Download Attributions - Ashton Southard


Things happen…
› Cars break down, people fail exams, sports teams win and lose,
people fall in love, marriages end in divorce, people lose their
jobs, loved ones die, people fight in the streets, people kill others
in war, ethnic groups try to eliminate other groups


Most people, most of the time, do not accept that the
world we live in is random, unpredictable, and unreliable
For most people, most of the time, things happen for a
reason
› Events are caused by something…

For life to be orderly and predictable, people attribute
causes and explanations to events and try to understand
why people behave the way they do
› The ways in which people do this, the reasons why they attribute,
how they attribute, the conditions under which they do and
don’t attribute, all constitute the subject matter of attribution
theory

Like chapter 4 on attitudes, we will review
the study of attributions from 3 perspectives
› Social cognition, social identity, and social
representations

As with most social psychological topics,
the study of attributions has largely been
dominated by the social cognitive tradition
› Which reflects the mainstream view of how
people attribute causes to everyday events and
behavior
› But, the other approaches have attempted to
provide a more social and contextual account
of attributing causality in everyday life

Attribution theory dominated social psychology during the
1970s and 1980s and in that time a massive body of
research was generated
› During the 1970s over 900 attribution studies had been published
› By 1994 it was estimated that this number had quadrupled
(3,600… wow)
› Today a search for attribution studies turns up 27,700 articles

More recently, research on causal attributions has slowed
down
› Although much of this tradition has come to be subsumed under
the field of “ordinary personology”
 Ordinary personology – the processes by which ordinary people
come to understand others by inferring their temporary states and
feelings and other stable enduring traits and characteristics
 However, this focus on understanding others shares many of the central
concerns of the attribution theory


Despite the enormous attention devoted to the study
of attribution, social psychology has failed to develop
a single, unifying, integrating theory of attribution
Rather, there are several “mini-theories” of
attributional processes
› Historically, 3 of these are considered central
 Heider’s (1958) Naïve Psychology
 Jones and Davis’ (1965) Corresponding Inference Theory
 Kelley’s (1967) Covariation Model

These mini-theories are not seen as competing with
each other or describing the same phenomena
› Rather, these mini-theories are seen as complementing
one another
› They could be integrated into a single over-arching
theory, but this has yet to be done


Fritz Heider was an Austrian Jew who fled to the
United States during WWII
His most important work is his 1958 book The
Psychology of Interpersonal Relations
› Set the stage for most of the work on attributions that
followed
› Described a “common sense psychology or a naïve
psychology of action”

This common sense psychology views people as
naïve scientists (remember from chapter 2?)
› People in an intuitive way, or in a common-sense way,
infer or deduce the causes of events around them
› They naturally view the world as sets of cause and effect
relations, even when there is no causal relationship at all

The arrangement of objects and events into
cause and effect relations constitutes a
causal system in our cognitive structure
› The question of which objects and events will be
seen as cause and which will be seen as effect is
crucial
 It almost defines the attributional process

Heider claimed that we tend to perceive a
cause and its effect as a perceptual unit
› Some objects and events combine more easily
than others to form a causal unit
 Especially when the object or cause is a human
actor and the event or effect is a social behavior

Two prime determinants of “unit perception” are similarity and proximity
›
In our causal systems, two events are more likely to be seen as causally related if
they are proximal rather than distal

›

Greater similarity between two events makes them more likely to be perceived as a
causal unit than dissimilar events
Two more important principles of causal inference
›
›
People tend to attribute behavior to a single cause rather than to multiple causes
Causes of behavior can be thought of as residing either within the actor or outside
the actor somewhere in the situation




Temporal proximity (being close in time) is especially potent in influencing causality
Dispositional causes – causes within the actor

Ex. Personality characteristics, motivation, ability, and effort

Ex. Social context and role obligations
Situational causes – causes outside the actor
According to Heider, the more one of these causes is favored as an
explanation of a particular behavior, the less likely the other will also be
used
Heider also noticed that people tend to emphasize dispositional or
internal causes and tend to overlook situational causes when explaining
behavior
›
Which has become known as the “fundamental attribution error”

Edward Jones and Keith Davis’ theory of correspondent
inferences (1965)
› First systematic theory of Heider’s earlier ideas

The basic premise is that, under certain conditions, people
tend to infer that people’s actions correspond to their
intentions and dispositions
› That is, people like to infer that a person’s behavior matches an
underlying stable quality in the person
› Ex. Correspondent inference would be to attribute someone’s
aggressive behavior to an internal stable trait, such as
“aggressiveness”

Jones and Davis argued that such inferences are
motivated by our need to view people’s behavior as
intentional and predictable, reflecting their underlying
character
› This in turn, enhances our sense of being able to predict and
control other people’s behavior and thus our social interactions
more generally

In everyday life, however, making such
correspondent inferences may not be straightforward
› The information needed to make such inferences may be
ambiguous, requiring us to draw upon cues that are
maximally informative
 i.e. that reduce uncertainty about the causes of those
behaviors

Jones and Davis outline 3 major factors affecting the
process of making correspondent inferences:
› The social desirability of the behavior
› A person’s choice in the behavior
› The motivational variables of hedonic relevance and
personalism

We will discuss each of these factors in more detail

Behaviors judged to be socially desirable are less
informative than behaviors judged to be socially
undesirable
› When a behavior is socially desirable in the context in which it
occurs, it is normal or expected

Observing such behavior is not informative to the
perceiver because there are several alternative, equally
probable, reasons why the behavior occurred
› Maybe because the actor is intrinsically a good person,
chronically prone to commit socially desirable behaviors
(dispositional attribution)
› But, the behavior may also have occurred because it was
expected; it was the right thing to do (situational or external
attribution)

Either explanation is equally likely
› The behavior is uninformative because it does not help the
perceiver decide between the two competing explanations of
the good, desirable, expected, normative behavior

This is not the case for socially undesirable behavior
›

These behaviors are counter-normative; they are not what is expected
For this reason they are more informative than socially desirable
behavior
For socially desirable behaviors, dispositional and situational explanations
are equally likely
› For socially undesirable behaviors, the situational explanation is
eliminated (because the behavior is not expected); therefore, it is less
likely than the dispositional explanation
›
Thus, the perceiver, the intuitive scientist, has data that reduce
uncertainty, which helps in deciding between competing
explanations
 Undesirable behaviors are more informative than desirable
behaviors, and allow the perceiver to make a dispositional
attribution about the actor with more confidence

›
The attribution about the actor’s disposition is likely to be as negative as
the observed behavior

The second important factor in determining
correspondent inferences is whether the actor
freely chose the behavior: the principle of noncommon effects
› This principle applies particularly when an actor has,
or at least is perceived to have, free choice in action
between several behavioral alternatives

Again, this principle works because when
someone is believed to have freely chosen this
behavior it is informative
› It reduces uncertainty by implicitly favoring one
explanation for the behavior over other, competing
explanations
The desirability of outcomes and the principle of non-common
effects are both cognitive factors influencing the attributional
process
 The third factor is motivational
 Includes 2 related constructs: hedonic relevance and
personalism

›
An action is said to be hedonically relevant for a perceiver if the
consequences of the action affect the perceiver
 The welfare of the perceiver is either harmed or benefited by the action
›
Personalistic actions are a subset of hedonically relevant actions, and are
characterized by the intention of an actor for the action to have
hedonic relevance for the perceiver
 Does the perceiver think the actor is intentionally trying to affect his/her
welfare

Actions that are perceived to be hedonically relevant or
personalistic are more likely to produce a correspondent
inference (the action matches some underlying trait in the
actor) about the actor than are other actions
The analogy between the professional
scientist and the everyday perceiver, is
emphasized in Kelley’s covariation model of
attribution
 The model rests on the principle of
covariation

› Asserts that before two events can be accepted
as causally linked they must covary with one
another
› If two events do not covary, they cannot be
causally connected
The principle of covariation was used by Kelley as an analogy for
the way people infer causation in their everyday lives
 Kelley suggested there are 3 factors crucial in assessing
covariation

›
And that different combinations of positions of these 3 factors lead to
different types of causal conclusions regarding the specific behavior in
question
 Consistency, distinctiveness, and consensus

The general context in which these 3 dimensions are applied is
one where a perceiver attributes a cause to a person’s response
to a particular stimulus as a particular time
Consistency – whether that person responds in the same way to the
same stimulus or similar stimuli at different times
› Distinctiveness – whether the actor acts in the same way to other,
different stimuli, or whether the actor’s response distinguishes between
different stimuli
› Consensus – is not a feature of the actor’s behavior, but of the behavior
of others: is there consensus across actors in response to the same
stimulus, or do people vary in response?
›

According to the covariation model, perceivers will
decide, almost in a dichotomous way:
› That the actor acts either in the same way at different times
(consistency is high) or in different ways (consistency is low)
› That the actor either shows similar responses to different stimuli
(distinctiveness is low) or acts this way only in response to this
particular stimulus (distinctiveness is high)
› That the actor either acts in the same way as most other people
(consensus is high) or acts differently (consensus is low)

Different combinations of positions on the 3 dimensions
lead to different attributions
› An internal or dispositional attribution is most likely when
consistency is high, distinctiveness is low, and consensus is low
› An external or situational attribution is most likely when
consistency is high, distinctiveness is high, and consensus is low
› Other combinations lead to less clear attributions
2 important factors were added to the covariation model by
Kelley five years after his original formulation: discounting and
augmentation
 An event can have many causes

›

It sometimes happens that several plausible causes co-occur
 But some would be expected to augment the given effect, or make it
more likely
 And some would be expected to inhibit the given effect, or make it less
likely
If the effect occurs even in the presence of inhibitory causes, the
augmenting cause will be judged as stronger than if the
augmenting cause and its effect had occurred without the
inhibitory cause
›
Ex. If you have the opportunity to help an elderly lady get her shopping
bags in her car but there are several inhibitory factors involved (ex. Its
raining heavily and you are running late and in a hurry) but you have
been raised to help the elderly (augmenting factor) and you end up
helping the elderly lady, the augmenting factor will be seen as stronger
 But if it wasn’t raining and you weren’t in a hurry, this factor may not be
perceived to be as strong



Kelley’s model has one requirement that is not included in either
Heider’s model or Jones and Davis’ model
›
Perceivers use information across times, situations, and actors
›
In contrast, perceiver in the naïve scientist model or in the correspondent
inference model makes causal attributions based on a single action
performed by a single actor on a single instance
Without this information it is not possible to make consistency,
distinctiveness, and consensus judgments
This is an important point when we try to evaluate how well the
theory relates to everyday life
People don’t assign cause to an action as though they are unaware or
ignorant of the likelihood that other people would perform the same
action in response to the same stimulus, or the same person would repeat
the behavior, or how that actor would perform in response to other stimuli
› People do not consider each event as if it were new
› On the other hand, people don’t engage in the complex mental
calculations described by Kelley’s covariation model every time they
assign a cause to an action
›

A resolution to this dilemma is offered by
the concept of causal schemas
› Kelley’s concept can be taken to refer to a set
of stored knowledge about the relations
between causes and effects
› We each acquire an implicit causal theory of
events through our socialization
› This implicit theory gives us ready-made
attributional accounts of most events we
encounter from one day to the next
› It allows us to run on default most of the time,
and we only have to devote attention to
unusual, exceptional or important cases
All three of the theories just discussed view people as
naïve scientists that systematically seek the causes of
events
 A consequence of this is that it views the human perceiver
as rational, as going about the attributional process in a
fairly systematic, logical fashion
 However, we all know that people do not typically act in
this way, not even scientists
 It is reasonable to think of attribution theory as being
prescriptive

› It describes how attribution perhaps should be made
Empirical research has found that people do not make
attributions in such a systematic and calculated way
 Rather, social perceivers demonstrate persistent biases
when attributing causality to events and behavior

When an attribution deviates from the
prescribed model, it is thought of as a
biased attribution
 Some attribution researchers refer to biases
as errors

› This implies that researchers know the true
causes of behavior
› But, in all probability, they don’t

There are no validity benchmarks for
assessing the authenticity of an attribution
› It is better, then, to refer simply to attributional
biases, rather than errors

Original conception
› [The fundamental attribution error] is the tendency for
attributers to underestimate the impact of situational
factors and to overestimate the role of dispositional factors
in controlling behavior


More recently, Gilbert has referred to this
phenomenon as the “correspondence bias”
The earliest empirical demonstration of the
fundamental attribution error (FAE) was produced by
Jones and Harris (1967)
› Participants were shown to make correspondent
inferences about an actor’s attitudes based on the actor’s
statements about an issue
› These inferences occurred even when the participants
knew the actor had no choice in making the statement

1st experiment (conducted in NORTH CAROLINA!!!)
Participants read a short essay on Castro’s Cuba and then indicated
what they thought the essay writers true attitude was
› Each participant only read one essay, but half of the essays were proCastro and the other half were anti-Castro
› Manipulation 1
›
 At the time, there were not many pro-Castro advocates in NC
 Thus, the direction of the essay (pro- or anti-) constituted a manipulation of
the probability of the behavior
 Meaning that the pro-Castro essay was improbable and therefore more
informative behavior
›
Manipulation 2
 Participants were led to believe that the essay’s position had been either
assigned (no choice, uninformative) or chosen by the writer (choice,
informative)
After reading the 200-word essay participants answered questions about
what they thought the essay writer’s TRUE attitude was towards Castro’s
Cuba
› And then indicated their own attitude
›

If the informativeness of the behavior was the most
important factor in determining whether or not a
correspondent inference was made by the observer
› Then such an inference should be most evident among
those who read a pro-Castro essay written by someone
who could have chosen to write an essay criticizing or
defending Castro’s Cuba
 Meaning that this condition should lead participants to
believe the essay truly reflected the attitude of the essay
writer
› And should be least evident among those participants
who read the anti-Castro essay written by someone who
was instructed what to write
 Meaning this condition should lead participants to believe
the essay did not reflect the true attitude of the essay
writer



The mean “attributed attitude scores,” the attitudes participants
attributed to the essay writers
›
Range from 10 (anti-Castro) to 70 (pro-Castro)
›
›
The inferred attitude matched the essay direction
Inferences are stronger in the choice conditions than in the no-choice
conditions
There is indeed evidence that participants made correspondent
inferences
But, importantly for the FAE
›
›

Correspondent inferences are still evident in the no-choice condition
Even when participants were told that the essay writer was instructed to
write either a pro- or and anti-Castro essay, they still infer that the essay
writer has an attitude consistent with the views expressed in the essay
This is the FAE: attributers (participants in the experiment) have
apparently underestimated the impact of situational factors and
over estimated the role of dispositional factors in determining
behavior
Mean attributed attitude scores (and variances in parentheses),
according to essay direction and degree of choice
Essay Direction
Choice Condition
Pro-Castro
Anti-Castro
Choice
59.62 (13.59)
17.38 (8.92)
No choice
44.10 (147.65)
22.87 (17.55)

Further, in a second experiment
› Demonstrated that emphasizing the choice
manipulation did not diminish the attitude
attribution effect
› Even under no-choice conditions
participants were correctly aware of the
essay writer’s choice or lack of choice

Recent research has drawn upon the work
on automaticity in social perception
› Suggesting that making correspondent
inferences is so pervasive that dispositional
attributions are made spontaneously and
without conscious awareness

In the view of social cognition
› Dispositional attributions are the “default option”
› Perceivers spontaneously attributed behavior to
people’s traits because these attributions are
fast and require little effort

Quattrone (1982) was the first to propose a sequential
model of attribution
› Once a behavior is identified, dispositional attributions are
always made first, spontaneously and without conscious
deliberation
› But, these can be subsequently corrected by situational
attributions if perceivers are motivated and have time to
consider alternative explanations

Thus, trait attributions are easy and effortless, and
situational attributions are corrections that require more
effort and cognitive resources
Identification
Attribution
Attribution
Automatic
dispositional
inference
Effortful
situational
correction

A review by Trope and Gaunt (2003)
concludes that situational attributions are
more likely to be made if perceivers:
› 1) are made accountable for their inferences
› 2) are not cognitively busy or distracted by
pursuing other goals; or
› 3) when situational attributions are made salient,
accessible and relevant

But why should dispositional attributions be
so effortless and less cognitively demanding
than situational attributions?

Explanations of FAE tend to belong to one of two types
› Explanations based on psychological or cognitive processes
› Explanations which seek the origins of the bias in social, cultural,
and ideological processes

Heider was the first to suggest a cognitive explanation
› Arguing that “behavior in particular has such salient properties it
tends to engulf the total (perceputal) field”

Fiske and Taylor support this cognitive explanation
› Describing situational factors that give rise to behavior, such as
social context or situational pressures, are relatively less salient
and unlikely to be noticed when compared with the dynamic
behavior of the actor

Thus, the fundamental attribution error has primarily been
explained by the dominance of the actor in the
perceptual field

Another explanation is a motivational one
› Emphasizes the degree to which attributions about a person give
us a sense of predictive control of other people’s behavior

Jones and Davis (1965) stressed this aspect of
correspondent inferences:
› That such inferences are motivated by our need to view
people’s behavior as intentional and reflecting their underlying
personality traits

The main point being:
› If we believed that people’s behavior was unstable and
fluctuated according to the situations people are in, then this
makes predicting their behavior and controlling our environment
more difficult

Thus, dispositional attributions or correspondent inferences
enhance our sense of prediction and control in everyday
life
› This is why we prefer them

Others have suggested that this bias toward dispositional
attributions is not a universal law of cognitive functioning
› But, rather, reflects the dominant ideology of individualism in
European and American culture

The tendency to favor personal over situational causation
was first noted by Ichheiser (1949)
› Instead of viewing this phenomenon as an individual “error” or
bias in cognitive judgment, he viewed it as an explanation
grounded in American society’s collective and cultural
consciousness
The dominant representation of the person in western
liberal democracies is that of an important causative
agent, over and above situational and contextual forces
 The FAE, then, may not be a cognitive or perceptual bias
alone

› Rather, it may largely be a product of western, industrialized
constructions of the “individual” as the source of behavior

If attributions and explanations are
grounded in cultural representations of the
person
› Then cross-cultural differences should be evident
in the prevalence of person attributions

Indeed, this has been largely confirmed by
studies comparing the prevalence of
dispositional attributions in individualist as
compared to collectivist cultures
› The “person” does not seem to enjoy the same
degree of perceptual dominance among nonwestern people living in collectivist societies
“There is a pervasive tendency for actors to attribute their
actions to situational requirements, whereas observers tend to
attribute the same actions to stable personal dispositions.”
(Jones & Nisbett, 1972, p. 80)
 Heider noted that actors and observers have different views of
behavior, of the situation, and of the causes of behaviors in
situations

›

Think of how easy it is for us to explain our own socially
undesirable behavior (e.g., being rude or impolite to a fast food
worker) in terms of extenuating, stressful circumstances (we’re
having a bad day)
›


“The person tends to attribute his own reactions to the object world, and
those of another, when they differ from his own, to personal
characteristics in [the other]” (1958, p. 157)
But we are less sympathetic when others behave in this way (the teller at
the bank was rude to you), and attribute the behavior to the person’s
character (the teller was just a mean person)
Heider referred to this as a “polar tendency in attribution”
Jones and Nisbett called it the “actor-observer effect” (AOE)

Classic AOE experiment Ross, Amabile, and Steinmez (1977)
›
Setup was a quiz game, involving pairs of same-sex participants
›
12 pairs of participants were in the experimental condition
›
6 pairs of participants were in the control condition
›
During the quiz game
 One member of the pair was randomly assigned the role of questioner, the
other the role of contestant (participants were aware the roles were
assigned randomly)
 Questioners were told to make up 10 “challenging but not impossible”
general knowledge questions to ask to the contestant
 While the questioner did this, the contestant was told to compose 10 easy
general knowledge questions, “just to get into the spirit of the study”
 The questioner and participant both made up 10 easy questions
 During the quiz game itself, participants were told the questioner would ask
the contestant 10 questions made up before the study by someone else
 Experimental condition: questioners asked the contestants the 10
“challenging but not impossible” questions they themselves had made up
 Control condition: questioners asked 10 questions given to them by the
experimenter
In the experimental condition the average
number of correct answers in the quiz
game was 4
 After the quiz game

› Questioners and contestants rated themselves
and their partner on a number of dimensions,
most importantly “general knowledge
compared to the average Stanford student”
(experiment was conducted at Stanford)
› Participants rated self and other on a scale from
0 to 100, with 50 being “the average Stanford
student”

Results
›
Control condition: questioners and contestants didn’t really distinguish
between self and other in terms of general knowledge relative to the
average Stanford student
 Everyone rated self and other as about average
›
Experimental condition: there are big differences in how each member of
the pair sees self and other
 The questioner doesn’t really distinguish between self and other
 The contestant devalues self relative to average and increases the rating of
the questioner relative to average
 Presumably because the contestant only got 4 out of 10 answers right on average
and also because they acknowledged the difficulty of the general knowledge
questions produced by the questioner
Mean ratings of self and others’ general knowledge by
questioners and contestants
Ratings of
Self
Other
Questioner
53.5
50.6
Contestant
41.3
66.8
Questioner
54.1
52.5
Contestant
47.0
50.3
Ratings by
Experimental Condition
Control Condition

What makes these results so cool?
› ROLE ASSIGNMENT WAS RANDOM (and the participants knew it
was)
› So presumably, if the roles were reversed, the contestant would
have made up 10 difficult questions and the former questioner
would have got about 4 of them right

There is an asymmetry between the roles in terms of
opportunity to express “smart” behavior
› The questioner gets to call the shots and the contestant has to
play along
› The role of questioner implies an advantage over the contestant

Questioners apparently recognized this
› They elevate neither their own status nor lower the contestants’
status
› But contestants appear to be unaware of, or to under-correct
for, the advantage of the questioner

In a 2nd experiment
› Confederates re-enacted some of the questioner-
contestant performances from study 1
› Real participants watched these interactions, believing
they were authentic
› Participants then rated both the questioner and the
contestant on general knowledge ability, relative to other
Stanford students (on the same rating scale used in
experiment 1)

Results
› Participants, acting as observers, apparently saw the quiz
game through the eyes of the contestant
 The average rating of the questioners was 82.08
 The average rating of the contestants was 48.92
› Thus, these results mirrored the ratings given in experiment 1,
by the contestants themselves


Of course, there are competing explanations
of the AOE
Like the FAE, one explanation is perceptual
› Essentially argues that actors and observers literally
have “different points of view”
 Actors cannot see themselves acting
 From an actor’s perspective what is most salient and available
are the situational influences on their behavior (the objects, the
people, the role requirements, and the social setting)
 From an observer’s point of view, the actor’s behavior is
more perceptually salient than the situation or context
› These different “vantage points” for actors and
observers lead to different attributional tendencies
 Situational attributions for actors
 Dispositional attributions for observers

Taylor and Fiske (1975) attempted to test the perceptual salience
hypothesis
›
›
›
›
2 male confederates were seated opposite one another, and conversed for 5
minutes
A participant sat behind confederate A, so that they could only see confederate B
A participant sat behind confederate B, so that they could only see confederate A
A participant sat at the table between confederates A and B and could see them
both
A
O
C
Blah
O
Blah
B
C
O

After watching A and B interact for 5 minutes all observers rated
each confederate on the dimensions of:
Friendliness
Talkativeness
Nervousness
The extent to which each confederates behavior was caused by
dispositional qualities and by situational factors
› How much each confederate set the conversation’s tone
› Determined the kind of information exchanged in the conversation
› Caused the other’s behavior
›
›
›
›

Results
›
›
›

The observers behind A watching B viewed B as more causal than A
The observers behind B watching A viewed A as more causal than B
Observers between A and B viewed A and B as about equally influential
Thus different vantage points do matter


Other individualistic explanations of AOE have been suggested
Jones and Nisbett (1972)
›

In this view, actors have access to their own feelings, desires and
motivations, as well as to their own cross-situational behavioral
history
›

Originally suggested that actors and observers may possess different
information about events and this is what leads to the different
attributions
Which observers are unaware of
This theory of informational differences between actors and
observers has been supported by research
Idson and Mischel (2001) found that observers were more likely to make
situational inferences and fewer trait attributions about an actor’s
behavior if that person is familiar and important to them
› Presumably, then, the longer we know someone, the more knowledge
we are likely to have about their behavior across different situations
›

Another individualistic (although more social and
interactive) explanation is based on the linguistic
practices of actors and observers
› Different linguistic categories convey different information
about an event

Semin and Fielder (1988) suggest there are 4 linguistic
categories referring to interpersonal relations
› Descriptive action verbs
 Ex. Sally is talking to Bob
› Interpretive action verbs
 Ex. Sally is helping Bob
› State verbs
 Ex. Sally likes Bob
› Adjectives
 Ex. Sally is an extroverted person

Adjectives convey more information about a person than do
verbs
›
And therefore lead to more dispositional inferences
 Ex. It’s hard to imagine making a dispositional correspondent inference
based on the statement “Sally is talking to Bob”
 Ex. But it’s hard not to make a dispositional correspondent inference with
the information “Sally is an extroverted person”
 Because this presumes a disposition in itself

Semin and Fiedler (1989)
›
Found that actors tended to use the more concrete linguistic forms
(descriptive and interpretative verbs)
 i.e. “Sally is talking to Bob,” and “Sally is helping Bob”
›

Observers tended to use the abstract forms (state verbs and adjectives)
 i.e. “Sally likes Bob,” and “Sally is an extroverted person”
In contrast to purely cognitive models of attribution, this work
emphasizes how language itself can shape attributions

Both the actor-observer effect and the fundamental
attribution error are two of the most widely investigated
attributional biases
› Actors and observers differ, sometimes drastically, in the
inferences they draw from and the attributions they make about
presumably the same event

But, the evidence we have discussed here cannot support
a strong form of either the AOE or the FAE
› It appears that attributers do not make either a dispositional or a
situational attribution

Rather, a weak form of the AOE and the FAE is more
consistent with research as a whole
› Attributers use both dispositional and situational factors in
constructing causal sense of events surrounding them
› But, they tend to rely on one relatively more than the other
depending on their perspective of events

While there is evidence that changing
people’s point of view alters their attributional
accounts of events in that view, this does not
imply that there are hard-wired, innate
cognitive attributional mechanisms
› Developmental and cross-cultural research suggests
that people learn the attributional accounts favored
by their social environment
 Ex. European and North American’s tend towards
dispositional attributions and people from more
collectivist societies tend towards situational attributions
› This learning is likely to be so efficient that particular
attributional accounts become automatic and
unconscious


Theories of attribution tend to view the
attributer as a uninvolved bystander observing
events around them
But, as we all know, this is far removed from
the heat of normal human interaction
› People are involved, passionately or not, in the
events around them
› People, and their attributions, affect and are
affected by others and by events

People often make attributions that reflect
self-serving biases
› Designed consciously or unconsciously to enhance
their self-esteem in their own eyes and in the eyes of
others

It is an all too common phenomenon that people
accept credit for success and deny responsibility for
failure
› Students do it after passing for failing a course
› Athletes do it after winning or losing an event
› Even academics do it after having a manuscript
accepted or rejected for publication


Although the strength of the effect varies across
cultures, the attributional asymmetry following
success or failure has been noted in cultures around
the globe
Again, both cognitive and motivational explanations
have been formulated to account for the
attributional difference

Example of cognitive explanation
› Weary (1981) suggested that focus of attention towards
self or away from self and informational availability may
be 2 strong cognitive mechanisms involved in this
phenomenon

However, most researchers advocate a motivational
explanation related to an almost self-evident,
common-sense explanation
› People accept credit for success and deflect responsibility
for failure because doing so makes them feel good and
look good
› Basically, it serves a self-enhancement motive

Ex. Miller (1976)
› Found that the attributional difference is greater when the
task participants succeed or fail at is important to them

Attributing egocentrically not only bolsters self-esteem, but also
influences the impressions others have of the attributer
›

Evidence for the impressions of others is clearer than bolstering one’s own
self-esteem
Ex. Schlenker and Leary (1982)
›
Found that audiences were generally most favorably impressed by
actors who make “accurate” attributional claims for their success
 i.e. actors who were modest about their superior performance were liked
more than actors who boasted about their performance
›

Also, audiences disliked actors who predicted that they would not do
well, even when that prediction turned out to be accurate
What is clear here, is that different attributional patterns following
success or failure create different impressions on an audience
Some kinds of attributions do seem to make the actor look good in the
eyes of others and others seem to make the actor look less favorable
› Whether they also make the actor feel good is another issue
›


Central to any self-enhancement explanation of attributional
biases are these predictions:
›
Self-esteem will increase following a self-serving attribution
›
Self-esteem will decrease following a self-deprecating attribution
 An internal attribution following success or an external attribution following
failure
 An external attribution following success or an internal attribution following
failure
Research does support this, Maracek and Metee (1972):
People with chronically high self-esteem make more self-serving
attributions
› People with chronically low self-esteem make more self-deprecating
attributions
›
This is an important finding with clinical implications for the history
and treatment of depression
 But, this finding is not quite the same thing as evidence that
changes in self-esteem follow particular attributions

›
Which is the core of any self-enhancement explanation

The absence of studies documenting
attribution effects on self-esteem is
curious, and perhaps due to 2 factors:
› First, many researchers appear to accept
such effects as obvious and hence not
needing empirical verification or falsification
› Second, it is methodologically difficult to
design an unconfounded experiment to test
the hypothesis

A pure, experimental investigation would require the experimenter to:
›

The problem is that participants make their own attributions
›


They cannot be assigned to an internal or external attribution condition in the same
way as they can to a success or failure condition
So direction of attribution cannot be experimentally controlled
›
It can only be investigated by allowing participants to make their own attributions
›
Because we know that people with chronic high self-esteem accept credit for
success and deflect blame for failure and people with chronically low self-esteem
tend to do the opposite
And who knows if these attributional styles cause or reflect differences in chronic
self-esteem
But, allowing this automatically introduces a confound between
participants’ attributional direction and their prior self-esteem
›

Assign participants randomly to either an internal or an external attribution
condition following either success or failure and to observe consequent effects on
self-esteem
Thus there is no direct test of the central hypothesis of a selfenhancement explanation
Implicit in the self-enhancement account of attributional biases
is the notion that it is functional, and biologically adaptive to
make these biased attributions because they help to create and
maintain a positive self-esteem
 Weiner’s attributional theory of motivation and emotion

Argues that the kinds of attributions people make for success and failure
elicit different emotional consequences
› And that these attributions are characterized by 3 underlying dimensions
›
 Locus: whether we attribute success and failure internally or externally
 Stability: whether the cause is perceived as something fixed and stable
(like personality or ability) or something changing and unstable (like
motivation and effort)
 Control: whether we feel we have any control over the cause
Locus
Stability
Control
Ability
Internal
Stable
Uncontrollable
Effort
Internal
Unstable
Controllable
Luck
External
Unstable
Uncontrollable
Task difficulty
External
Unstable
Uncontrollable

Consistent with the self-enhancement bias
› People who attribute their achievements to internal, stable and
controllable factors are more likely to feel good about
themselves
 Ex. After making an A on an English essay exam, a student
attributes success to personal ability and effort (I’m good at
writing English essays and I studied hard)
› In contrast, attributing negative outcomes to internal, stable and
uncontrollable factors is associated with negative emotions such
as hopelessness and helplessness
 Ex. After failing an English essay exam, a student attributes failure
to personal lack of ability (I’m never going to be good at writing
English essays because I’m just a dumb student)

This attributional pattern for negative events and
outcomes has been referred to as a “depressive
attributional style”
› This style has been strongly linked with clinical depression


The learned helplessness model of depression
views this attributional style as directly causing
depression
Others have argued this attributional tendency
is merely a symptom of depression
› Reflecting the affective state of the depressed
person

Whether a cause or a symptom
› Attributional retraining programs, in which people
are taught how to make more self-enhancing
attributions, are being widely accepted as an
important clinical intervention in the treatment of
depression

Some psychologists have suggested that attribution
theory exaggerates the extent to which people seek
causal explanations for everyday occurrences and
events:
› That the degree of attributional activity suggested by
attribution research may simply be an artifact of the
reactive methodologies used in such studies


Keep in mind that most attribution studies require,
and instruct, participants to indicate their agreement
or disagreement with attributional statements
provided by the researchers
So, do people spontaneously engage in causal
thinking and, if so, under what conditions do they
make causal attributions?
› 2 studies address these questions directly

Lau and Russell (1980):
› Examined newspaper reports of 33 sporting events
 The 6 baseball games in the 1977 World Series, and a
number of college and professional football games
› Found that more causal attributions were made after
an unexpected outcome than an expected
outcome

Taylor (1982):
› Found that 95% of a sample of cancer victims
spontaneously made attributions about the cause of
their cancer
› Also that 70% of close family members of cancer
victims made spontaneous attributions about the
cause of their loved one’s cancer





These two studies suggest that people do in fact spontaneously
make causal attributions about events around them
›
At least when those events are either unexpected or negative
›
But mostly for unexpected events and especially when confronted with
failure
Weiner (1985) similarly concludes that people do indeed
engage in spontaneous causal thinking
This conclusion is consistent with that of others who have argued
that people actively look for causal explanations for the
unexpected or different
›
And that in such situations the complexity of attributions increases
›
For such events, people probably function mindlessly or essentially run on
automatic
No doubt many events in social life are common, routine,
everyday, and give no need for any sort of attributional analysis
However, people do make causal attributions under some
conditions
›
And even when operating mindlessly people probably could generate
causal attributions for the events passing them by if they were required to
Thus far in this chapter, attributing causes to behavior and
events has been presented primarily as an intraindividual
cognitive phenomenon
 In traditional attribution models individuals are construed as
information processors:



›
›
›
Who attend to and select information from the environment
Process the information
And then arrive at a causal analysis of the behavior or event in question
›
As we all know, social life is an intricate complex mass of individuals,
couples, groups, sects, ethnicities, nations, etc. all interacting and
negotiating an ever-changing social reality which is reproduced,
represented, and reconstructed
This is too simple
None of us interacts with any one other person as if that person
were an abstracted, fixed and given individual
›

We all are social, contextualized, and cannot interact with or even
perceive others as if we, or they, were anything else
According to your textbook:
›
Attribution theory persists in theorizing the asocial, decontextualized
fiction called the “individual”
Hewstone has argued that the bulk of attributional
research has been articulated at the intrapersonal and
the interpersonal levels
 Kelley’s covariation model is a good example of the
intrapersonal level

› An individual perceives an event – usually behavior enacted by
another individual
› Engages in a mental calculus estimating the consistency,
consensus and distinctiveness of that event
› Then arrives at a conclusion regarding the cause of the event

The attributer turns only inward in this attributional search
› Without reference to interpersonal relations, dominant social
representations, the language of causation, and their relative
group memberships and identifications
› Everything, apart from the event that triggered the attributional
search, takes place internally within the mind of the individual

The actor-observer effect and the
fundamental attribution error are
examples of interpersonal attribution
research
› Even here, the individuals in the interaction
come to the interaction strictly as individuals:
 The individuals have no history, no power or
status differentials, no social context
 They are interchangeable, asocial,
decontextualized, often disembodied
individuals
These models thus far have had little to say
about the social, interactive, and cultural
context within which causal attributions are
made
 Attribution theory has therefore been
criticized for being predominantly an
individualistic theory

› Requiring a greater social perspective

Several social psychologists have
attempted to develop a more social
account of attributions
The extraordinary events of September 11, 2001 – the
attacks on the World Trade Center in New York and the
Pentagon in Washington – will undoubtedly be
remembered as one of the most politically salient and
defining events at the beginning of the 21st century
 These attacks and their graphic portrayal on television (as
they occurred in real time) stunned the world

The research on spontaneous attributions
would suggest that such a negative and
unexpected event is likely to generate
considerable attributional activity
 Indeed, as the enormity and significance of
the attacks began to sink in, people tried to
make sense of this event by looking for
reasons as to why it occurred

› Why would a group of individuals plan and
execute such a brazen act of mass murder and
suicide by flying planes into tall buildings?

Here is how Tony Blair, the Prime Minister
of Britain, attempted to make sense of
this event in a speech he delivered the
day following September 11, 2001:
› “The world knows the full evil and capability
of international terrorism which menaces the
whole of the democratic world. The terrorists
responsible have no sense of humanity, of
mercy, or of justice. To commit acts of this
nature requires a fanaticism and wickedness
that is beyond our normal comprehension.”

Over time, of course, we have been provided with a
range of explanatory accounts, by a variety of
expert sources including other world leaders,
politicians, the media, and social analysts
› These explanations included accounts such as that of
Tony Blair that attributed the cause(s) of September 11
primarily to the religious fanaticism and extremism of “evil”
terrorists
› Accounts that attributed the cause(s) to geopolitical
factors and the current state of international relations

Importantly, people’s attributions for this significant
event were not arrived at simply through a solitary
cognitive process of information processing:
› People’s social, cultural and political identifications
significantly shaped their causal analysis and response to
the events of September 11

A social identity, or intergroup, approach
to attributions examines:
› How group memberships, social
identifications, and intergroup relations
affect the sorts of attributions people make

Although we will discuss this work in more
detail later in the chapters on
stereotyping and prejudice
› Here we will cover what Pettigrew (1979) has
coined the “ultimate attribution error”

Pettigrew combined the fundamental attribution error and
Allport’s classical work on intergroup relations:
› Formulated an analysis of how prejudice shapes intergroup
“misattributions”

Pettigrew (1979) noted how people typically make
attributions that:
› Favor and protect the group to which they belong (ingroup)
› Derogate groups to which they do not belong (outgroup)

The title of the universal attribution error was meant partly
as a joke about the rather grand title Ross gave to the FAE
› It’s very unlikely that anything in the social sciences deserves the
label “fundamental”
› Thus, Pettigrew names this ingroup serving and outgroup
derogating attributional pattern the “ultimate attribution error”

When a person is confronted with an obviously positive behavior
committed by a member of a disliked outgroup
That person will have trouble reconciling this with their negative
stereotype of the group
› And is unlikely to make a dispositional attribution
›

This positive outgroup behavior is likely to be explained away or
dismissed as either being:
›
›
Due to external situational pressures OR
Due to the exceptional and thus unrepresentative nature of the outgroup
member
In contrast, negative behavior by an outgroup member will be
attributed to stereotypic dispositions and traits that are
associated with the outgroup
 This pattern of attributions is completely reversed for ingroup
behavior

›
›
Positive ingroup behavior is attributed to dispositional traits
Negative ingroup behavior is attributed to external situational factors
Indeed studies from around the world have
demonstrated what may be a universal,
certainly a pervasive, self-serving and
ethnocentric pattern in the way we see
and explain the events around us
 These studies will be discussed in detail in
chapter 7
 But we will discuss one interesting study
regarding the real world of intergroup
conflict in Northern Ireland


Hunter, Stringer, and Watson (1991): simple study demonstrating
the social psychological processes underlying the markedly
different perceptions of sectarian violence in Northern Ireland
›
Gathered newsreel footage of Catholic and Protestant violence in
Northern Ireland
 One scene showed a Protestant attack on a Catholic funeral
 The other scene showed two soldiers in a car being attacked by a group
of Catholics
Both scenes had been rated as being comparable in their degree of
violent content by a sample of Spanish and German foreign exchange
students
› Then the two clips were shown to Catholic and Protestant students at the
University of Ulster (located in Northern Ireland)
›
 The clips were shown without sound to control for any possible effects of
media bias
Students were then asked to “explain in their own words what they
thought was happening in the videos, and why they thought those
involved had behaved as they had”
› Students’ reasons for the behavior of the people shown in the video were
coded as either an internal or an external attribution
›

Catholic students:
Causes of violent acts committed by Catholics were viewed as being
somewhere in the situation (21 vs. 5)
› Causes of violent acts committed by Protestants were viewed as being caused
by dispositional factors (19 vs. 5)
›

Protestant students:
›
›

Causes of violent acts committed by Protestants viewed as situational (15 vs. 6)
Causes of violent acts committed by Catholics viewed as dispositional (15 vs. 6)
Yet the two groups were watching the same acts of the same people
which had been judged to be equal in violence
Pattern of internal and external attributions
Catholic violence
Protestant violence
Catholic
students
Protestant
students
Catholic
students
Protestant
students
Internal
5
15
19
6
External
21
6
5
15
Attribution

The behavior of participants in the Hunter et al. experiment isn’t
unusual
Studies of this type demonstrate that social perception, especially in
situations involving partisanship, is rarely, if ever, neutral and
dispassionate
› Also, the possibility of ever being able to find a single “true” account of
social “reality” is highly questionable
›

These studies also suggest that differential attributions of this kind
for ingroup and outgroup violence are probably linked to the
maintenance and perpetuation of intergroup conflict
›

External attributions for violence committed by ingroup members by both
groups may serve to justify violence committed by one’s own group and
to view it as legitimate
Further, internal attributions for the other group’s violence may
perpetuate hostilities and perhaps even lead to a self-fulfilling
prophecy
›
Members of the outgroup come to act in ways the ingroup expects



A considerable body of research has indicated that
social groups prefer different explanations for a
range of social issues and problems (ex. Poverty and
unemployment)
Explanations for these social issues are linked to
political identifications and voting behavior
Ex. In Britain
› Conservatives rate individualistic explanations for poverty
and unemployment as more important than Labor voters
› Labor voters rate societal-structural factors as more
important

Explanations are therefore not purely cognitive
phenomena
› But, are collectively shared by those with similar political
and social identities
How do social representations affect
attribution processes and outcomes?
 If we accept that explanations for everyday
events and experiences are social
phenomena, which are negotiated and
communicated during social interaction

› Then, we require an approach that emphasizes
the contents of social knowledge
› Enter social representations theory

Like attribution theory, social representations theory also
emphasizes the fundamental human need to understand
and explain events in everyday life
› Difference:
 Attribution theory seeks to identify the internal cognitive processes
involved in making causal explanations
 Social representations theory seeks to locate these causal
attributions not in individual minds, but in the cultural meaning
systems embodied by social representations

So, both theories emphasize the importance of
explanation in social life
› But, the two theories are articulated at different levels of analysis

Unlike traditional attribution theory, social representations
theory emphasizes the social and collective nature of
explanations

Moscovici and Hewstone have proposed that
social representations should be viewed as the
foundations on which attributions are built
› “A theory of social causality is a theory of our
imputations and attributions, associated with a
representation… any causal explanation must be
viewed within the context of social representations
and is determined thereby.”

Meaning that, when we attribute causality, our
attribution stems from an overall social
representation
› And thus, any attribution we make has to be viewed
in the context of the social representation it came
from

In a similar vein, Lalljee and Abelson (1983) suggest a
“knowledge structure” approach to attribution
› Well-learned consensual structures, like highly organized event
schemas or scripts, do not usually evoke causal explanations
because people expect the sequence of events that occur
› People’s prior expectations, beliefs knowledge or schemas will
determine what parts of social information we need an
attribution for
 Information that is consistent with a person’s schema or
representation won’t require an in-depth search for causality,
because that information is expected and therefore
automatically processed
 In contrast, information that is inconsistent with expectations or
existing knowledge will require a more detailed search to an
explanation

Ex. The Walmart greeter scenario
› The greeter is nice, polite, and helpful = expected and
processed automatically
› The greeter is mean and insulting = need to know why the
greeter acted this way, thus attributional processes are needed
“… social representations impose a kind of automatic
explanation. Causes are singled out and proposed prior to a
detailed search for and analysis of information. Without much
active thinking, people’s explanations are determined by their
social representations”
 The social function of such automatic explanations is that they
are learned and thus socially communicated through language
 It is suggested that the use of cultural hypotheses to explain
behavior and events can be regarded as a kind of “socialized
processing”

Culturally agreed upon explanations eventually become common-sense
explanations
› Each society has its own culturally and socially sanctioned explanation or
range of explanations for phenomena such as illness, poverty, failure,
success, and violence
›

Point being, people don’t always need to engage in an active
cognitive search for explanations for all forms of behavior and
events
›
Instead, people can use their socialized processing for social
representations
The study of perceived causality in attribution theory is
primarily concerned with what passes as everyday social
explanation
 There are 2 kinds of attributions central to attribution
theory: dispositional (personal) and situational (contextual)
 These 2 types of explanation correspond to what has been
referred to as “individual” and “social” principles
 Earlier we discussed one of the most consistent findings in
attribution research: the FAE

› We also discussed that this bias has primarily been explained by
cognitive and perceptual factors (like the dominance of the
actor in the perceptual field)
› But, we also talked about how others have suggested that this
bias may be due to our individualistic culture
The importance of individualism specific to liberal democratic
societies and, most particularly, in American social, cultural, and
political life, has been emphasized by political philosophers
 Ex. Lukes (1973) pointed out how political, economic, religious,
ethical, epistemological, and methodological domains have
been filled with individualist principles

›

This representation of the person may seem pretty self-evident
and not particularly controversial, But, the anthropologist Geertz
emphasizes its uniqueness:
›

Liberal individualism’s central principles emphasize the importance of the
individual over and above society, and view the individual as the center
of all action and processes
“The western conceptions of the person as a bounded, unique, more or
less integrated motivational and cognitive universe, a dynamic center of
awareness, emotion, judgment, and action organized into a distinctive
whole and set contrastively both against a social and natural
background is, however incorrigible it may seem to us, a rather peculiar
idea within the context of the world’s cultures.”
Meaning that the emphasis on individualism in western culture is
odd given that the majority of societies around the globe don’t
think this way

If, indeed, attributions and explanations are grounded in
social knowledge
› Then, cultural variations in the representation of the person
should yield cross-cultural differences in the prevalence of
person attributions

The overcomplicated point the authors of this textbook
are trying to make:
› The focus on individualism in western society, which is part of our
social representations, is not the general view of the rest of the
world
› Emphasis on individualism leads to more personal attributions
and less situational attributions
› So, if attributions do stem from social knowledge, or social
representations, cultures that place emphasis on individualism
should make different attributions from cultures than do not
› Thus, there should be differences in prevalence of personal
attributions between cultures

Supporting research…

Before any research was specifically designed
to examine cultural influences on attributions:
› Developmental research had found a significant
tendency for dispositional attributions to increase
with age in western cultures
 Young western children predominantly make references
to contextual factors to explain social behavior
 Western adults are more likely to stress dispositional
characteristics of the actor
› Anthropologists had also found that that:
 Non-western adults place less emphasis on the
dispositional characteristics of the agent and more
emphasis on the contextual or situational factors,
compared to western adults

At first, social psychologists didn’t attempt to explain these
differences in a social way
› Rather, they tried to explain them in terms of individual cognition
and experience

Ex. Early explanation of why young children don’t make
dispositional attributions
› Young children have a limited cognitive capacity to make
dispositional attributions because it requires the cognitive
competence to generalize behavioral regularities over time
› It was argued that children did not acquire the cognitive
capacity to do this until they were older

Similarly, it was argued that non-western adults are less
likely to make dispositional attributions because the
cognitive capacity to do so is more likely to be associated
with the experience of living in complex modernized
societies

Joan Miller was among the first social psychologists to
point out that these explanations disregard the possibility
that developmental and cultural differences may:
› “result from divergent cultural conceptions of the person
acquired over development in the two cultures rather than from
cognitive or objective experiential differences between
attributors”

Cultural differences
› Western notions of the person are essentially individualistic,
emphasizing the centrality and autonomy of the individual actor
in all action
› Non-western notions of the person tend to be holistic, stressing
the interdependence between the individual and his/her
surroundings

Developmental/age differences
› Reflect the enculturation process – the gradual process by which
children adopt the dominant conception of the person within
their culture


Miller’s 1984 research does support this theory
Cross-cultural study comparing the attributions made
for prosocial and deviant behaviors
› American vs. Indian Hindu
› Sample: 8, 11, and15 year olds, and a group of adults
(mean age = 40.5 years)

Results:
› At older ages, Americans made significantly more
references to dispositions (M = 40%) than did Hindus (M <
20%)
 Most of the dispositions referred to personality
characteristics of the actor
› There were no significant differences that distinguished the
responses of 8-11 year old American children from the
responses of Hindu children (difference was an average of
2%)
Proportion of references to
dispositions
Cultural and developmental patterns of
dispositional attribution
0.4
0.35
0.3
0.25
Americans
0.2
Indian Hindus
0.15
0.1
0.05
0
8
11
15
Adult

Within each culture developmental trends indicated a
significant linear age increase in reference to:
›
›

These attributional patterns reflect Indian cultural conceptions
›

General dispositions among Americans
Context among the Hindus, emphasis on social roles and patterns of
interpersonal relationships
Emphasis on locating a person, object, or event in relation to someone or
something else
Ex. Explanations given by an American and a Hindu participant
regarding a story about an attorney who, after a motorcycle
accident, left his injured passenger in a hospital without
consulting with the doctor while he went on to work
American: “The driver is obviously irresponsible (disposition). The driver is
aggressive in pursuing career success (disposition).
› Hindu: “It was the driver’s duty to be in court for the client whom he’s
representing (context). The passenger might not have looked as serious
as he was (context/aspects of another person).
›



More recent cross-cultural studies have generally
confirmed Miller’s cultural hypothesis
It appears that the tendency to overrate
personal/dispositional factors of the agent in western
adults cannot be completely explained by cognitive
and experiential interpretations
The attribution “bias” may not simply be a cognitive
property or universal law of psychological functioning
› It may be culture-specific

Though the agent of action tends to dominate the
perceptual field for Americans
› The “person” does not seem to enjoy the same degree of
perceptual dominance among non-western people

It’s clear that attributions or lay explanations
for behavior and events are not only the
outcome of internal cognitive processes
› Rather, they are social phenomena that are
based on widely held and shared beliefs in the
form of social and collective representations

Hewstone (1989) suggested the concept of
an “attributing society” – our tendency to
seek explanations within our predominant
cultural framework

Our explanations for social phenomena are shaped not
only by culture but also by scientific and expert
knowledge
› The diffusion and popularization of scientific concepts
throughout society is occurring at a rapid rate through the mass
media
› Increasingly, expert knowledge contributes to the stock of
common sense which people draw upon to understand social
reality
› Thus, people can be regarded as “amateur” scientists,
“amateur” economists, “amateur” psychologists, etc., as they
draw upon this information to explain a range of phenomena
 Such as the causes of cancer, economic depression, or problems
in personal relationships
› Some of this knowledge becomes an integral part of mass
culture and, ultimately, what will come to be regarded as
“common sense”
The attributions that people make for societal events such as
social issues provide us with rich insight into a society’s prevailing
explanations or meaning systems
 Research on causal attributions for social issues has included
everyday explanations for poverty, unemployment, riots, and
health and illness
 Research has found that people in western industrialized
societies are more likely to attribute poverty to individualistdispositional causes, such as lack of effort and laziness, than
situational-societal causes

›
›
People primarily hold the poor responsible for their dilemma
In contrast, unemployment is predominantly attributed to social and
structural causes such as economic recession and government policies