similar situations, and individual
differences in the belief that reinforcements are
under personal control. And for Atkinson
(1964), expectancy is influenced by the number of individuals against whom one is competing, prior reinforcement history, and communications from others concerning the likelihood of success. It therefore is evident that
consensus does not exist about the antecedents
of goal expectancy, although all theorists would
agree that past reinforcement history does play
some important role.
A second possibility is initially to find relations between attributions and changes in
expectancy, and then use this information to
determine the relation between causal ascriptions and absolute expectancy of success. Inasmuch as attributional search is initiated following an outcomeso that change can be examinedthis seems to be a promising lead.
In addition, perhaps change in goal expectancy,
as opposed to absolute expectancy level, is
more amenable to a general law that transcends
the situational context.
Investigations of Expectancy Change
Three psychological literatures are directly
related to changes in goal expectancy, and
again the pertinent research has primarily been
conducted in achievement-related contexts.

556

BERNARD WEINER

One set of investigations is associated with level


of aspiration; the second concerns the effects
of outcomes at chance tasks on probabilities
of future success; and the third research endeavor is linked with resistance to extinction
and beliefs about locus of control.
Level of aspiration. A number of quite replicable findings emerged from level of aspiration research. Among the most important for
present purposes is that subsequent aspiration
level is in part dependent on the prior outcome.
In the vast majority of instances, aspiration
increases after goal attainment and decreases
if a prior aspiration has not been fulfilled.
These so-called goal discrepancies are referred
to as "typical" aspiration shifts.
It has been assumed that aspiration level in
good part reflects the subjective expectancy of
success: The higher the expectancy, the higher
the aspiration level. Hence, the aspiration literature can be interpreted as revealing that increments in expectancy follow success, whereas
expectancy decrements follow failure. This
conclusion also has been documented extensively in contexts where expectancy is directly
measured, rather than inferred from statements about goal aspiration (see, e.g., Diggory,
Riley, & Blumenfeld, I960; Montanelli & Hill,
1969; Zajonc & Brickman, 1969).
This is not the complete story, however, for
in games of skill "atypical" reactions also are
sometimes observed. In these instances, there
is a decrease in aspiration level following success and an increase after failure. For example,
Lewin, Dembo, Festinger, and Sears (1944)
noted,
In the case of nonachievement which is linked, for instance,
to outside disturbances, the subject is not likely to lower
his aspiration in a way that he would if he believed that
the nonachievement reflected a genuine decrement in his
performance ability, (p. 367)

Chance tasks. A divergent pattern of data


emerged from research on the subjective probability of success at games of chance. Here the
gambler's fallacy often is observed. That is,
after winning, a loss is expected, and after losing, a win is anticipated (see Cohen & Hansel,
1956). A related phenomenon at games of
chance is labeled the negative recency effect.
This is illustrated in the increased expectancy
of a heads after the appearance of a tails on a
coin toss. That is, atypical shifts are frequent

at games of chance. There are, however, some


exceptions to this general rule. At times, gamblers exhibit the belief that they are on a winning or losing streak and anticipate repetitions
of the prior win or loss. Thus, typical shifts
also are observed in chance settings, but with
less frequency than are atypical shifts. Note
that this is the mirror image of the data pattern
given skill tasks.
Social learning theory integration. The
problem is to create a conceptual framework
able to incorporate the observations of typical
and atypical shifts in situations of skill and
chance. Social learning theorists attempted to
do just that, primarily examining resistance to
extinction while manipulating skill and chance
task perceptions. They contended that expectancy change following success or failure is influenced by the perceived locus of control of
the outcome, with internal or personal beliefs
about causality (skill tasks) producing typical
shifts, while external perceptions of causality
(chance tasks) generate atypical shifts. In addition, given that some individuals might perceive skill tasks as determined by chance, and
chance tasks as determined by personal factors,
occasional reversals in the usual pattern of data
would be observed. In sum, social learning
theorists were the first to relate the structure
of perceived causality (the locus dimension) to
expectancy change (see Rotter, 1966).
Attributional Approach to
Expectancy Change
In this article I have reasoned, however, that
Rotter and his colleagues gave insufficient attention to the richness of causal explanation
and confounded dimensions of causality.
Ability (skill), in addition to being internal,
also is perceived as relatively stable. On the
other hand, in addition to being external, luck
is perceived as relatively unstable. Hence, ability and luck differ in subjective stability and
not merely on the locus dimension of causality.
The observed differences in expectancy shifts
given skill versus chance tasks may therefore
either be attributed to the locus or to the stability dimension of causality.
The attributional position is that the stability
of a cause, rather than its locus, determines
expectancy shifts. If conditions (the presence
or absence of causes) are expected to remain

ACHIEVEMENT MOTIVATION AND EMOTION

the same, then the outcome(s) experienced in


the past will be expected to recur. A success
under these circumstances would produce relatively large increments in the anticipation of
future success, and a failure would strengthen
the belief that there will be subsequent failures.
On the other hand, if the causal conditions are
perceived as likely to change, then the present
outcome may not be expected to repeat itself
and there is likely to be uncertainty about subsequent outcomes or a belief that something
different will result. A success therefore would
yield no increments in subsequent expectancy
and could give rise to decrements in the subjective probability of future success. Similarly,
a failure will not augment the belief that there
will be future failures.
These principles are able to explain the data
in level of aspiration research and in studies
involving chance tasks. Success and failure at
skill tasks most usually are ascribed to ability
and effort. Ability is thought to be a relatively
fixed property, and the belief that success was
caused by hard work usually results in the intent to again work hard in the future (Dalai et
al., 1985). Inasmuch as the causes of a prior
success are perceived as relatively stable given
skill-related tasks, future success should be
anticipated with greater certainty and there will
be increments in aspiration level and expectancy judgments. Occasionally, however, outcomes at skill tasks are ascribed to unstable
factors, such as the "disturbances" noted by
Lewin et al. (1944). In addition, if failure is
ascribed to low effort, then the failing person
may anticipate working harder in the future.
In these circumstances there would be atypical
or minimal shifts in expectancy following
failure.
Conversely, success at chance tasks tends to
be ascribed to an unstable factor. The actor is
likely to reason, "I had good luck last time,
but that probably will not happen again." Expectancy therefore should not rise and indeed
could drop following a positive outcome. But,
occasionally, one might conclude that he or
she is a lucky or an unlucky person or is on a
winning or losing streak. In these instances,
the cause of the outcome is perceived as stable,
so that typical shifts will be displayed. In sum,
the attributional position can account for the
observed typical and atypical shifts in chance
as well as in skill settings.

557

These ideas gave rise to a wealth of pertinent


research, primarily in achievement-related
contexts. Two research strategies were representedcorrelational, and the manipulation
of causal ascriptions. In the correlational research, subjects were induced to succeed or
fail at some laboratory task and their expectancies of future success as well as causal ascriptions were assessed. In the causal manipulation procedure, perceptions of task outcomes as caused by ability, effort, luck, and
so forth were induced, and expectancy of success was ascertained following success or failure. Table 3 includes a considerable (but far
from complete) sample of investigations reporting that stable, relative to unstable, ascriptions are related to high expectancies of
success after goal attainment and to low expectancies of success following a failure. None
of the studies are definitive in that often other
possible interpretations of the data have not
been ruled out; nonetheless, the consistency of
the findings is quite compelling.
In addition to the data reviewed in Table 3,
a number of "real-life" behaviors without apparent phenotypic overlap have been shown to
be genotypically comparable and explainable
given an attributional perspective. These studies have taken place in nonachievement as well
as in achievement-related contexts. More specifically, for example:
1. Parole decisions are in part based on the
perceived stability of the cause of the crime
(Carroll, 1978; Carroll & Payne, 1976, 1977).
A criminal perceived as committing a crime
because of stable causes (e.g., a psychopathic
personality) is considered more of a risk than
is one perceived as committing a crime because
of unstable causes (e.g., temporarily unemployed). In a similar manner, criminals as well
as prison officials believe a crime due to stable
causes is more likely to be repeated when the
offender is released than is a crime perceived
as due to unstable causes (Saulnier & Perlman,
1981). Criminals judge the causes of their
crimes to be more unstable than do prisonrelated officials; they also are less likely to expect recidivism.
2. Rejection when soliciting blood from
potential donors that is ascribed to unstable
causes (poor soliciting strategy, lack of effort)
results in higher expectancy of success and
greater persistence at solicitation than does re-

558

BERNARD WEINER

jection attributed to stable factors such as an


inability to persuade others (Anderson, 1983b;
Anderson & Jennings, 1980).
3. Intentions to reenter school after dropping out, to enroll in another psychology
course although doing poorly in one, or to resubmit a rejected manuscript for publication
are greater when the prior "failures" are ascribed to unstable causes (e.g., need a break
from school, poor manuscript reviewers) rather
than to stable causes (e.g., cannot benefit from

school, poor research training; see Day, 1982;


Pancer, 1978; Grittended & Wiley, 1980).
4. When product failure is perceived as due
to stable factors (e.g., a bad company), consumers anticipate future product dissatisfaction and express a desire for a monetary refund; if the failure is perceived as due to unstable causes (e.g., bad shipment of goods),
then subsequent product satisfaction is expected and there is a preference for product
exchange (Folkes, 1984).

Table 3

Research Relating Attributions to Expectations


Experiment

Subjects

Attribution
measurement

Task

Expectancy
measure

Correlational studies
Meyer, W. (1973)

German high
school students

Digit-symbol
substitution

Percentage rating

Probability of
future success

McMahan(1973)

American
grammar, high
school, &
college students

Anagrams

Paired comparison

Confidence of
future success

Weiner, Nierenberg,
& Goldstein
(1976)

American college
students

Block design

Within-dimension
scale rating

Anticipated performance

Inagi(1977)

Japanese college
students

Puzzle

Percentage rating

Probability of
future success

Kovenklioglu &
Green haus
(1978)

American college
students

Test performance

Paired comparison

Anticipated performance

Ronis, Hansen, &


O'Leary(1983)

American college
students

Unspecified
achievement
stories

Dimension scale
rating

Confidence in
outcome
repetition

Research manipulating attributions


Rosenbaum(1972)

American college
students

Unspecified "project"
(hypothetical)

Causes given in
description

Expected project
outcome

Neale & Friend


(1972)

American college
students

School exam
(hypothetical)

Causes given in
description

Anticipated grade

Fontaine (1974)

Australian college
students

Unspecified "tasks"

Fictitious ascriptions
of others

Expected score

Valle(1974)

American college
students

Sales (hypothetical)

Causes given in
description

Anticipated performance

Pancer & Eiser


(1977)

British college
students

Anagrams

Fictitious
information from
others

Performance
prediction

Heilman & Guzzo


(1978)

American college
students

Job performance
(hypothetical)

Causes given in
description

Predicted personnel action

ACHIEVEMENT MOTIVATION AND EMOTION

Summary and Conclusions


Individuals classify their thoughts into broad
categories. Hence, phenotypic dissimilarities
might be connotatively, or genotypically, similar. Failure in athletics because of lack of
height, failure in math because of low aptitude,
failure in politics because of poor charisma,
and social rejection because of unattractive
features are phenotypically different events
with diverse specific causes. Yet the causes are
likely to be similarly categorized as enduring
or stable. Hence, future hopes in these heterogeneous contexts will be minimized. On the
other hand, failure in athletics because of insufficient practice, failure at math because of
temporary illness, failure in politics because
of a current recession, and interpersonal rejection because the desired partner is ill are
diverse events that are likely to be categorized
as due to unstable causes. Hopes for the future
therefore are likely to be maintained.
The amount, extensity, and consistency of
the empirical findings, in conjunction with the
logical analysis, documents a fundamental
psychological law relating perceived causal
stability to expectancy change:
Expectancy Principle. Changes in expectancy of success following an outcome are influenced by the perceived stability of the cause
of the event.
This principle has three corollaries:
Corollary 1. If the outcome of an event is
ascribed to a stable cause, then that outcome
will be anticipated with increased certainty, or
with an increased expectancy, in the future.
Corollary 2. If the outcome of an event is
ascribed to an unstable cause, then the certainty or expectancy of that outcome may be
unchanged or the future may be anticipated
to be different from the past.
Corollary 3. Outcomes ascribed to stable
causes will be anticipated to be repeated in the
future with a greater degree of certainty than
are outcomes ascribed to unstable causes.

559

tainly affect other thoughts and actions. But it


also is quite evident that goal expectancies are
not sufficient determinants of action. After all,
there are an infinite number of actions not undertaken in which expectancy of the goal is
absolutely certain.
Both cognitive and mechanistic conceptions
of behavior have identified another class of
variables with motivational impact. They are
called goal incentives, or the properties of the
goal object. Motivation is believed to be determined by what one can get (incentive) as
well as by the likelihood of getting it (expectancy). This is the essence of the position of
Expectancy X Value theorists.
There seems to be no blatant reason to believe that objective value (i.e., the inherent
properties of a goal object) is influenced by
perceived causalitythe reason why the goal
was reached. A dollar has the value of one dollar whether it is attained because of good fortune, hard work, or as a gift from another. On
the other hand, instead of conceiving incentive
values in terms of the objective properties of
the goal, consider incentive to mean the consequences of goal attainment for the actor, or
the subjective value of the goal. We prefer a
dollar to a nickel because the anticipated consequences will make us happier, give greater
satisfaction, and the like. Although causal ascriptions do not influence the objective properties of goal objects, they do determine or
guide emotional reactions, or the subjective
consequences of goal attainment. For example,
a dollar attained because of good luck could
elicit surprise; a dollar earned by hard work
might produce pride; and a dollar received
from a friend when in need is likely to beget
gratitude. In a similar manner, a gift from a
beloved will have different affective significance
than does this same gift from an enemy (Heider, 1958). These diverse affective reactions
could generate quite disparate actions. For example, gratitude but not pride might give rise
to reciprocal helping behavior. It therefore
seems reasonable to pursue the idea that causal
ascriptions influence emotions, and that emotional reactions play a role in motivated behavior.

Motivational Dynamics of Perceived


Causality: Affective Reactions
It has been rather definitively documented
that causal attributions influence expectancy
of success. This is a necessary linkage for the The Attribution-Emotion Process
development of an attributional theory of moThe field of emotion is vast and complex;
tivation, inasmuch as goal anticipations cer- the formulation of a complete theory of emo-

560

BERNARD WEINER
.General positive or negative emotions
Outcome

-Outcome evaluation

Causal attribution and dimensions -

-Distinct emotions

Figure I. The cognition-emotion process.

tion is not my goal. Rather, the aims of this


section of the article are to offer an attributional view of the emotion process and to propose and document laws linking attributional
thinking and specific feelings (for a discussion
of the assumptions guiding this approach to
emotion see Weiner, 1982; Weiner & Graham,
1984).
Most emotion theorists with a cognitive
persuasion conceive of emotional experience
as a temporal sequence involving cognitions
of increasing complexity. Arnold (1960) and
Lazarus (1966), for example, contend that the
perception of a distal stimulus gives rise to a
primary appraisal and to a rather primitive
emotional reaction. Primary appraisal is believed to be followed by a secondary appraisal
that often involves ego-related or more advanced psychological mechanisms such as ego
defenses. The elicitation of these processes can
intensify or modulate the emotional experience or alter the quality of the emotion.
Schachter and Singer (1962) proposed the
most oft-cited emotion sequence. They hypothesized that the initial step in this sequence
is the experience and recognition of nondifferentiated internal arousal. Then the source
of the arousal is determined on the basis of
situational cues, and this cognitive labeling
plus the arousal give rise to emotional states
(although the cognition of the arousal and the
cognition of the source of the arousal often
take place simultaneously).
The attributional framework advanced here
also assumes a sequence in which cognitions
of increasing complexity enter into the emotion process to further refine and differentiate
experience. It is contended that, following the
outcome of an event, there is a general positive
or negative reaction (a "primitive" emotion)
based on the perceived success or failure of the
outcome (the "primary appraisal"). These
emotions, which include happy for success and
frustrated and sad for failure, are labeled as
outcome dependent-attribution independent,
for they are determined by the attainment or

nonattainment of a desired goal, and not by


the cause of the outcome.
Following outcome appraisal and the immediate affective reaction, a causal ascription
will be sought. A different set of emotions is
then generated by the chosen attribution(s).
For example, success perceived as due to good
luck produces surprise, whereas success following a long-term period of effort expenditure
results in a feeling of calmness or serenity.
Emotions such as surprise and serenity are labeled attribution dependent, inasmuch as they
are determined by the perceived cause of the
prior outcome. Note that increasing cognitive
involvement generates more differentiated
emotional experience (for somewhat related
conceptions see Abelson, 1983; Roseman,
1984; C. Smith & Ellsworth, 1985).
Additionally, causal dimensions play a key
role in the emotion process. Each dimension
is uniquely related to a set of feelings. For example, success and failure perceived as due to
internal causes such as personality, ability, or
effort respectively raises or lowers self-esteem
or self-worth, whereas external attributions for
positive or negative outcomes do not influence
feelings about the self. Hence, self-related
emotions are influenced by the causal property
of locus, rather than by a specific cause per se.
The cognition-emotion process suggested in
the above paragraphs is depicted in Figure 1.
It is evident from Figure 1 that this approach
to emotions assumes that feelings arise from
how an event is construed or evaluated. The
figure also indicates that general and distinct
emotions are independent, but this remains to
be fully resolved (see, e.g., McFarland & Ross,
1982).
The cognition-emotion process that has
been proposed provides the focus and outline
for the following pages. I first briefly examine
outcome-related affects and then consider in
detail the associations between causal dimensions and affects, ignoring here the relations
between specific causes and emotional reactions (see Weiner, Russell, & Lerman, 1978,

ACHIEVEMENT MOTIVATION AND EMOTION

561

1979). I focus on dimension-linked affects be- Dimension-Related Emotions


cause they have had the most extensive emAs previously indicated, the bulk of the perpirical support. These associations, just as the
one between causal ascription and expectancy tinent attribution-emotion research relates
change, form powerful and general laws.
causal dimensions, rather than specific causes,
to affects. The emotion of pride and feelings
of self-esteem are linked with the locus diOutcome-Generated Emotions
mension of causality; anger, gratitude, guilt,
Two research paradigms, one simulational pity, and shame all are connected with the
and reactive, the other retrospective and op- controllability dimension; and feelings of
erant, were first used to document the asso- hopelessness (hopefulness) are associated with
ciations between outcomes and emotions. causal stability. These relations are described
Again these studies were conducted in here, but without detailed documentation (see
achievement-related contexts. In the former Weiner, 1982, and Weiner & Graham, 1984,
paradigm, participants were asked to imagine for fuller discussions).
that a student succeeded or failed an exam for
Pride (self-esteem).
A relation between
a particular reason, such as hard work or bad causal locus and self-esteem has been long recluck. The subjects then reported the intensity ognized by many well-known philosophers.
of the affective reactions that they thought Hume, for example, believed that what one is
would be experienced in this situation (Weiner proud of must belong to the person; Spinoza
et al., 1978). Intensity was indicated on rating reasoned that pride consists of knowing one's
scales for a number of preselected affects. In merits; and Kant nicely captured the locusthe second paradigm, participants were asked pride union by noting that everyone at a meal
to recall a time in life when they succeeded or might enjoy the food, but only the cook of that
failed for a specified reason. They also re- meal could experience pride.
counted the affects they experienced at that
It is therefore reasoned that pride and postime (Weiner et al., 1979).
itive self-esteem are experienced as a conseThese studies revealed that one determinant quence of attributing a positive outcome to
of affect is the outcome of an action: Success the self and that negative self-esteem is expeat achievement-related activities was associated rienced when a negative outcome is ascribed
with the affect of happy regardless of the cause to the self (Stipek, 1983; Weiner et al., 1978,
of that outcome, and failure seemed to be re- 1979). The relation between causal locus and
lated to frustration and sadness. Thus, for ex- feelings of self-worth also is part of naive psyample, given athletic competition, one tends chology and is used by the layperson to influto experience happiness following a victory ence the emotions of others. Thus, individuals
whether that win was due extra training, the tend to communicate ascriptions external to
poor play of the competitor, or good luck. the requester when rejecting that person for a
Outcome-dependent affects also have been social engagement so that "feelings of self-esdocumented in quality of life research. Inves- teem are not hurt" (e.g., they indicate that they
tigators have reported that satisfaction, un- are ill rather than truthfully telling the rehappiness, and frustration are related to ob- quester that he has a poor personality; see
jective life outcomes, such as income level, in- Folkes, 1982). Children as young as 5 years of
dependent of attributions (see Bryant & Veroff, age have demonstrated an understanding of the
1982; E. Smith & Kluegel, 1982). In addition, relation between causal locus and hurt feelings,
outcome-dependent affects have been postu- given a rejection (Weiner & Handel, 1985). A
lated in the interpersonal domain. Kelley voluminous attributional literature also doc(1983) stated,
uments existence of a hedonic bias, or a tenI am pleased or displeased by the more specific and concrete dency for individuals to ascribe success to inthings I experience [in close relationships]. So when my ternal factors and failure to external factors.
wife prepares a picnic lunch for the afternoon's outing, As Harvey and Weary (1981) noted, "By takmy pleasure-displeasure comes partly from the quality of
the lunch itself, and also (as a partly separate matter) from ing credit for good acts and denying blame for
the quality of love and thoughtfulness I attribute to her bad outcomes, the individual presumably may
effort, (p. 15)
be able to enhance or protect his or her self-

562

BERNARD WEINER

esteem" (p. 33). Pride and personal esteem


therefore are self-reflective emotions, linked
with the locus dimension of causality.
Anger. A large survey study by Averill
(1982, 1983) illustrates the attributional antecedents of anger. Averill asked his respondents to describe a situation in which they were
made angry, and then examined the characteristics of these situations. He concluded,
The major issue for the person in the street is not the specific
nature of the instigating event; it is the perceived justification for the instigator's behavior. Anger, for the person
in the street, is an accusation . . . Over 85% of the episodes
described by angry persons involved either an act that they
considered voluntary and unjustified (59%) or else a potentially avoidable accident (e.g., due to negligence or lack
of foresight, 28%) . . . More than anything else, anger is
an attribution of blame. (Averill, 1983, p. 1150)

Many others have reached a similar conclusion. For example, among the very first of the
pertinent investigations, Pastore (1952) demonstrated that aggression (and, by implication,
anger) is not merely the result of nonattainment of a desired goal, but rather follows when
a barrier imposed by others is arbitrary (e.g.,
"Your date phones at the last minute and
breaks an appointment without adequate explanation") rather than nonarbitrary (e.g.,
"Your date phones . . . and breaks an appointment because he (she) suddenly became.
ill)." To summarize, the attributional antecedent for anger is an ascription of a negative,
self-related outcome or event to factors controllable by others (see Weiner, 1980a, 1980b;
Weiner, Graham, & Chandler, 1982).
Pity. In contrast to the linkage between
controllability and anger, it is hypothesized that
uncontrollable causes are associated with pity.
It is said that when Helen Keller began her
training, her teacher stated to Ms. Keller's
family: "We do not want your pity," thus conveying that a target of pity is associated with
an uncontrollable deficit. This analysis is similar to Hoffman's (1982) conception, for he
stated, "It is only when the cues indicate that
. . . the victim had no control that the . . .
partial transformation of empathic into sympathic distress may apply" (p. 296).
A number of research studies support this
contention. Another's loss of a loved one because of an accident, or difficulties because of
a physical handicap, are prototypical situations
that elicit pity (see Graham, Doubleday, &

Guarino, 1984; Weiner, 1980a, 1980b; Weiner,


Graham, & Chandler, 1982). Note, therefore,
that the perceived controllability of a cause for
a negative outcome in part determines whether
anger or pity is directed toward another. We
feel anger toward the lazy and therefore punish
lack of effort, but we feel pity toward the unable and therefore do not punish lack of ability
(Weiner & Kukla, 1970).
The relations between controllability-anger
and uncontrollability-pity also are part of naive psychology and are used in everyday life to
control or manipulate the emotions of others.
Thus, when providing an excuse (ex = from;
cuse = cause) for failing to appear at a social
engagement, uncontrollable causes tend to be
communicated (e.g., "My car had a flat tire")
rather than controllable ones (e.g., "I decided
to watch TV"; see Weiner, Amirkhan, Folkes,
& Wachtel, 1985). One hopes that this communication defuses anger and perhaps even
alters the reaction to pity. Similar interpersonal
strategies are understood and used by children
as young as 5 years of age (Weiner & Handel,
1985).
Guilt and shame. Philosophers and social
scientists have devoted considerable attention
to the experience of guilt, its antecedents, and
its consequences. Reviewing the guilt literature, Wicker, Payne, and Morgan (1983) concluded, "In general, guilt is said to follow from
acts that violate ethical norms, principles of
justice . . . or moral values. Guilt is accompanied by feelings of personal responsibility"
(p. 26). In a similar manner, Izard (1977) concluded that "Guilt occurs in situations in
which one feels personally responsible"
(p. 423), and Hoffman (1976) more precisely
reasoned, "Blaming oneself becomes possible
once one has acquired the cognitive capacity
to recognize the consequences of his action for
others and to be aware that he has choice and
control over his own behavior" (p. 139). In
support of these interpretations, my colleagues
and I have found that the most prevalent guilteliciting situations among college students involve lying to parents, cheating on an exam,
or being disloyal to a dating partner (Weiner,
Graham, & Chandler, 1982), although it is evident that guilt may be evoked by either the
commission or the omission of particular actions (see Hoffman, 1970).
Guilt and anger therefore are elicited by

ACHIEVEMENT MOTIVATION AND EMOTION

controllable causes, but guilt is directed inward, whereas anger is typically (but not necessarily) directed outward. Thus, for example,
we tend to feel guilty when we have lied to
others, but angry when we have discovered that
someone has lied to us (see Weiner, Graham,
& Chandler, 1982). In a similar manner, lack
of effort toward an important goal tends to
elicit anger from others (such as teachers) and
also generates personal guilt.
Shame frequently is contrasted with guilt,
although both involve "negative self-evaluations that are painful, tense, agitating. . . depressing" (Wicker et al., 1983). Although there
appear to be different kinds of shame, it is believed that one antecedent is an attribution for
failure that is self-related and uncontrollable,
such as lack of ability. In studies testing uncontrollability-shame and controllability-guilt
associations, Brown and Weiner (1984), Covington and Omelich (1984), and Jagacinski and
Nicholls (1984) have reported that shame-related affects (disgrace, embarrassment, humiliation, and/or shame) are linked with failure due to low ability, whereas guilt-related
affects (guilt, regret, and/or remorse) are associated with failure due to lack of effort. It
also has been documented that shame-related
emotions give rise to withdrawal and motivational inhibition, whereas guilt-related emotions promote approach behavior, retribution,
and motivational activation (Hoffman, 1982;
Wicker et al., 1983). Hence, there are linkages
between low-ability-shame-inhibition and between lack-of-effort-guilt-augmentation. It
also is of interest to repeat that anger tends to
motivate aggression, so that three patterns of
behavior noted by Horney (going toward, going
away from, going against) are related to causal
controllability and the respective affects of.
guilt, shame, and anger.
Gratitude. There is relatively little research
concerned with gratitude, but the evidence
suggests that gratitude toward another is elicited if and only if the act of the benefactor was
under volitional control and was intended to
benefit the recipient. For example, Tesser,
Gatewood, and Driver (1968) presented subjects with scenarios that involved a benefactor
and asked the subjects how grateful they would
feel under the various circumstances that were
portrayed. They found that reported gratitude
was maximized when the gift was intended to

563

benefit only the receiver (as opposed to a situation in which the gift enhanced the reputation of the giver). In other supporting research it has been documented that reciprocity
is more likely when a gift is given deliberately
rather than accidentally (Greenberg & Frisch,
1972) and when help is voluntary rather than
compulsory (Goranson & Berkowitz, 1966).
Hopelessness. It has been convincingly
documented that causal stability in part determines expectancies regarding future success
and failure. Thus, any emotion involving anticipations of goal attainment or nonattainment will likely be influenced by causal stability. One such affect has been labeled hopelessness. It has been found that hopelessness
and resignation are elicited given an attribution
for a negative outcome to stable causes (Weiner
et al., 1978, 1979). That is, if the future is anticipated to remain as bad as the past, then
hopelessness is experienced. In addition, affects
such as pity are exacerbated when the cause
of the negative state is stable rather than unstable (e.g., we tend to pity the blind more than
we pity those with temporary eye problems).
Similarly, we tend to be more angry at others
when perceived controllable behavior, such as
lack of effort, is stable (a trait) rather than an
unstable state (Weiner, Graham, & Chandler,
1982).
Summary and conclusions. Attributions
play a key role in affective life. Seven emotions
were briefly examined that relate to causal
structure: pride (self-esteem), anger, pity, guilt,
shame, gratitude, and hopelessness. These are
among the most frequently reported and written-about affective experiences (see Bottenberg, 1975; Davitz, 1969). Sociobiologists have
specified that four of these emotionsanger,
pity, guilt, and gratitudeare of special importance in promoting gene survival (see Trivers, 1971). These four emotions are related to
the causal dimension of controllability, which
is consistent with the sociobiological position
that emotions are used to aid in maintaining
the social order. Finally, although the bulk of
the supporting data have been generated in
achievement-related contexts, the relations
specified above do not seem to be confined to
a particular motivational domain. As previously suggested, aptitude as a cause of achievement success, and physical attractiveness as a
cause of social success, are conceptually similar

564

BERNARD WEINER

in that both are internal, stable, and uncontrollable. Thus, success or positive outcomes
due to these factors should enhance pride and
positive self-esteem, just as negative outcomes
because of their absence should lower esteem.
But nonattainment of a goal for these reasons
should provoke neither anger from others nor
personal guilt. Rather, failure given both these
specific ascriptions is likely to elicit pity from
others and produce feelings of shame and
hopelessness in the frustrated individual.
A word of caution, however, is needed about
the preceding discussion. Given a causal ascription, the linked emotion does not necessarily follow. For example, one may not have
put forth effort at something important, yet
still be free from guilt. Or one may attribute
success to help from others, yet not feel grateful. Furthermore, an emotion may be experienced in the absence of its linked antecedent.
For example, one may not be responsible for
an outcome, but will experience guilt (see
Hoffman, 1976). Hence, the position being espoused is that the dimension-affect relations
are not invariant, but are quite prevalent in
our culture, and perhaps in many others as
well. This position is similar to the argument
that there is a linkage between frustration and
aggression, although frustration elicits reactions other than aggression, and aggression has
other antecedents in addition to frustration.
The Complete Theory
It is now possible to present an attributional
theory of motivation and emotion based on
the prior discussion of the theoretical components. The theory is presented in Figure 2.
In contrast to other Expectancy X Value approaches, this conception is represented as a
historical or temporal sequence; motivation is
not conceived as an "ahistorical problem"
(Atkinson, 1964, p. 146). In addition, the theory to be proposed departs from prior Expectancy X Value conceptions by linking value to
the affect elicited following goal-directed activity. Other theories of motivation have been
remiss by virtually ignoring the emotions, save
for an acceptance of the general pleasure-pain
principle. The sequence depicted in Figure 2
will be used to discuss the following contrived
(but surely extant) scenario: "A Little League
baseball player performs very poorly during a

game. Instead of appearing for the next contest, the boy stays at home." Other scenarios,
such as the boy taking extra batting practice
following failure (rather than missing the game)
or taking extra batting practice after playing
well (success), could have readily been used to
portray how the theory shown in Figure 2 conceptualizes an achievement-related motivational episode. This is followed by an examination of achievement change programs, for
these therapeutic attempts illustrate both how
the theory has been used and document its
incomplete utilization. After these analyses, I
consider the generality of the theory beyond
the achievement domain.
Figure 2 reveals that a motivational sequence is initiated by an outcome that individuals interpret as positive (goal attainment)
or negative (nonattainment of the goal). Inasmuch as affects are directly linked with outcomes (the primary appraisal), Figure 2 includes a connection between outcome and the
reactions of happy (for success) and frustrated
or sad (if the outcome was interpreted as a
failure). These associations are designated with
a 1 in the figure. In the baseball scenario, the
boy performed poorly at the game and this
will elicit general negative reactions.
A causal search is then undertaken to determine why the outcome occurred (Linkage
2). Some of the conditions that particularly
promote this search, which were not discussed
in the present article (see Weiner, 1985), are
indicated in the figure. In our example, failure
at a subjectively important act should result
in the boy overtly or covertly wondering, "Why
did I perform so poorly?" A large number of
antecedents influence the causal explanation(s)
reached. This popular topic also was not discussed in the present article. Some of the
known attributional antecedents are included
in Figure 2, such as specific information (e.g.,
past personal history, performance of others;
see Kelley & Michela, 1980). The blanket etcetera at the bottom of the antecedents merely
conveys that there are many unlisted determinants of the selected attribution.
The causal decision is biased toward a relatively small number of causes such as ability
and effort in the achievement domain (see Table 1). Again Figure 2 is not complete, as denoted by the etcetera at the bottom of the
causal lists. In our example, assume that the

565

ACHIEVEMENT MOTIVATION AND EMOTION

S
5(A
-8
2S
1!
S S
S3

B 00 0
41 B 41
B i4 >0

.
\

>.1

<0
B

ar uxS?

oo 8 >

S
,v.

C > -H 41
*4 41 1* .-4
0.-H *J 0

**

-H

1-4
.e
1* U
u
y

<J

<

c -^
*> ** w u

O M

^ J 0. U

/k

,
o
4>

u
u

i
J

'

4>

1
u

3
"41 09

S
4*
W

41
4) |
T3 "*4
*4 *-4
U 4J

U
U
4)
4t
>
IH
-H
*-
t)
U
I
4 > * M

B
V

O.JS

ss

00

(X

J2

**

U Q

jo

<-<

W h Q

<H Q
4 -H
3 U
4 SO

C
41

O
i-l
U
4
^4
~4

4 U 4
U 4 B
^ M0
4 4
XJB **
f U 41

*4
*W

ft,

<3

A
4
^4
i-t
4
>

0. X

*^

"rt

4
U
M4 .
O O
W
U

"C

1
> V

u
Sfi

0 S

^ o

i2

D:

Ihs
3

CN/T*

,'S 8.
C224
u
U

**

U ^J

B
41

u|
i-t C

*
1

0*

1
|

"t

S 2
25

32

M W <*4 41
H -H -4 00

< bj MH >j vi

A U
0) 4)

t*
3s
S's

f4

SxS'.,

1 U

-H

n
u

X
M
4V* U H
-H W *>
-4 0 4jrf Jl
-H IM t- *t 0 O
JO <M W 4 3 U

B>

m
TJ

o
E

X
*J

1 5

'
B

.0

"3 > o

-t
C
O

5*35^

SSx

*
X

^. X

\\ //
\ /

i
-4 tO
4 B

'S

f t"\\

A\

41

sl 81 !

w
r*

ij
U

"t

(0

8:3

B
0
*4 41

4 v4
3 fci
CL
r-t U
4)3

B 41
tO 6
W -<
41 3
^4 M
U
(X a
O Q

41

3J

51

"SI

S X
(X CX

B 4

M J5

M-Hl

Si'

566

BERNARD WEINER

boy has played quite poorly in the past and


that other children on the team are playing
well. The boy also practiced many hours. On
the basis of the past outcome history, social
comparison, and effort expenditure, the boy
decides that he is low in baseball-playing ability. That is, he thinks, "I failed because I am
not any good at baseball" (Linkage 3). A
unique affective reaction may be elicited by
this causal decision (Linkage 4).
The cause is then located in dimensional
space. This is depicted as Linkage 5 in the figure. As documented in Table 2, the three main
properties of causes are locus, stability, and
controllability, with globality and intentionality considered possible causal properties (and
therefore accompanied by question marks).
The Little Leaguer ascribed his performance
to lack of ability, which is likely to be perceived
as internal, stable, and uncontrollable (although that placement must be analyzed from
the phenomenology of the perceiver). It also
might be unintentional and global ("I am poor
at sports").
Causal dimensions have psychological consequences, being related to both expectancy
and affect (which is presumed in this conception to be the value of goal attainment). The
stability of a cause influences the relative expectancy of future success (Linkage 6). This
association is documented in Table 3. In our
scenario, the boy anticipates repeated failure
inasmuch as low ability is perceived as a stable
cause. He also might have increased expectancy of failure in other sporting activities if
the cause is perceived as global. That is, stability influences temporal aspects of expectancy, whereas globality influences cross-situational expectancies.
Turning to affective consequences, the locus
of a cause exerts an influence on self-esteem
and prideinternal ascriptions elicit greater
self-esteem for success and lower self-esteem
for failure than do external attributions (Linkage 7). The boy in our story failed because of
a cause considered internal, and therefore he
should be experiencing low self-esteem. The
stability of the cause, by affecting expectancy,
also fosters feelings of hopelessness (or hopefulness); this is indicated in Linkage 8. The
Little Leaguer, with a history of failure and
ascription of the current failure to low ability,
should be feeling hopeless. Finally, controllability influences social emotions; controllable

causes of personal failure promote feelings of


guilt, whereas uncontrollable causes generate
sham,e (Linkage 9). These are represented in
the figure as self-directed affects, as are the
specific attribution-linked emotions of relaxation and surprise. Among the affects directed
toward others are anger (given a cause of failure
controllable by others), pity (given an uncontrollable cause of failure), and gratitude (given
a controllable cause; Linkage 10). The failing
Little Leaguer is likely to be feeling ashamed
of himself and humiliated (but not guilty),
whereas his coach or his mother feels pity or
feels sorry for him (but not angry).
Finally, expectancy and affect are presumed
to determine action (Linkages 11,12, and 13).
The actions can be described according to their
intensity, latency, and so on. In the baseball
scenario, the boy has a low expectancy of future success and is feeling sad, low self esteem,
ashamed, and hopeless. These conditions promote withdrawal and behaviors that are not
instrumental to the attainment of the desired
goal. He then stays home from the next game.
Although Figure 1 appears to depict an exhaustive conceptual analysis, even greater
complexity has been documented. The linkages in the figure all are unidirectional, although it is known that this is not the case.
For example, expectancy of success influences
attributions (see Feather & Simon, 1972).
Thus, if our baseball player succeeded, his low
expectancy of success would foster an attribution to an unstable cause such as good luck.
In addition, affects such as pity and anger are
important attributional cues (Graham, 1984;
Weiner, Graham, Stern, & Lawson, 1982). For
example, directing pity toward the Little Leaguer will increase his belief that personal failure
was due to low ability. And feelings of happiness and sadness influence outcome perceptions (see Bower, 1981). Hence, the boy in the
story might perceive an ambiguous outcome
during the game as a failure because of his
unhappiness and other negative affective experiences initiated prior to the outcome in
question. These added intricacies are neglected
here, but deserve full incorporation into the
theory.
Achievement Change Programs
I now turn from the devised scenario to an
ongoing topic of research. There is an increas-

ACHIEVEMENT MOTIVATION AND EMOTION

ingly popular therapeutic treatment that induces participants to alter their attributions
for success and failure (see Foersterling, in
press). Often the participants in these programs were selected because they ascribe personal failure to low ability. The main empirical
finding in these studies is that persistence in
the face of failure is enhanced when attributions for failure are changed from low ability
to lack of effort (Andrews & Debus, 1978;
Chapin & Dyck, 1976; Dweck, 1975; Zoeller,
Mahoney, & Weiner, 1983), to poor strategy
(Anderson, 1983b; Anderson & Jennings,
1980), or to temporary external barriers (Wilson & Linville, 1982, 1985).
To alter attributions, in the treatment techniques the experimenter often directly communicates to the participants the attribution
that is desired to be induced (e.g., "You failed
because you did not try hard enough" or
"Success depends entirely on finding the right
strategy"). Following the logic of Figure 2, the
participants use this information to reach a
causal conclusion. Furthermore, they apparently accept the communicated ascription,
rather than attributing induced failure to low
ability. Attributions thus are altered from stable to unstable, which should (and does) result
in the maintenance of goal expectancy (see
Anderson, 1983b; Anderson & Jennings,
1980). This sequence is depicted in Linkages
3, 5, and 6 in the figure. The investigators all
reported increments in persistence of achievement strivings in the face of failure following
the treatment and assumed that the behavior
change was mediated, in part, by shifts in the
subjective expectancy of success, as indicated
in Linkage 11 (although Wilson & Linville,
1982, questioned the consciousness of this estimate, and Dweck, 1975, stressed the controllability rather than the stability of the induced causal ascription).
It is therefore evident that the researchers
have focused attention on expectancy of success. Although consistent with the theory, it
also is clear that the conceptualization is not
fully brought to bear on the phenomena, inasmuch as emotions are entirely neglected. As
previously revealed, ascriptions to ability, effort, strategy, and external barriers have disparate affective consequences. For example, a
program that induces effort rather than ability
ascriptions for failure theoretically is altering

567

reactions of shame and humiliation to guilt.


This new emotional reaction, rather than (in
addition to) a change in expectancy, may be
responsible for the increments in motivated
behavior. On the other hand, a program that
promotes task difficulty ascriptions (Wilson &
Linville, 1982, 1985) theoretically is enhancing
the self-esteem of the participants, for the attribution is being shifted from internal to external (in contrast to the controllability alteration that precedes the hypothesized affective
shift from shame to guilt). Perhaps increments
in self-esteem rather than (in addition to) expectancy maintenance is responsible for the
augmented achievement strivings. In sum, the
attributional conception in Figure 2 suggests
that the change programs may be more complex, and less similar to one another, than has
been recognized.
Theoretical Generality
It has been intimated throughout this article
that the theory shown in Figure 2 is conceived
as a general conceptual framework, although
it has been acknowledged that the vast amount
of supporting data has been generated in
achievement-related contexts. Thus, although
the focus of the theory concerns achievement
strivings, it is tentatively believed that the conception has a wide range of applicability. This
is similar to the position espoused by Atkinson
(1964), who also assumed that he was developing a general theory of motivation, although
achievement settings provided the site of the
experimental research.
The foundation for generality in the present
approach is provided by two conceptual
mechanisms. First, it is proposed that a motivational episode is initiated following any
outcome that can be construed as attainment
or nonattainment of a goal. Achievement success and failure clearly capture this requirement, but acceptance and rejection in the affiliative domain provide a ready parallel (see
Anderson, 1983a; Sobol & Earn, in press). In
addition, the conception has been used to examine a number of social and personal "failures," including, for example, alcoholism
(McHugh, Beckman, & Frieze, 1979); crime,
and parole decisions (Carroll, 1978); depression (Abramson et al., 1978); deprivation
(Mark, 1985); loneliness (Peplau, Russell, &

568

BERNARD WEINER

Helm, 1979); need for help (Betancourt, 1983;


Reisenzein, in press; Weiner, 1980a, 1980b);
maladaptive reactions to rape (Janoff-Bulman,
1979); smoking (Eiser, Van der Pligt, Raw, &
Sutton, in press); and wife battering (Freize,
1979). It is especially worth noting that the
popular attributional analysis of depression
advanced by Abramson et al. (1978) and the
oft-cited distinction between characterological
versus behavioral self-blame (Janoff-Bulman,
1979) both have the present attributional approach as their source.
In these analyses, the authors first determine
the perceived cause of the outcome, such as
the cause of a crime or the cause of depression.
Although these causes vary widely, both within
and between the domains under consideration,
they can be described according to their structural properties of locus, stability, and controllability. The dimensional analysis furnishes
the second key mechanism for theoretical generality, for once the structure of the cause is
ascertained, then its impact on expectancy, affect, and action can be tested. Consider, for
illustrative purposes, the application of the
theory to the disparate areas of criminal behavior (parole decisions), smoking cessation,
and help giving. The latter two research topics
have been subject to examination by means of
path-analytic techniques, thus providing a full
or partial test of the proposed temporal sequence outlined in Figure 2,
Parole decisions. According to Carroll
(1978) and Carroll & Payne (1976, 1977), parole decision makers search for a cause of a
crime when reaching their decision, utilizing
and integrating a variety of available information such as the past criminal record, circumstances at the time of the crime, and so
on. According to Figure 2, the perceived stability of the cause determines the risk of the
criminal to society, that is, the expectancy that
another crime might be committed. The controllability and/or intentionality of the crime
influence(s) the anger at the criminal. Risk and
anger (expectancy and affect), in turn, are hypothesized to affect the parole judgment.
Carroll (1978) and Carroll and Payne (1976,
1977) furnished evidence that this analysis does
capture the parole decision process. It was
found, for example, that an individual with a
record of conviction who committed a longplanned crime is less likely to be paroled than

is one without a previous history who impulsively committed the same crime. This is in
part because criminal history is a cue used to
determine the stability of the cause of the
crime; an extensive history results in the perception of the cause of the current crime as
stable and recidivism is therefore anticipated.
Parole board members do consider nonattributional factors in their decisions and, as opposed to college students simulating parole officials, base their decisions entirely on risk factors. However, the decisions made by judges
and by college students also take into account
beliefs regarding "deserved" punishment.
Crimes committed because of intentional and/
or controllable factors are believed to be more
deserving of punishment than are crimes
due to unintentional and/or noncontrollable
causes.
In this research, as in the work on achievement change programs, the full theory has not
been applied because of the neglect of emotions. This is not the fault of the researchers,
for the introduction of emotion into this theory
came well after their studies. Hence, a reasonable direction for these investigations is to incorporate feelings, particularly anger and pity,
into the determinants of sentence and parole
decisions.
Cessation of smoking. A great deal of attention has been devoted by psychologists to
an examination of why people do not give up
smoking, given the known negative consequences of this behavior. Eiser and Sutton
(1977) argued that the decision facing a wouldbe quitter is not whether to smoke or to quit,
but whether to smoke or to try to quit. This
shifts the theoretical focus from the determinants of quitting to the subjective expectancy
that an attempt at quitting will be successful.
In a large survey study, Eiser et al. (in press)
examined the attributions that smokers give
for the failure of others to give up smoking, as
well as the reasons for their own personal failure at cessation attempts. Path analyses revealed that the perceived stability of the cause
of prior failure attempts of both others and
oneself was related to personal confidence
about giving up smoking in the future. Confidence, in turn, was related to the behavioral
intention to try and quit, and intention was
associated with actual abstinence attempts.
Hence, the temporal sequence of cause, causal

ACHIEVEMENT MOTIVATION AND EMOTION

stability, expectancy of success, and behavior


was confirmed in this health-related context.
Again, the role of emotions was neglected, and
the possibility arises that affects such as shame
or guilt might also be predictive of attempts
at quitting (also see Goldstein, Gordon, &
Marlatt, 1984).
Helping behavior. According to the present
attributional analysis, when a person is in need
of aid, the potential helper attempts to determine why help is needed. If the cause is uncontrollable, then pity is experienced and help
should be offered. On the other hand, if the
cause is perceived as controllable, then the
person is held responsible, anger is experienced, and help should be withheld. Extensive
research has revealed positive associations between perceived controllability-anger-neglect
and between perceived uncontrollability-pityhelp (see J. Meyer & Mulherin, 1980). For example, it has been documented that individuals on a subway are more likely to help a falling person who is ill (uncontrollable) rather
than drunk (controllable; Piliavin, Rodin, &
Piliavin, 1969; Reisenzein, in press; Weiner,
1980a); that students are more likely to lend
another student their class notes if the other
student has an eye problem (uncontrollable)
rather than if the student needs notes because
he or she went to the beach (controllable; Betancourt, 1983; Reisenzein, in press; Weiner,
1980b); and that teachers are more likely to
help a shy (uncontrollable) rather than a hyperactive (controllable) student (Brophy &
Rohrkemper, 1981). Note that the attributional approach points out the similarity, and
like consequences, between drunkenness and
going to the beach, or between illness and shyness, by indicating their comparable subjective
placements on the causal dimension of controllability.
This research, in direct opposition to the
study of achievement change programs, parole
decisions, and attempts to quit smoking, has
typically ignored the role of causal stability and
expectancy in the motivational sequence and
has focused on affect. Thus, the entire theory
depicted in Figure 2 again has not been engaged. Aid may be more likely to be extended
if the cause of the need is perceived as stable
as well as uncontrollable, so that there is relative certainty that the needy person will be
unable to help him- or herself in the future.

569

Thus, students may be more likely to lend their


notes to a blind person rather than to a person
with a temporary eye problem. In addition,
stable uncontrollable causes of need elicit
greater pity than do unstable uncontrollable
causes.
Concluding Comments
During the decades between 1930-1950 the
field of motivation was central in psychology.
At present, this field is not particularly active.
I suggest that one reason for the relative demise
in the perceived importance of motivational
thinking has been the unreliability of the "reference experiments," that is, the basic investigations that provide the empirical foundations for the theories. For example, regarding
the unequal recall of incompleted versus completed tasks, or what is known as the Zeigamik
effect, Lewin (1935) stated, "All later experimental investigations were built upon this"
(p. 240). But the differential task recall observed by Lewin and Zeigamik is not a reliable
finding. In a similar manner, Atkinson (1964)
contended that individuals classified as high
versus low in achievement needs exhibit opposing risk preferences, given tasks differing
in perceived difficulty. This central prediction
from Atkinson's conception is not reliably
found (see W. Meyer, Folkes, & Weiner, 1976);
one suspects this is partially responsible for
the lessening influence of this conception. And
differences in expectancy shifts between people
labeled as internal and external in perceptions
of control has not been reliably demonstrated,
although this is a fundamental prediction of
Rotter's (1966) conception.
The empirical foundation for the theory
presented herethe existence of causal search,
the dominant causal perceptions, the structure
of perceived causality, the relation between
causal stability and expectancy change, and the
associations between causal structure and the
emotions of pride, anger, pity, guilt, gratitude,
shame, and hopelessness, is robust. I believe
that these facts and relations will survive, independent of the fate of the entire theory. In
addition, the present conception has other virtues perhaps less evident in prior motivational
conceptions: A full range of cognitions and
emotions are incorporated and there is an explicit concern with the self. Furthermore, an
attempt has been made to relate the structure

570

BERNARD WEINER

of thought (in this case, causal thinking) to the


dynamics of feeling and action. This is one of
the basic tasks that motivational theorists must
solve.
In conclusion, I believe that some attention
also must be paid to the "nothing but common
sense" criticism leveled at times against attributional approaches. When critics charge that
an attributional approach is "mere" common
sense, they are exclaiming that the relations
pointed out or predicted by the theory represent shared knowledge (see Fletcher, 1984). I
agree that the linkages in Figure 2 between,
for example, stable causes and repeated effects,
internal locus and self-esteem, and causal controllability and anger, gratitute, and guilt, generally are known or at least will be positively
acknowledged when presented to the lay public. What is not shared knowledge, however, is
the conceptual analysisthe linking of various
"understood" empirical relations and the use
of similar principles to explain a vast array of
phenotypic observations. The layperson does
not appreciate that expecting to be rejected
for a social engagement because of prior attribution to lack of attractiveness and feeling
grateful and returning a favor because of a volitionally given gift are part of the same conceptual network. It is this systemization, that
is, the higher order relations between associations realized in everyday life, that represents
much of this attributional contribution. That
the individual parts or components are naively
shared underscores their veridicality, thus further supporting the certainty of the empirical
relations and thereby providing a strong foundation for theory building.
References
Abelson, R. P. (1983). Whatever became of consistency
theory? Personality and Social Psychology Bulletin, 9,
37-54.
Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. (1978).
Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74.
Anderson, C. A. (1983a). The causal structure of situations:
The generation of plausible causal attributions as a
function of type of event situation. Journal of Experimental Social Psychology, 19, 185-203.
Anderson, C. A. (1983b). Motivational and performance
deficits in interpersonal settings: The effects of attributional style. Journal of Personality and Social Psychology,
45, 1136-1147.
Anderson, C. A., & Jennings, D. L. (1980). When experiences of failure promote expectations of success: The

impact of attributing failure to ineffective strategies.


Journal of Personality, 48, 393-407.
Andrews, G. R.( & Debus, R. L. (1978). Persistence and
causal perceptions of failure: Modifying cognitive attributions. Journal of Educational Psychology, 70, 154166.
Arnold, M. B. (1960). Emotion and personality. New York:
Columbia University Press.
Atkinson, J. W. (1964). An introduction to motivation.
Princeton, NJ: Van Nostrand.
Averill, J. A. (1982). Anger and aggression. New York:
Springer Verlag.
Averill, J. A. (1983). Studies on anger and aggression.
American Psychologist, 38. 1145-1160.
Bar-tal, D., Goldberg, M., & Knaani, A. (1984). Causes of
success and failure and their dimensions as a function
of SES and gender: A phenomenological analysis. British
Journal of Educational Psychology, 54, 51-61.
Betancourt, H. (1983). Causal attributions, empathy, and
emotions as determinants of helping behavior: An integrative approach. Unpublished doctoral dissertation,
University of California, Los Angeles.
Bottenberg, E. H. (1975). Phenomenological and operational characteristics of factor-analytically derived dimensions of emotion. Psychological Reports, 37, 12531254.
Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129-148.
Brophy, J. E., & Rohrkemper, M. M. (1981). The influence
of problem ownership on teachers' perceptions of and
strategies for coping with problem students. Journal of
Educational Psychology, 73, 295-311.
Brown, J., & Weiner, B. (1984). Affective consequences of
ability versus effort ascriptions: Controversies, resolutions, and quandaries. Journal of Educational Psychology, 76. 146-158.
Bryant, F. B., & Veroff, J. (1982). The structure of psychological well-being: A sociohistorical analysis. Journal
of Personality and Social Psychology, 43, 653-673.
Burger, J. M., Cooper, H. M., & Good, T. L. (1982). Teacher
attributions of student performance: Effects of outcome.
Personality and Social Psychology Bulletin, 4, 685-690.
Carroll, J. S. (1978). Causal attributions in expert parole
decisions. Journal of Personality and Social Psychology,
36, 1501-1511.
Carroll, J. S., & Payne, J. W. (1976). The psychology of
the parole decision process: A joint application of attribution theory and information processing psychology.
In J. S. Carroll & J. W. Payne (Eds.), Cognition and
social behavior (pp. 13-32). Hillsdale, NJ: Erlbaum.
Carroll, J. S., & Payne, J. W. (1977). Judgements about
crime and the criminal: A model and a method for investigating parole decision. In B. D. Sales (Ed.), Prospectives in law and psychology: Vol. 1. The criminal
justice system (pp. 191-240). New York: Plenum Press.
Chapin, M., & Dyck, D. G. (1976). Persistence in children's
reading behavior as a function of N length and attribution
retraining. Journal of Abnormal Psychology, 85, 511515.
Cohen, J., & Hansel, C. E. M. (1956). Risk and gambling.
London: Longmans Green.
Collins, B. E., Martin, J. C., Ashmore, R. D., & Ross, L.
(1974). Some dimensions of the internal-external metaphor in theories of personality. Journal of Personality
and Social Psychology, 29, 381-391.

ACHIEVEMENT MOTIVATION AND EMOTION


Cooper, H. M., & Burger, J> M. (1980). How teachers explain students' academic performance: A categorization
of free response academic attributions. American Educational Research Journal, 17, 95-109.
Covington, M. V., & Omelich, C. L. (1984). An empirical
examination of Werner's critique of attribution research.
Journal of Educational Psychology, 76, 1199-1213.
Crittended, K. S., & Wiley, M. G. (1980). Causal attributions and behavioral response to failure. Social Psychology Quarterly, 43, 353-358.
Dalai, A., Weiner, B., & Brown, J. (1985). Issues in the
measurement of causal stability. Unpublished manuscript, University of California, Los Angeles.
Davitz, J. R. (1969). The language of emotion. New York:
Academic Press.
Day, V. H. (1982). Validity of an attributional model for
a specific life event. Psychological Reports, 50, 434.
Diggory, J. C., Riley, E. J., & Blumenfeld, R. (1960). Estimated probability of success for a fixed goal. American
Journal of Psychology, 73, 41-55.
Dweck, C. S. (1975). The role of expectations and attributions in the alleviation of learned helplessness. Journal
of Personality and Social Psychology, 31, 674-685.
Eiser, J. R., & Sutton, S. R. (1977). Smoking as a subjectively rational choice. Addictive Behaviors, 2, 129-134.
Eiser, J. R., Van der Pligt, J., Raw, M., & Sutton, S. R. (in
press). Trying to stop smoking: Effects of perceived addiction, attributions for failure and expectancy of success. Journal of Behavioral Medicine.
Elig, T. W, & Frieze, I. H. (1979). Measuring causal attributions for success and failure. Journal of Personality
and Social Psychology, 37, 621-634.
Falbo, T, & Beck, R. C. (1979). Naive psychology and the
attributional model of achievement. Journal of Personality, 47, 185-195.
Feather, N. T, & Davenport, P. (1981). Unemployment
and depressive affect: A motivational and attributional
analysis. Journal of Personality and Social Psychology,
41, 422-436.
Feather, N. T, & Simon, J. G. (1972). Causal attributions
for success and failure in relation to initial confidence
and success and failure of self and other. Journal of Personality and Social Psychology, 18, 173-188.
Fincham, F. D., & Jaspers, J. M. (1980). Attribution of
responsibility: From man the scientist to man as lawyer.
In L. Berkowitz (Ed.), Advances in experimental social
psychology (Vol. 13, pp. 82-139). New York: Academic
Press.
Fletcher, G. O. (1984). Psychology and common sense.
American Psychologist, 39, 203-213.
Foersterling, F. (1980). A multivariate analysis of perceived
causes for success and failure. Archives of Psychology,
133, 45-52.
Foersterling, F. (in press). Attributional retraining: A review.
Psychological Bulletin.
Folkes, V. S. (1982). Communicating the reasons for social
rejection. Journal of Experimental Social Psychology,
18, 235-252.
Folkes, V. S. (1984). Consumer reactions to product failure:
An attributional approach. Journal of Consumer Research, 11, 398-409.
Fontaine, C. (1974). Social comparison and some determinants of expected personal control and expected performance in a novel task situation. Journal of Personality
and Social Psychology, 29, 487-496.

571

Forsyth, D. T. (1980). The function of attributions. Social


Psychology Quarterly, 43, 184-189.
Frieze, I. H. (1976). Causal attributions and information
seeking to explain success and failure. Journal of Research in Personality. 10, 293-305.
Frieze, I. H. (1979). Perceptions of battered wives. In Frieze,
I. H., Bar-Tal, D., & Carroll, J. S. (Eds.), New approaches
to social problems (pp. 79-108). San Francisco: JosseyBass.
Frieze, I. H., & Snyder, H. N. (1980). Children's beliefs
about the causes of success and failure in school settings.
Journal of Educational Psychology, 72, 186-196.
Furnham, A. (1982a). Explanations for unemployment in
Britain. European Journal of Social Psychology, 12, 335352.
Furnham, A. (1982b). Why are the poor always with us?
Explanations for poverty in Britain. British Journal of
Social Psychology, 21, 311-322.
Goldstein, S., Gordon, J. R., & Marlatt, C. A. (1984, August). Attributional processes and relapse following
smoking cessation. Paper presented at the 92nd Annual
Convention of the American Psychological Association,
Toronto, Ontario, Canada.
Graham, S. (1984). Communicating sympathy and anger
to black and white children: The cognitive (attributional)
consequences of affective cues. Journal of Personality
and Social Psychology, 47, 40-54.
Graham, S., Doubleday, C., & Guarino, P. A. (1984). The
development of relations between perceived controllability and the emotions of pity, anger and guilt. Child
Development, 55, 561-565.
Goranson, R. E., & Berkowitz, L. (1966). Reciprocity and
responsibility reactions to prior help. Journal of Personality and Social Psychology, 3, 227-232.
Greenberg, M. S., & Frisch, D. M. (1972). Effect of intentionality on willingness to reciprocate a favor. Journal
of Experimental Social Psychology, 8, 99-111.
Harvey, J. H., & Weary, G. (1981). Perspectives on attributional processes. Dubuque, IA: Wm. C. Brown.
Heider, F. (1958). The psychology of interpersonal relations.
New York: Wiley.
Heilman, M. E., & Guzzo, R. A. (1978). The perceived
causes of work success as a mediator of sex discrimination in organizations. Organizational Behavior and
Human Performance, 21, 346-357.
Hoffman, M. L. (1970). Conscience, personality, and socialization techniques. Human Development, 13,90-126.
Hoffman, M. L. (1976). Empathy, role-taking, guilt, and
development of altruistic motives. In T. Likona (Ed.),
Morality: Theory, research and social issues (pp. 124143). New York: Holt, Rinehart & Winston.
Hoffman, M. L. (1982). Development of prosocial motivation: Empathy and guilt. In N. Eisenberg-Borg (Ed.),
Development of prosocial behavior (pp. 281-313). New
York: Academic Press.
Inagi, T. (1977). Causal ascription and expectancy of success. Japanese Psychological Research, 19, 23-30.
Izard, C. E. (1977). Human emotions. New York: Plenum.
Jagacinski, C. M., & Nicholls, J. G. (1984). Conception of
ability and related affects in task involvement and ego
involvement. Journal of Educational Psychology, 76,
909-919.
Janoff-Bulman, R. (1979). Characterological versus behavioral self-blame: Inquiries into depression and rape.

572

BERNARD WEINER

Journal of Personality and Social Psychology, 37, 17981809.


Kelley, H. H. (1971). Attributions in social interactions.
Morristown, NJ: General Learning Press.
Kelley, H. H. (1983). The situational origins of human
tendencies. Personality and Social Psychology Bulletin,
9, 8-30.
Kelley, H. H., & Michela, J. L. (1980). Attribution theory
and research. In M. R. Rosenzweig & L. W. Porter (Eds.),
Annual review of psychology (Vol. 31, pp. 457-501). Palo
Alto, CA: Annual Reviews.
Kovenklioglu, G., & Greenhaus, J. H. (1978). Causal attributions, expectations and task performance. Journal
of Applied Psychology, 63, 698-705.
Lazarus, R. S. (1966). Psychological stress and the coping
process. New York: McGraw-Hill.
Lee, Y. K. (1976, April). Construct validation of a new
locus of control scale by multidimensional unfolding.
Paper presented at the meeting of the Western Psychological Association, Los Angeles.
Lewin, K. (1935). A dynamic theory of personality. New
York: McGraw-Hill.
Lewin, K., Dembo, T, Festinger, L., & Sears, P. S. (1944).
Level of aspiration. In J. McV. Hunt (Ed.), Personality
and the behavioral disorders (Vol. 1, pp. 333-378). New
York: Ronald Press.
Mark, M. M. (1985). Expectation, procedural justice, and
alternative reactions to being deprived of a desired outcome. Journal of Experimental Social Psychology, 21,
114-137.
Martinez, A. (1982, November 1). Out-of-work lumbermen
not out of the woods yet. Los Angeles Times, Pt. 5,
p. 1.
McFarland, C, & Ross, M. (1982). Impact of causal attributions on affective reactions to success and failure.
Journal of Personality and Social Psychology, 43, 937946.
McHugh, M., Beckman, L., & Frieze, I. H. (1979). Analyzing alcoholism. In I. H. Frieze, D. Bar-Tal, & J. S.
Carroll (Eds.), New approaches to social problems
(pp. 168-208). San Francisco: Jossey-Bass.
McMahan, I. D. (1973). Relationships between causal attributions and expectancy of success. Journal of Personality and Social Psychology, 28, 108-115.
Meyer, J. P. (1980). Causal attributions for success and
failure: A multivariate investigation of dimensionality,
formation, and consequences. Journal of Personality arid
Social Psychology, 38, 704-715.
Meyer, J. P., & Koelbl, S. L. M. (1982). Dimensionality of
students' causal attributions for test performance. Personality and Social Psychology Bulletin, 8, 31-36.
Meyer, J. P., & Mulherin, A. (1980). From attribution to
helping: An analysis of the mediating effects of affect
and expectancy. Journal of Personality and Social Psychology, 39, 201-210.
Meyer, W. U. (1973). Leistungsmotiv und-Ursachenerklarung von Erfolg und Misserfolg [Achievement motivation and causal attributions for success and failure].
Stuttgart: Ernst Klett.
Meyer, W. U., Folkes, V. S., & Weiner, B. (1976). The perceived informational value and affective consequences
of choice behavior and intermediate difficulty task selection. Journal of Research in Personality, 10, 410-423.
Michela, J. L., Peplau, L. A., & Weeks, D. G. (1982). Per-

ceived dimensions of attributions for loneliness. Journal


of Personality and Social Psychology, 43, 929-936.
Montanelli, D. S., & Hill, K. T. (1969). Children's achievement expectations and performance as a function of
two consecutive reinforcement experiences, sex of subject, and sex of experimenter. Journal of Personality and
Social Psychology, 13, 115-128.
Musashi, M. (1974). A book of five rings. Woodstock, NY:
Overlook. (Original work written in 1645)
Neale, J. M., & Friend, R. M. (1972). Attributional determinants of reactions to performance in academic situations. Perceptual and Motor Skills, 34, 35-40.
Pancer, S. M. (1978). Causal attributions and anticipated
future performance. Personality and Social Psychology
Bulletin, 4, 600-603.
Pancer, S. M., & Eiser, J. R. (1977). Expectation, aspirations
and evaluations as influenced by another's attribution
for success and failure. Canadian Journal of Behavioral
Science, 9, 252-264.
Passer, M. W. (1977). Perceiving the causes of success and
failure revisited: A multidimensional scaling approach.
Unpublished doctoral dissertation, University of California, Los Angeles.
Passer, M. W, Kelley, H. H., & Michela, J. L. (1978). Multidimensional scaling of the causes for negative interpersonal behavior. Journal of Personality and Social
Psychology, 36, 951-962.
Pastore, N. (1952). The role of arbitrariness in the frustration-aggression hypothesis. Journal of Abnormal and
Social Psychology, 47, 728-732.
Peplau, L. A., Russell, D., & Heim, M. (1979). The experience of loneliness. In I. H. Frieze, D. Bar-Tal, &
J. S. Carroll (Eds.) New approaches to social problems
(pp. 53-78). San Fransisco: Jossey-Bass.
Piliavin, I. M., Rodin, J., & Piliavin, J. A. (1969). Good
samaritanism: An underground phenomenon? Journal
of Personality and Social Psychology, 13, 289-299.
Reisenzein, R. (in press). A structural equation analysis of
Weiner's attribution-affect model of helping behavior.
Journal of Personality and Social Psychology,
Robert, R. (1982, November 24). Malavasi questions character of some, says coaching is tough. Los Angeles Times,
Pt. 3, p. 3.
Ronis, D. L., Hansen, R. D., & O'Leary, V. B. (1983).
Understanding the meaning of achievement attributions:
A test of derived locus and stability scores. Journal of
Personality and Social Psychology, 44, 702-711.
Rosenbaum, R. M. (1972). A dimensional analysis of the
perceived causes of success and failure. Unpublished
doctoral dissertation, University of California, Los Angeles.
Roseman, I. J. (1984). Cognitive determinants of emotion.
Review of Personality and Social Psychology, 5, 11-36.
Rotter, J. B. (1966). Generalized expectancies for internal
versus external control of reinforcement. Psychological
Monograph, 80, 1-28.
Saulnier, K., & Pearlman, D. (1981). The actor-observer
bias is alive and well in prison: A sequel to Wells. Personality and Social Psychology Bulletin, 7, 559-564.
Schachter, S., & Singer, J. E. (1962). Cognitive, social and
physiological determinants of emotional state. Psychological Review, 69, 379-399.
Schiitz, A. (1967). Collected papers I: The problem of social
reality. The Hague, Netherlands: Martinus Nijhoff.
Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive

ACHIEVEMENT MOTIVATION AND

appraisal in emotion. Journal of Personality and Social


Psychology, 48, 813-838.
Smith, E. R., & Kleugel, J. R. (1982). Cognitive and social
bases of emotional experience: Outcome, attribution,
and affect. Journal of Personality and Social Psychology,
43, 1129-1141.
Sobol, M. P., & Earn, B. M. (in press). Assessment of children's attributions for social experiences: Implications
for social skills training. In B. H. Schneider, J. E. Ledingham, & K. H. Rubin (Eds.), Research strategies in
children's social skills training. New York: Springer-Verlag.
Staff. (1926, April). Fifty and 100 years ago. Scientific
American, p. 228.
Stern, P. (1983). A mullimethod analysis of student perceptions of causal dimensions. Unpublished doctoral
dissertation, University of California, Los Angeles.
Stipek, D. J. (1983). A developmental analysis of pride and
shame. Human Development, 26, 42-54.
Tesser, A., Gatewood, R., & Driver, M. (1968). Some determinants of gratitude. Journal of Personality and Social
Psychology, 3, 233-236.
Tolman, E. C. (1925). Purpose and cognition: The determinants of animal learning. Psychological Review, 32,
285-297.
Triandis, H. C. (1972). The analysis of subjective culture.
New York; Wiley.
Trivers, R. L. (1971). The evolution of reciprocal altruism.
Quarterly Review of Biology, 46. 35-57.
Valle, V. A. (1974). Attributions of stability as a mediator
in the changing of expectations. Unpublished doctoral
dissertation, University of Pittsburgh.
Weiner, B. (1979). A theory of motivation for some classroom experiences. Journal of Educational Psychology,
71, 3-25.
Weiner, B. (I980a). A cognitive (attribution )-emotion-action model of motivated behavior: An analysis of judgements of help-giving. Journal of Personality and Social
Psychology, 39, 186-200.
Weiner, B. (1980b). May I borrow your class notes? An
attributional analysis of judgements of help-giving in an
achievement related context. Journal of Educational
Psychology, 72.676-681.
Weiner, B. (1982). The emotional consequences of causal
ascriptions. In M. S. Clark & S. T. Fiske (Eds.), Affect
and cognition: The 17th Annual Carnegie Symposium
on Cognition (pp. 185-200). Hillsdale, NJ: Erlbaum.
Weiner, B. (1983). Some methodological pitfalls in attributional research. Journal of Educational Psychology,
75, 530-543.
Weiner, B. (1985). "Spontaneous" causal search. Psychological Bulletin, 97, 74-84.
Weiner, B, Amirkhan, J., Folkes, V. S., & Wachtel, S.
(1985). An attributional analysis of excuses: Studies of
a naive psychology of emotion. Unpublished paper, University of California, Los Angeles.
Weiner, B., Frieze, I. H., Kukla, A., Reed, L., Rest, S., &
Rosenbaum, R. M. (1971). Perceiving the causes of success and failure. Morristown, NJ: General Learning
Press.
Weiner, B., & Graham, S. (1984). An attributional ap-

EMOTION

573

proach to emotional development. In C. Izard, J. Kagan,


& R. Zajonc (Eds.), Emotion, cognition and behavior
(pp. 167-191). Cambridge, MA: Cambridge University
Press.
Weiner, B., Graham, S., & Chandler, C. (1982). Causal
antecedents of pity, anger and guilt. Personality and Social Psychology Bulletin, 8, 226-232.
Weiner, B., Graham, S., Stern, P., & Lawson, M. E. (1982).
Using affective cues to infer causal thoughts. Developmental Psychology, 18, 278-286.
Weiner, B., & Handel, S. (1985). Anticipated emotional
consequences of causal communications and reported
communication strategy. Developmental Psychology, 21,
102-107.
Weiner, B., & Kukla, A. (1970). An attributional analysis
of achievement motivation. Journal of Personality and
Social Psychology, IS, 1-20.
Weiner, B., Nierenberg, R., & Goldstein, M. (1976). Social
learning (locus of control) versus attributional (causal
stability) interpretations of expectancy of success. Journal of Personality, 44, 52-68.
Weiner, B., Russell, D., & Lerman, D. (1978). Affective
consequences of causal ascriptions. In J. H. Harvey,
W. J. Ickes, & R. F. Kidd, (Eds.), New directions in attribution research (Vol. 2, pp. 59-88). Hillsdale, NJ:
Erlbaum.
Weiner, B., Russell, D., & Lerman, D. (1979). The cognition-emotion process in achievement-related contexts.
Journal of Personality and Social Pscyhology, 37, 12111220.
White, R. W. (1959). Motivation reconsidered: The concept
of competence. Psychological Review, 66, 297-333.
Wicker, F, W., Payne, G. C., & Morgan, R. D. (1983).
Participant descriptions of guilt and shame. Motivation
and Emotion, 7, 25-39.
Willson, V. L., & Palmer, D. J. (1983). Latent partition
analysis of attributions for actual achievement. American
Educational Research Journal, 20, 581-589.
Wilson, T. D., & Linville, P. W. (1982). Improving the
academic performance of college freshmen: Attribution
theory revisited. Journal of Personality and Social Psychology, 42, 367-376.
Wilson, T. D., & Linville, P. W. (1985). Improving the
performance of college freshmen with attributional
techniques. Journal of Personality and Social Psychology,
49, 287-293.
Wimer, S., & Kelley, H. H. (1982). An investigation of the
dimensions of causal attribution. Journal of Personality
and Social Psychology, 43, 1142-1162.
Zajonc, R. B., & Brickman, P. (1969). Expectancy and
feedback as independent factors in task performance.
Journal of Personality and Social Psychology, 11, 148156.
Zoeller, C, Mahoney, G., & Weiner, B. (1983). Effects of
attribution training on the assembly task performance
of mentally retarded adults. American Journal of Mental
Deficiency, 88, 109-112.

Received January 9, 1985


Revision received April 11, 1985

You might also like