Is
Isn’t
Not
anymore
But,
used
to be
Wish
they
were…

Of course, I’m talking about being “COOL”
› Cool is a highly prized virtue
› It has replaced goodness, quality, and servitude as the
modern, secular virtue

Cool is…
› An attitude of defiance to authority and tradition
› A “permanent state of private rebellion”
› An unashamed amalgam of “narcissism, ironic
detachment, and hedonism”

Cool is not just a contemporary concept
› Is has a long and distinguished heritage over many
centuries and continents

Of course if hasn’t always been called “cool”
› But, the attitude of cool can be identified in many places
and in many time periods

When we talk about attitudes we talk about
them as something people “have”
› “She has an attitude.” or “He has an attitude
problem!”
› We talk as though people can have an attitude
in the same way they have an ear, a toe, or a
new car
› We give “attitude” the status of a noun, like it is
something real and tangible, and that influences
the behavior of the “attitude-owner”
› We don’t really stop to think about what we
mean when we talk about attitudes

But, it hasn’t always been that way
› The concept of attitude can be traced back
to 1710 when it was first introduced to the
English language
› It was used by sociologist Herbert Spencer
and biologist Charles Darwin in the mid to
late 1800s
› Used in the early 19th century to refer to a
physiological state or physical orientation

So, “attitude” has not always been a
part of the common sense we take it to
be now



To date, attitudes have been the single most
researched topic in social psychology
But, the definitions, models, and theories of attitudes
used by social psychologists are more often than not
vague and inconsistent
So it is useful for us to be clear about what social
psychologists typically mean by “attitude”
› “… attitudes are defined at least implicitly as responses
that locate ‘objects of thought’ on ‘dimensions of
judgment’.”
› And…
› “…[an attitude is] a general and enduring positive or
negative feeling about some person, object or issue.”

Thus, attitudes are first and foremost
evaluations
› They convey what we think and how we feel about
some object, or attitude referent

All attitudes have a referent, an “object of
thought,” a “stimulus object”
› Referents may be specific and tangible like…

But, referents may also be abstract or intangible like
liberalism, equality and “awesomeness”
By indicating the attitude-holder’s
“orientation” to the referent, an attitude
conveys that person’s evaluation of the
referent
 Attitudes are expressed in the language of
“like/dislike,” “approach/avoid,” and
“good/bad”

› i.e. they are evaluative

Thus, when the object of the attitude is
important to the person, the evaluation of
the object produces an affective, or
emotional, reaction in that person

Important features of attitudes
› First, attitudes have specific referents, and thus
will only be relevant when a particular object,
person or issue is categorized as being attituderelevant
 This implies that activation of an attitude requires
at least some minimal cognitive effort
 However, this effort is so minimal that attitudes can
be activated and can function automatically
 It has been argued that an attitude is represented
in memory by
 An object label and rules for applying that label
 An evaluative summary of that object
 A knowledge structure supporting that evaluation

Second, attitudes are relatively enduring
› Although attitudes can change, as a result
of new experience or following some
persuasive communication, an attitude is not
usually considered to be a transitory
evaluation
 Rather, it is considered an expression of a
largely stable body of knowledge and
experience with a particular object, person or
issue

The dimensions of judgment upon which
attitudes fall may be universal or specific,
socially shared or unique to a single
individual
› Ex. All referents: George W. Bush, brussels sprouts,
Jack Daniels, liberalism, equality, and
“awesomeness” can be placed somewhere on
a dimension ranging from bad to good, or like to
dislike
› But, not all referents can be located on a
dimension from stupid to smart, or “cool” to
“lame”
The definition of attitude as evaluation is
becoming increasingly common, but still
not universal
 It is replacing a previously widespread 3part explanation of attitudes: the ABC
model

› Divides attitudes into 3 components: affect,
behavior, and cognition
› For this model, attitudes are predispositions to
respond to some class of stimuli with certain
classes of responses
› The 3 major classes of response are cognitive,
affective, and behavioral



Cognitive responses are the knowledge and beliefs a
person has about a particular stimulus object
Affective responses are simply how the person feels
about a stimulus object
Behavioral responses are simply overt behaviors
ATTITUDE
COGNITIVE
AFFECTIVE
Knowledge and
beliefs
Feelings and
emotion
BEHAVIORAL
Overt behavior


The model allows for these 3 responses to be inconsistent with
one another
›
Which is helpful because more often than not they are
›
Ex. Sally does not like olives (cognitive and affective), based on this, one
would expect Sally not to eat olives (behavior)
But, the degree of discrepancy between the 3 presumed
components of the same attitude held by one person towards a
single object is usually so large that the tri-component model has
largely been dismissed
 But, Sally is dating Tommy and really wants his mother to like her (cognitive)
 Sally goes to Tommy’s mother’s house for dinner and eats a dish filled with
olives (behavior) rather than tell Tommy’s mother she doesn’t like the food
Also, by defining behavior as just a component of attitude, any
supposed relationship between attitude and behavior is simply
defined away
 This is not a helpful resolution of one of social psychology’s major
problems:

›
Do attitudes predict behavior?

Attitudes are conceptualized as having a definite
structure with 2 parts
› Intra-attitudinal structure
 The structure that derives from the relationships between
the affective and cognitive (and perhaps behavioral)
components of a particular attitude
› Inter-attitudinal structure
 The structure that exists across attitudes and that organizes
the associations between attitudes to a range of different
attitude objects

Additionally, the concept of attitudes as knowledge
structures (or schemas) gives rise to important
properties of attitudes, including their accessibility
and potential for automatic activation

In many ways an attitudes is like a
schema
› Except that attitudes have the evaluation of
that referent as their defining and central
element

3 aspects of attitudes deriving from their
structure as schemas are particularly
important:
› Accessibility, activation, and the possibility of
ambivalent attitudes


Some attitudes come to mind more easily than others
Social psychology distinguishes between the availability of an
attitude and its accessibility
›
An attitude is available if it exists within a person’s cognitive structure
›
Attitude accessibility is the ease with which a particular attitude can be
retrieved from memory
 That is, if the person actually possesses that attitude
 Usually operationalized as the speed with which the attitude can be
assessed

The accessibility of an attitude at any given time is affected by:
›
›

The structural properties of the attitude (attitude strength)
And aspects of the context that serve to “prime” particular attitudes as
being relevant to that context
Accessibility is an important property of attitudes because
accessible attitudes govern behavior more strongly than do less
accessible attitudes
Most of the enduring differences between people in the
accessibility of particular attitudes are considered to be due to
the strength of their attitudes
 An attitude is said to be strong if there is a consistent, wellrehearsed link between an attitude object and its evaluation

›

The strength of an attitude does not refer to its extremity (i.e. the
degree of positivity or negativity in the evaluation)
›

Ex. If someone thinks about George W. Bush regularly, and each time
evaluates him in a moderately negative way, that person will be said to
develop a strong attitude towards George W. Bush
Rather, it refers to the frequency and consistency of its expression
The association between an object and its evaluation becomes
well developed through repetition making an attitude strong
›
Thus, strong attitudes are more accessible (more easily and quickly
activated) than weaker attitudes, and are therefore more influential on
behavior

Not all our attitudes are active at any one time
› It almost goes without saying that only a small set of our stock
pile of attitudes is active at any given moment


Attitudes must be activated, or “turned on” somehow
The process of attitude activation has received
considerable research attention and much of this work
has drawn from principles of cognitive psychology
› Attitudes are conceived as nodes in memory, connected in an
associative network
› Nodes are activated when we categorize some experience in
terms of them, and multiple nodes become connected through
experience of co-activation
 Ex. “kitten” and “cute” or “reality television” and “crazy”
› The more frequently any connection is experience the stronger
that connection becomes
Reality TV
Crazy
It has been argued, based on the associative network of
attitudes, that an attitude is an association between a
referent (object, person, issue) and its subjective
evaluation
 Thus, noticing an object in the environment and
categorizing it as an instance of a particular type of
object activates the node corresponding to that type of
object in memory
 After this initial activation, activation will “spread” from the
object node to the evaluation node associated with it, in
proportion to the strength of the association
 When the association between the object and its
evaluation is strong enough, simply noticing the object will
cause the evaluation to become activated

Reality TV
Identified as a reality show
Crazy
Some connections become so well
rehearsed that when one node is
activated, the other is automatically
activated
 Automatic activation is said to occur if
merely thinking about or noticing the
attitude object produces the evaluation
of that object even if evaluation of the
object is not intended at that time

Crazy

Several studies have found evidence for automatic
activation of attitudes
› Positive or negative attitude objects are used as primes for
positive or negative adjectives
› Results have shown clear evidence for automatic activation of
the evaluation associated with an attitude object
 Although, there is some disagreement about whether all attitudes
are automatically activated, or if this effect holds only for strong
attitudes

Bargh et al (1996)
› Found that both strong and weak attitudes produced an
automatic evaluation effect
 Positive attitude object primes facilitated faster reaction times for
responding to positive adjectives
 Negative attitude object primes facilitated faster reaction times
for responding to negative adjectives
Following the evidence that attitudes can be activated without
deliberate intention or conscious awareness, in recent years
social psychologists have begun to consider whether we may
have attitudes that are outside of our conscious awareness
 If conscious processing is not required for an attitude to be
activated and have an effect, maybe we don’t need to be
consciously aware of all the attitudes we have
 Attitudes we are unaware that we have are called implicit
attitudes

Implicit attitudes are like conscious attitudes in that they consist of a
relatively stable association between an object and its evaluation
› The BIG difference is that with implicit attitudes, the person holding the
attitude may be unaware that they hold it
›

Implicit attitudes are “…introspectively unidentified (or
inaccurately identified) traces of past experience that mediate
favorable or unfavorable feeling, thought, or action toward
social objects.”
›
In other words, implicit attitudes are evaluations of objects that people
are unaware they hold

If we accept that:
› A) we may hold attitudes that we are unaware
of, AND/OR
› B) that our automatic attitudes may be different
from those attitudes that we are willing to
deliberately endorse

A whole new problem for attitude theorists
is created:
› If our implicit and explicit attitudes are
inconsistent or even contradictory, which should
be considered to be our “true” attitude?

Wegner and Bargh
› Argue that people are inclined to see automatic responses as
more genuine because they are not subject to the selfpresentation concerns that may affect the expression of more
controlled responses
 Meaning, since we can’t change an automatic response to make
ourselves look better, it’s probably more genuine
› Also argue that automatically activated attitudes are more
important in some respects than controlled attitudes, because
they are more likely to predict behavior

Dovidio et al.
› Argued against a universal preference for automatic attitudes
› Propose instead that implicit (automatic) attitudes may be
important in predicting automatic behavior, while explicit
(controlled) attitudes may be more likely to predict deliberate,
considered behavior



If attitudes have multiple cognitive and
affective elements, it follows that these
different elements may not always lead to the
same evaluation
People are said to have ambivalent attitudes
when they have both positive and negative
evaluations of the same target
Ambivalent attitudes are perhaps particularly
likely for targets that are complex and
differentiated and with which we have
numerous encounters
› Such as groups of people rather than for simple
targets like olives

The possibility of attitude ambivalence has long been
acknowledged by psychologists
› Ambivalence has been associated with attitude instability and
amplification

Ambivalent attitudes are considered unstable
› The evaluation that is expressed in a particular moment will
depend on which elements of the attitude are most accessible
at that time
› And, as there are large variations in the evaluations associated
with different elements of ambivalent attitudes, the expressed
evaluations of the target are likely to be correspondingly
variable

Amplification
› Refers to the tendency for people to make more extreme
evaluations of targets toward which they hold ambivalent
attitudes and less extreme evaluations of targets when attitudes
are more straightforward


The realization that people often hold
ambivalent attitudes has led to a rethinking of
how attitudes should be measured
Traditionally, attitudes were measured by rating
objects on bipolar evaluative dimensions
› Good/bad, pleasant/unpleasant, warm/cold, etc.

However, these types of bipolar attitude
measure facilitate a “midpoint problem”
› Should neutral ratings on bipolar scales be
interpreted as reflecting ambivalence or indifference
towards the attitude object?

A person who selects a midpoint rating on
a bipolar scale that rates politicians from
good to bad may:
› A) believe that some politicians are very good
and others are very bad
› B) believe that politicians have some very good
qualities (like intelligence, public mindedness,
enthusiasm) and other very negative qualities
(like vanity, dishonesty, ruthlessness)
› C) not have strong opinions about politicians

So, how do we know?

Kaplan’s solution:
› Separate traditional bipolar semantic
differential scales into separate unipolar
measures of positive and negative attributes

Although this method has been adopted
by researchers in a number of areas,
notably in the measurement of
intergroup attitudes
› The use of bipolar semantic differentials is still
common in many areas of attitude
assessment

Work on the spread of activation of attitudes
largely assumes that each attitude exists as a
discrete node in an associative network
› Which has no structure other than horizontal
associations formed through repeated co-exposure
and rehearsal

However, in addition to these associated
connections, we can also think of attitudes as
existing in hierarchical relations to each other
› In this view, some specific attitudes are thought to be
instances of broader, more generalized attitudes

Ex. A person’s
attitude towards
paid maternity
leave may reflect
and/or be
derived from the
person’s more
general attitude
towards working
mothers
› Which may in turn
reflect the
person’s attitudes
towards various
aspects of
feminism

Kerlinger (1984) structure of political
attitudes
› Concerned with how social and political
attitudes are organized
› Prior research had suggested that these
attitudes could be arranged in a bipolar way,
ranging from liberal to conservative
 In this view, liberalism is the opposite of
conservatism
 Someone who agrees strongly with a liberal item in an
attitude scale is also presumed to disagree strongly with
a conservative item in the scale
Kerlinger suggested that the two ideologies of liberalism and
conservatism do not exist in opposition to one another, but,
rather are independent of one another
 This model starts with social referents – the objects of social and
political attitudes, such as abortion, real estate, trade unions,
money, racial equality and patriotism
 Some of the referents are said to be criterial for liberals and are
criterial for conservatives

›
A referent is said to be criterial for someone if it is significant, or salient, to
that person
Bipolar models would assume that referents criterial for liberals
are also negatively criterial for conservatives, and vice versa
 Kerlinger argues that liberals do not care about conservative
referents and conservatives do not care about liberal referents

›
›
In other words, criteriality is generally positive or neutral, not negative
As an ideology, liberalism has one set of criterial referents and
conservatism has its own set, and the two are independent

Evidence for this model relies on the factor analysis of
criterial ratings (both liberal and conservative) of a large
number of referents by a large number of people
› That is, the structure Kerlinger talks about is identified across, not
within people
 Although, it may also be represented as a structure within one
person
Factor analysis of criterial ratings typically produces about
a dozen first-order factors (like religiosity, racial equality,
civil rights, morality, etc.)
 When these first-order factors are themselves factor
analyzed, they produce two orthogonal second-order
factors: liberalism and conservatism

› These second-order factors are what Kerlinger labels as
ideologies – a collection of shared beliefs, attitudes and values
organized around some coherent core and often associated
with a particular group in a social structure

Ideologies are shared
› It is not possible for one person to “have” an
ideology
› They do not “exist” or “reside” within any one
person
› Rather, they are bodies of thought
themselves
 They only have life to the extent they are
shared, an hence can be said to be truly and
only social – they are the product of social
relations

Considering the structure of social and
political attitudes as being built upon
ideologies returns us to issues raised at
the start of the chapter
› Most of the work on attitudes has concerned
the intra-individual structure of attitudes
 Their accessibility
 Whether they are changed to maintain
consistency
 How, if at all, attitudes are related to behaviors
Work on the ideological nature of attitudes
is relatively scarce, but still important
 Intra-individual and inter-individual research
are complementary to each other

› Intra-individual, or micro-level, focuses on how
attitudes work
› Inter-individual, or macro-level, places attitudes
in a social context, and illustrates their
fundamental social character
 Attitudes are social, in origin, function and
consequence
 They originate in social life, they communicate meaning,
they are shared, and they have social consequences
What are the functions of attitudes?
 Psychologists have attempted to answer
this question in two different eras

› In the 1950s and from the mid-1980s on
 Not much was done in between
› In both of these eras, the focus is on the
functions attitudes serve for the individual
attitude-holder, and has ignored the
broader social functions

Katz (1960) proposed 4 functions of
attitudes
› 1) Knowledge function
 Similar to the common understanding of what an
attitude does
 Attitudes help us explain and understand the world
around us
 Definition: an attitude is a memorial representation
of an object, and associated with that
representation are rules about the labeling of the
object, an evaluative summary of the object and
a knowledge structure about the object
 The knowledge function of attitudes helps us know
the world around us

2) Utilitarian function
› Attitudes help us gain rewards and avoid
›
›
›
›
punishments
Utilitarian function emphasizes the social
consequences that follow from the expression of
certain attitudes
To be “politically correct” is to hold and display
attitudes for utilitarian reasons
The idea that attitudes can have utilitarian functions
underscores the flexibility in people’s expression of
their attitudes
People are likely to alter their “attitude” to some
social object according to the social context they
are in

3) value-expressive function
› The expression of an attitude can sometimes be
no more than a public statement of what a
person believes or identifies with (probably
strongly)
 Ex. Political statements

Stickers on car windows

Slogan T-shirts

Uniforms or sports teams

Displayed clothing labels




These are all public signs intended to convey a
message about the owner
They signal to the world that you’re a Lions fan,
a mother, support world peace, miss Dale
Earnhardt, that you can afford designer cloths
and don’t buy from K-Mart, etc.
There is no real point to these expressions other
than to tell the world something about who
you are
Thus, attitudes serve a value-expressive
function

4) ego-defensive function
› These attitudes are usually deep-seated, difficult
to change and hostile to the attitude object
 Ex. Homophobia and xenophobia
› According to Katz, at least some people who
hold such attitudes do so because they are
unconsciously denying some aspect of their own
self
 Ex. Homophobics may be so hostile to homosexuals
and homosexuality because they deny and don’t
want to confront aspects of their own sexuality
› Attitudes that serve this function thus project
outward what are really internal, intrapsychic
conflicts

Ex. Adams, Wright and Lohr (1996)
› Measured the sexual and physiological arousal of
heterosexual men while they viewed sexually explicit
videos of men engaged in homosexual activity
› Results:
 Participants who had previously identified as holding
negative attitudes towards homosexuality and
homosexual men showed greater sexual arousal while
watching the videos than did the participants with more
positive attitudes towards homosexuality
› The researchers argued that the expression of
negative attitudes towards homosexuality provides a
defense against the assumed ego-threat
experienced by straight men who experience some
level of homosexual arousal

Attitudes may simultaneously serve more than one
function and may be held or expressed for different
reasons at different times
› Ex. A person’s attitudes to medically assisted reproduction
may largely serve a knowledge function, being based on
what that person knows about IVF, donor insemination,
etc. and depending on how much that person needs to
formulate an attitude to assisted reproduction
› But, the same attitude may also aid that person’s
relationship with an infertile relative
› Or, it may also be the expression of more deeply held
beliefs about God’s will and the Chruch’s position on
procreation
› Or, it may also reflect that person’s own, perhaps
unconscious, conflicted sense of their own sexuality or
fertility, or doubts about their genealogy

Several theorists have reformulated Katz’s typology of
attitudes
› Smith et al.
 Object-appraisal function = Katz’s knowledge function
 Externalization function = Katz’s ego-defensive function
 Social adjustment function = Katz’s value-expressive and utilitarian
functions
› Shavitt
 Makes a more definite tie between attitudes and individual and
social identities
 Has shown how the success of attempts to change an attitude
depends on the function that attitude serves the owner
 Utilitarian function = Katz’s knowledge and utilitarian function and Smith et
al.’s object-appraisal function
 Social identity function = Katz’s value-expressive function and Smith et al.’s
social adjustment function
 Self-esteem maintenance function = Katz’s ego-defensive function and Smith
et al.’s externalization function

Herek proposes 2 different kinds of attitudes
›
Evaluative attitudes: attitudes in which the attitude object is an end in itself
and the attitude functions to allow the individual access to the object itself
 Experimental and specific – based on and restricted to a single object (Ex.
The dog next door that bit you)
 Experimental and schematic – based on experience with particular objects,
but generalized to a class of objects (Ex. Generalizing your attitude towards
the dog next door to all dogs)
 Anticipatory – based on expected, rather than direct, experience (Ex. You’ve
never been bitten by a dog, but expect to be bitten if you come into
contact with one)
›
Expressive (or symbolic) attitudes: Attitudes in which the attitude object is a
means to an end, by providing social support, increasing self-esteem or
reducing anxiety
 Social expressive – based on the individual’s need to be accepted by others
(Ex. Dressing a certain way to fit in with a new group of friends)
 Value expressive – based on the individual’s need to define self by expressing
important values and aligning self with important reference groups (Ex. After
joining a Greek organization, adopting that organization’s ideals and wearing
their letters)
 Defensive – based on the individual’s need to reduce anxiety associated with
intrapsychic conflicts (Ex. Expressing hostility towards homosexuals to reduce
anxiety caused by one’s own homosexual feelings)


Herek suggests that strategies to change
attitudes must consider whether attitudes are
held for evaluative or expressive/symbolic
reasons
Also, attitudes formed for different functional
reasons are likely to be more or less resistant to
change
› Ex. Few Australians have a lot of direct contact with
Aboriginal people, but this does not prevent them
from forming strong anti-Aboriginal attitudes
 Because these attitudes are based on anticipatory
rather than experience, they are hard to disconfirm
through direct experience, and hence are hard to
change


Note that all the functions of attitudes we have
discussed emphasize the functions for the individual
attitude-holder
These functions are mostly theorized at an individual
level
› That means the “social” functions are conceptualized in
terms of the social consequences for an individual holding
or expressing a particular attitude

Attitudes also serve social functions
› But, there is relatively little research in the social cognitive
tradition that directly investigates the social functions
attitudes serve
› So, we’ll talk about the social understandings and their
functions a little later when we discuss other theoretical
perspectives

One of the most enduring questions for social psychologists
regards the relationship between attitudes and behavior
›
The common-sense view would suggest that attitudes directly cause a
person to act in a particular way
 Ex. If you know someone feels strongly about practicing safe sex, you could
predict certain things about this individual’s sexual activity like using
condoms and birth control

But, social psychologists know that the relationship between
attitudes and behavior isn’t this simple
In reality, behaviors appear to be unrelated to attitudes just as often as
they appear to be related
› Also, behaviors can “cause” attitudes as much as attitudes can “cause”
behaviors
› Additionally, some researchers have argued that we need to distinguish
between different types of attitudes, like attitudes towards targets and
attitudes towards behaviors
›
 Ex. In trying to predict people’s sun protection behavior, it is necessary to
consider both their attitudes towards skin cancer AND their attitudes
towards applying sunscreen

Richard LaPiere (1934) was the first to present evidence
that the attitudes a group of people express toward a
certain object do not particularly correspond to their
behavior towards the same object
› In the early 1930s LaPiere traveled with a Chinese couple around
›
›
›
›
the US west coast, staying in inns and campsites
This was a time of strong anti-Chinese feelings throughout the US,
yet the trio was only refused accommodation once
After the trip, LaPiere wrote to all the managers of the
establishments they had visited, and others they hadn’t stayed
at, asking if they would accept Chinese guests
More than 90% claimed they would not accept Chinese guests
Thus, there was a disparity between the expressed attitudes of
the managers and their overt behaviors

This disparity has been found in many studies
over the years
› Wicker (1969) summarized the results of 32 different
studies, each of which measured individual’s
attitudes toward an object and a direct measure of
behavior towards the same object
 The attitude-behavior correlations reported in these
studies rarely exceeded +0.3, were often close to zero,
and were even negative on some occasions

Thus, at best, attitudes appear to explain up to,
but rarely more than, 10% of the variance in
behavior
› So if attitudes are so worthless in predicting behavior,
what use is this construct to us?

Social psychologists have attempted to
understand Wicker’s results
› Largely examining under what conditions we do
expect attitudes to be related to behavior and
under what conditions the two should be unrelated

Social psychology has produced 2 broad
classes of response to Wicker
› Many have attempted to work with the relationship
between a single attitude-behavior pair, attempting
to find when the link is strong and when it is not
› Others have attempted to formulate and test a more
elaborated model of the general link between
attitudes and behavior

Many variables influencing the strength of the
attitude-behavior link have been identified
› First, attitudes about an object formed through direct
experience appear to be more strongly associated
with behavior towards that object than attitudes not
formed through direct experience
 It has been suggested that this stronger link is due to:
 Attitudes formed through experience are held with more
confidence, clarity and certainty
 These attitudes are more accessible (able to be brought into
consciousness easily) and stronger
 And these attitudes are automatically activated in the
presence of the attitude object

Second, it has been suggested that
more stable attitudes show greater
attitude-behavior consistency than less
stable attitudes
› This proposal has 2 components
 1) the greater the time between measuring
the attitude and measuring the behavior, the
less strong the attitude-behavior link will be
 2) the attitude-behavior link will be stronger for
more general attitudes, as opposed to more
specific attitudes

Lastly, several individual differences have been found to affect
the strength of the attitude-behavior link
›
Self-awareness
›
Self-monitoring
›
Reasons for attitudes
›
Should attitudes correspond to behavior?
 People who have been made self-aware (usually by placing a mirror next
to them while they complete attitude scales) typically display much
greater attitude-behavior consistency than do people not made selfaware
 People who are high in self-monitoring (people who monitor and regulate
their own reactions through the reactions of others) typically show lower
attitude-behavior consistency than those people who are low in selfmonitoring (people who monitor internal reactions, rather than others’
reactions)
 Being asked to provide reasons for attitudes may lower the consistency
between people’s attitudes and behaviors
 There is also evidence of cross-cultural differences in the tendency to
believe that attitudes should correspond to behaviors

Festinger’s (1957) theory of cognitive
dissonance helps explain how people change
their attitudes because of their behaviors,
rather than the other way around
› The theory simply states that if a person holds two
cognitions that are psychologically inconsistent (not
necessarily logically inconsistent), this inconsistency
create dissonance that is uncomfortable, and the
person is motivated to reduce the dissonance

Dissonance can be reduced by changing
either or both of the cognitions, or by
introducing a new cognition

Ex. If I smoke and if I also know that
smoking is bad for my health I should
experience dissonance because these two
cognitions are psychologically inconsistent
›

The dissonance I experience can be
alleviated by changing one of the two
cognitions or by introducing some new
cognition
›
›

Note that there is nothing logically inconsistent,
just a psycho-logical discrepancy
Ex. I could give up smoking, but that’s a difficult
and unlikely thing to happen
Ex. I could alter my cognition that smoking is
bad for my health (it’s not uncommon for
smokers to argue that the evidence against
smoking is not as strong as public health
campaigns make out)
Or, I could introduce some new cognition
›
Ex. I could accept that smoking is bad for my
health, but then get out of it by claiming that I
smoke to relieve stress and gain pleasure, or
that I am addicted to smoking and therefore
unable to quit
Thus, we can see that if people engage in a particular
behavior, for whatever reason, they are likely to alter their
attitudes to correspond to the committed behavior
 Ex. A young child attending a mostly white primary school
with few minority children joins in with a group of older
children to tease one particular minority child

› The young child may not have had any attitude at all toward
minorities before teasing this one child, but it is unlikely that this sill
be the case after the incident
 Especially if the unpleasantness of the behavior is pointed out by
a teacher or parent or another child
› Rather, the child will alter (or in this case invent) his/her attitude
to correspond to the behavior
› Any dissonance will be displaced by another cognitive change –
in this case, by developing a negative attitude to the minority
child, and maybe toward minorities in general

On a bigger scale, this principle has been
applied to the dehumanization of the Viet
Cong by American soldiers and the
massacre of Vietnamese civilians at My Lai
by American infantrymen
› “It never happened and besides they deserved
it”
Attitudes follow behavior, not the other way
around
 According to cognitive dissonance
theorists, this principle applies to any
instance where a person engages in a
behavior that does not correspond with any
attitude held before that behavior occurs
 Cognitive dissonance is especially likely
when one of the conflicting cognitions is
about the self


Cognitive dissonance theory is known as a
consistency theory
› It assumes that inconsistency is unpleasant, and
that people are motivated to achieve
consistency and balance

This assumption has been challenged by
several critics
› Argue that the desire to achieve and maintain
consistency is a western phenomenon
› That cross-culturally, people are far more
tolerant of cognitive and interpersonal
inconsistencies than cognitive dissonance theory
assumes
Cognitive dissonance theory has been
challenged in its explanation of attitude
change and the relationship between
attitudes and behavior
 One challenge is self-perception theory
developed by Daryl Bem (1967, 1972)

› Developed to explain the same events as
cognitive dissonance theory, but without
including an explanation of the psychological
processes involved
› Bem argued that we infer our own attitudes to
object in the same way we infer others’
attitudes, through the process of attribution

Attribution theory suggests that observers attribute
attitudes to an actor which corresponds with the actor’s
behavior, and that this tendency is stronger when the
action is chosen freely by the actor
So we form an attitude about a person that is similar to their
behavior
 And, we tend to do this more often when the actor isn’t
forced to do the behavior, but is doing it of their own freewill


Bem proposes a similar process in inferring our own
attitudes

The major hypothesis of self-perception theory is that “in
identifying his [or her] own internal states, an individual
partially relies on the same external cues that others use when
they infer his [or her] internal states.”

Meaning, we think about how others would feel about us due
to how we are behaving, and use the same information to
figure out how we feel
Cognitive dissonance theory vs. selfperception theory… which, if either is
right???
 To figure it out, we should probably
design and conduct studies that pit the
two theories against each other

› Sadly… this is impossible
› Many have tried, but there is always some
other explanation that could explain the
results… and that’s not how science works

A review of research on attitude change finds that:
› Self-perception processes seem to operate when the
behavior is seems acceptable, or falls in the “latitude of
acceptance”
› But when the behavior seems unacceptable, or falls
outside the “latitude of acceptance,” cognitive
dissonance processes appear to operate

“latitude of acceptance” and “latitude of rejection”
are terms from social judgment theory
› Social judgment theory suggests that the dimension
characterizing the range of possible attitudes to a
particular object may be divided into these two latitudes
› Any one person’s latitude of acceptance includes all
those attitudes that person finds acceptable
› All those attitudes the person finds unacceptable fall into
the latitude of rejection

The second reaction to Wicker’s (1969)
conclusion that attitudes are worthless
› Attempted to theorize the attitude-behavior
relationship more fully than previous research

Major theory by Fishbein and Ajzen:
theory of reasoned action
› Theory about behavior (“action”), and thus
the attitudes that are relevant to this model
are attitudes towards behavior


Argued that attitudes do not predict behaviors per se,
but, rather, behavioral intentions
› In this theory behavioral intentions directly predict
behavior
Behavioral intentions themselves are a function
of attitudes to the behavior and what Fishbein
and Ajzen called subjective norms
› Subjective norms refer to what the individual actor
believes his or her significant others believe he or she
should do

Thus, the theory of reasoned reaction is really
only applicable to behaviors under voluntary
control
Attitudes and subjective norms make up
behavioral intentions
 Behavioral intentions predict behavior

Behavioral Intentions
Attitudes
about the
behavior
Behavior
Subjective
Norms


The theory of reasoned action has been widely used and has considerable
empirical support
Has been applied to a wide range of behaviors (actions)
›
›
›
›

The decision to have an abortion
The decision to breast or bottle-feed a baby
Smoking marijuana
Attending church
Meta-analyses have found strong support for this model in correlations
between constructs
These correlations are much
larger than the .30 maximum
reported by Wicker (1969)
Behavioral Intentions
.68
Attitudes
about the
behavior
.53
Subjective
Norms
.51
Behavior

Of course, this theory also has critics
› It has been argued that personal norms (individual beliefs
about the appropriateness of particular behaviors) and
behavioral norms (what everyone else does) are just as
important in the formation of behavior as are subjective
norms
› It has also been suggested that even behavior that is
under voluntary control does not necessarily conform to
the model
 Some behavioral routines are so scripted and rehearsed
that they are done mindlessly
› Similarly, it has been shown that behaviors that have
occurred in the past are more likely to happen again,
simply because they have been performed before and
despite the person’s intentions to behave differently
 Ex. New Year’s resolutions… its difficult to stop certain
behaviors despite all the best intentions to change

Finally, Ajzen revised the original model of the
theory of reasoned action into the theory of
planned behavior
› This theory accommodates the fact that behaviors
are often not under voluntary as assumed by the
theory of reasoned reaction

The theory of planned behavior still holds that
behavior intentions are the central link
between attitudes and behavior, and that
behavioral intentions are the product of
attitudes toward the behavior and subjective
norms
› But adds an important third factor: perceived
behavioral control

Perceived behavioral control refers to
the person’s perception of the ease or
difficulty of performing the behavior
› Some behaviors are easy to do once you
decide to do them, others are hard
› Some behaviors are easy not to do once
that has been decided, others are much
harder not to do

Perceived behavioral control affects the formation of
behavioral intentions, and also, importantly, directly
affects the production of behavior itself,
independent of behavioral intentions
Behavioral Intentions
Behavior
Attitudes
Subjective
Norms
Perceived
Behavioral
Control

A meta-analysis of 185 studies that used the theory of planned
behavior
The model overall accounted for 27% of the variance in people’s
behavior
› Accounted for 39% of the variance in people’s behavioral intentions
› Also, the theory of planned behavior predicted people’s behavioral
intentions better than the theory of reasoned action
›
 That is, perceived behavioral control added to the prediction of
behavioral intentions over and above the effects of attitudes toward the
behavior and subjective norms
Behavioral Intentions
Attitudes
Perceived
Behavioral
Control
39%
Subjective
Norms
Behavior
27%
So far we have discussed attitudes in terms
of the functions they serve in “locating
‘objects of thought’ on ‘dimensions of
judgment’”
 However there is another important sense in
which attitudes are “locating”:

› Expressing (and concealing) certain attitudes
are a powerful way people can locate
themselves relative to others in the social
environment
 That is, attitudes serve to locate an individual within
the social matrix
Earlier when we talked about the value-expressive function of
attitudes, it was from the point of view of the attitude-holder
 But viewed from the other side, the expression of values through
attitudes is required for social cohesion and evaluation

A group member who is reluctant about expressing an attitude on an
issue important to the group will not usually be allowed by the group to
remain silent on the issue
› The group will enforce or extract an expression of the attitude from the
group member
› This expression is an important sign of the individual’s position relative to
the group
› It is important to the group that individual’s positions are not far off from
the group’s “prototypical” position
›

Enforcing or extracting attitudes can be a potent form of social
control, requiring a demonstration of group allegiance from the
individual
›
Ex. The social group to which you belong has a general negative attitude
towards heavy metal music, if you do not express a similar opinion,
sooner or later a group member will ask you your opinion, and will expect
it to be similar to the overall opinion of the group
Attitudes have long been considered to be embedded in
social relationships
 Heider (1958) influential balance theory:

› Specifically, argues that we are motivated to hold attitudes that
are similar to those of people we like, and different from those of
people we dislike
› Also, that our liking or disliking of others can arise from our
perception of the extent to which we share the same attitudes

Social identity theory and self-categorization theory have
taken this idea further
› Arguing that similar attitudes can provide the basis for
psychological group formation
 In other words, that shared attitudes can provide the ground on
which we divide the world into “us” and “them”
› Ex. Pro-choice and pro-life groups
 Attitudes towards abortion are the basis of these groups


Groups can also provide a context in which
people can compare their attitudes to those
held by similar others, to evaluate the
appropriateness of their own attitudes
Festinger highlighted the fundamentally social
nature of attitudes and the effect of social
groups on the development of attitudes
› Stating that “an attitude is correct, valid, and proper
to the extent that it is anchored in a group of people
with similar beliefs, opinions, and attitudes”
› Simply put, knowing that people similar to us have
attitudes like ours validates our attitudes

In thinking about how SIT and SCT can help us
understand people’s attitudes it is helpful to
distinguish between attitudes towards social
groups and attitudes towards other kinds of
social and non-social objects
› So, attitudes toward social groups vs. attitudes
toward individuals and other objects

SIT and SCT are fundamentally concerned with
the development and expression of attitudes
towards one’s own and other social groups
› But, they have had much less to say about how
group membership might influence a person’s
attitudes towards other things

SCT can be thought of as providing a
theoretical account of the influence of social
identities on attitudes:
› To the extent that the expression of certain attitudes
constitutes a group norm, SCT would predict that
identification as a member of that group should lead
to the expression of those normative attitudes

But, empirical research into this proposal has
been very limited
› Tends to focus on the ways shared attitudes in a
group tend to become more extreme (group
polarization)
› And the role of social identity in making certain
attitudes and norms salient in certain social contexts
Social psychologists have known for a long time that people’s
attitudes tend to become more extreme following discussion
 The group polarization effect refers to the phenomenon in which
attitudes become more extreme following contact with others
who share the same initial evaluation in their attitude (i.e. an
initially positive or negative evaluation)
 Although not all group discussion of attitudes produces group
polarization, it is a robust phenomenon that many theorists have
attempted to explain
 Most SIT accounts of the group polarization effect rely on the
notion that group discussion provides either

›
A) informational influence: a new source of persuasive arguments in
favor of the initial position
 More information supporting the initial attitude
›
B) normative influence: information about others’ view that provides a
basis for social comparison
 Knowing that other people also hold the same view, which validates the
view further

Informational influence
› According to the persuasive arguments
account, discussion of attitudes with others
provides exposure to arguments and information
that may not have been considered in
developing the initial attitude
› Given the tendency for initial attitudes to be in a
particular direction (which is required for group
polarization to occur)
 It follows that these new arguments are likely to
support the initial attitude, and lead to a more
extreme position in the same direction as the initial
attitude

Normative influence
› The social comparison account of group polarization argues that
people seek to enhance their positive differentiation from the
group by enhancing their difference from other group members’
attitudes in the direction favored by the group
 That is, people want to have the most extreme attitude in the
group
› According to this view, polarization occurs because people
initially underestimate the extremity of others’ attitudes, and so
“re-adjust” their own attitudes to maintain a more favorable
(more extreme) position than other group members
 So, when everyone in the group is trying to have the most extreme
attitude, they think they have underestimated how extreme others
feel, and then make their attitudes more extreme… like a vicious
cycle
 Ex. In discussion of environmental conservation among people
who initially express “green” attitudes, people may see the
attitudes expressed by other people in the group becoming
increasingly “pro-green” as people compete for the position of
“greener than thou!”

Both of these accounts have been challenged by selfcategorization theory
›

According to SCT, group polarization is produced by group processes,
and cannot be reduced to either informational or social comparative
processes
Group polarization will only occur to the extent that people
psychologically identify with the group in which they are
discussing their attitudes and, will occur as a result of the
processes of group identification
In other words, when people identify themselves as members of a group,
they apply the prototypical characteristics, in this case attitudes, of the
group to themselves
› Prototypical characteristics are those that follow the principle of
maximizing between-group differences and minimizing ingroup
difference
› Thus, in cases where initial attitudes are already tending in a particular
direction (positive or negative), the more extreme attitudes (rather than
average group attitudes) will tend to be seen as prototypical
›
 Because they differentiate the group from outgroups taking an opposite
view

The SCT explanation thus explicitly
contradicts the social comparison
explanation
› By arguing that group polarization is
produced by conformity to a polarized
group norm
› Where social comparison suggests that
polarization is produced by the need for
positive distinctiveness

So to sum this up…
› Social identity theory: group polarization is due
to:
 New information supporting the attitude
 Or people’s desire to have the “biggest badest”
(i.e. most extreme) attitude in the group (i.e.
positive distinctiveness)
› Self-categorization theory: group polarization is
due to:
 The desire to maximize differences between groups
and minimize ingroup difference
 So attitudes become more extreme because they
serve to maximize the difference between groups
taking opposite perspectives
As we have already seen, the relationship between attitudes
and behavior is problematic
 Researchers working from the perspectives of SIT and SCT have
argued that understanding relationships between attitudes and
behavior requires careful attention to:

›
›

The salience of social identities
And the attitudes and behavioral norms associated with these social
identities
Different social identities are associated with different attitudinal
and behavioral norms
›
Ex. Soccer hooligan social identity is associated with extreme positive
attitudes towards a favorite team and extreme negative attitudes
towards every other team, as well as hostile behavior towards fans of
other teams


Some researchers have concluded that the influence of
social factors (ex. Subjective norms) is relatively weak
compared to the influence of personal factors
However, Terry et al. have challenged this conclusion
› Arguing that the concept of social norms in these models is
problematic and is inconsistent with the way social norms are
›
›
›
›
understood in SIT and SCT
Social norms are tied to specific reference groups which are
behaviorally relevant in particular situations
As such, particular social norms would only be expected to
influence behavior when the social identities with which they are
associated are salient in a particular instance
Ex. Soccer hooligans probably wouldn’t do this…
when they are at work, and their social identity of “soccer
hooligan” isn’t salient

Terry et al. (2000) examined the role of group norms in
moderating the attitude-behavior link
› Manipulated normative information about the intended
career choices of psychology students
 Students were asked to nominate their preferred career in
psychology from a list of three options (clinical psychology,
organizational psychology, and sports psychology)
 Students were then provided with normative information
that suggested the career preferences of other
psychology students were either similar (norm congruent)
or different (norm incongruent) from their own
› Students were then given an opportunity to attend an
information session on only one of the three career paths,
› And were asked about their willingness to take a variety of
steps to gain further information about their preferred
career path

Results
› Most students chose to attend a session about the
career they had selected
› BUT participants in the norm congruent condition
were significantly more likely to attend an
information session on their original career
preference than were participants in the norm
incongruent condition
 Meaning that information about norms effected the link
between initial attitudes (career preference) and
behavior (which information session was attended)
 When participants believed that their attitude was the
same as other students, it was more likely that they
would behave in congruence with that attitude

In an earlier study Terry et al. (1999) found:
› Participants’ attitudes towards household recycling were
predictive of intentions to recycle
› But only when they were led to believe that their social group
held similar attitudes, as opposed to when they were led to
believe that a group of random strangers also held this attitude

Taken together, these findings provide some support for
the argument that:
› The role of attitudinal norms in guiding behavior depends on the
extent to which a person identifies with the reference group with
which the norms are associated
 Students’ initial attitudes better predicted behavior when they
thought that their attitudes was the norm of other students
 People were more likely to intend to recycle when their social
group held similar attitudes
Throughout this chapter it should be clear
that social psychology has primarily treated
the attitude construct as an individual
phenomenon
 Attitudes have primarily been
conceptualized as individual and internal
cognitive and affective states

› Or, as behavioral intentions and predispositions

This contradicts how attitudes were
originally formulated when the construct
entered the social sciences

Early social scientific approaches to attitudes argued that
attitudes provide the links that tie individuals to their social
groups
Giving them a social position and social heritage, and allowing them to
live socially
› Attitudes were NOT, for early theorists, mental structures
›
Sociologists accepted this view and extended it to argue that
attitudes, like all forms of meaning, arise through social
interaction and communication
 Mainstream social psychology has increasingly individualized the
attitude construct
 Only recently have analyses emerged which reestablish the
social nature of attitudes

An increasing number of theorists are again emphasizing that attitudes
originate in and emerge from social life itself, through our everyday
interactions and communications with others
› Further, some attitudes are widely shared, providing cultural meaning
and substance to everyday life
› Shared attitudes are relied on to make sense of the social world and to
orient ourselves to that world
›


The increasing individualization of the attitude
construct can be attributed to Gordon Allport’s
original edition of A Handbook of Social Psychology
Allport defined an attitude as a global stimulusresponse disposition for the purpose of explaining
differences in behavior in objectively similar situations
› That is, attitudes are why different people behave in
different way in similar situations

Allport’s view, along with the important development
of techniques to measure attitudes, was the start of
the individualization of attitudes
› The strong desire to measure individuals’ attitudes to just
about any and every topic
› And to search for individual differences in attitudes that
could predict differences in behavior
Attitudes had become objectified, reified cognitive
entities with a life of their own inside people’s heads
 As an individual cognitive and emotional predisposition,
the attitude construct took on a methodological
individualism that shaped the subsequent nature of
attitude theories in social psychology

› Meaning, because attitudes became individual predispositions,
research methods aimed to study attitudes at the individual level
and thus, theories of attitudes were conceptualized at the
individual level as well

It was argued that this lead not only to the
“individualization of the social” but also to the
“desocialization of the individual”
› Making social things individual and removing the individual from
society

Contemporary (individualistic) theories of
attitudes contrast the more recent emergence
of social representations theory
› Social representations theory reinstates the collective
and social nature of cognitive constructs like
attitudes, beliefs, and values

As previously defined, social representations
refer to the stock of shared common-sense
knowledge and beliefs people within a
collective use to orient themselves to the social
world
› Social representations are seen as the building blocks
used to construct and thereby understand social
reality



Some critics have argued that social representations
may be the same thing as attitudes
However Moscovici warns that social representations
are not simply “attitudes” about social objects
The concept of social representations has been
endowed with a different status than the concept of
attitudes
› In contrast to the traditional attitude construct, social
representations are theorized to be more complex
cognitive structure akin to “theories” or “branches of
knowledge” and beliefs
› Social representations are therefore much more than just
evaluations or judgments about specific objects or
referents
 They are frameworks of understanding that give rise to
more specific evaluative judgments

Ex. Discussion of emerging representations of biotechnology in
Europe in Chapter 2
The Eurobarometer Survey, a quantitative questionnaire, was not on its
own able to shed light on how favorable, negative, or ambivalent
attitudes towards biotechnology are generated
› In order to understand where these attitudes came from, researchers
had to examine underlying representations and understandings of
complex notions such as “nature” and “life” as well as past experiences
› These, in turn, were shaped by religious, scientific, and popular “sci-fi”
accounts and narratives that were present throughout society
› As such, “attitudes” towards biotechnology were shaped and framed by
pervasive cultural understandings
›

Thus, representations can help us understand how attitudes are
formed
›
Hierarchical models of attitudes, such as Kerlinger were developed to
account for the complex organization and structure of attitudes

Although social representations cannot simply be equated
with attitudes, they do have an evaluative dimension
› “…the shared understandings encapsulated in social
representations also provide a way of expressing evaluative
judgments in ways that appear grounded in objective features
of the target rather than subjective opinion.”
 Meaning, if a collective shares a representation of some object
that is either positive or negative, a person in that collective can
express an evaluation that appears to be based on objective
“facts” rather than the individual’s personal opinion

Ex. To the extent that “family values” are seen as positive,
describing a political party as having policies that
promote family values is a subtle way of expressing
endorsement of that party
Unless you’re Newt
Gingrich

Moliner and Tafani (1997): attempted to theorize the
relationship between attitudes and social
representations
› “… the evaluative components of a representation can
be regarded as pieces of information upon which
individuals rely when manifesting their attitudes towards
the represented object. In this view, the evaluative
components of the representations form the underlying
structure of the attitude.”

Importantly, this explanation claims that social
representations can have multiple evaluative
components
› Ex. Representations of abortion can comprise evaluations
of the rights of unborn fetuses, as well as evaluations of
women’s rights to control their bodies (just to name 2)

Moliner and Tafani argue that even when there is broad
social consensus about the positiveness or negativeness of
each component of a social representation
› Attitudes towards the object of that representation can still differ
if those attitudes are based on different components of that
representation
› Ex. People can base their attitudes toward abortion on different
components of the representation of abortion
 Someone who bases their attitude on the rights of women to
control their bodies, will likely have a different attitude towards
abortion than someone who bases their attitude on the rights of
unborn fetuses
› Also, in abortion debates, people do not usually try to undermine
the arguments of the opposing side
 Ex. Pro-lifers generally do not argue that women should not have
control over their bodies
› Instead, they work to promote the centrality of the component
of the representation their attitude is based on

An important and central function of representations and
attitudes is that they are a mechanism for the transmission
and communication of social beliefs and knowledge
› Communication and interaction are social processes that
functional approaches to attitudes have relatively neglected

The public expression of an attitude by an individual
usually provokes some form or reaction from those around
› The public reaction to an expressed attitude engages both the
individual and the public in a rhetorical dialogue
› Positions, views, beliefs, doubts, inconsistencies, related issues,
etc. are exchanged and debated

These processes force the individual to resolve
inconsistencies, to consider one attitude in relation to
many, to figure out what he/she believes in and how
strongly, to commit publicly to a position
› In sort, to think critically about his/her attitude and its object

Another social function of social beliefs and
representations is that they play an explanatory, and
hence justificatory, role in orienting the individual to
the social world
› An “attitude” of dislike and disdain of the poor, of the
unemployed, of people of a different class, of people of a
different color, serves not only to orient the individual to
that particular social object, but also to position that social
object, be it a person or a group, in social space

This helps to explain, justify and reproduce the social
system which produced those social positions, and to
defend the individual’s own social position
› We’ll discuss this topic more in depth in chapter 8

To recap:
› A social representation is a collectively constructed and shared
knowledge and understanding about a particular theme or issue
or topic
 It is almost atmospheric, and is certainly cultural
› It is something which all members of a collective can access and
use in understanding the events around them
› It orients people to social objects, and is then, in this sense, a
social “attitude”
› Importantly, it is the shared nature of attitudes as social
representations that allows members of a collective to identify
particular kinds of language and behaviors as evaluations
 Ex. In our society, the word “jock” is associated with being athletic
and generally less intelligent, this is a social representation and
there are certain attitudes associated with the word “jock”
 If we didn’t have this shared representation, we wouldn’t be able
to understand that calling someone a “jock” might be a
derogatory statement about a person’s intelligence
› Thus, without this sharedness, many of the sophisticated means
by which we are able to communicate our attitudes would
simply not work, because they couldn’t be recognized by others