9 Emotional factors in prosocial behavior

Introspection will convince most people that more than cognition is in-
volved in much prosocial behavior; often we engage in prosocial be-
havior because of what we feel, or what we think we shall feel, if we are
kind or considerate (or what we might feel if we failed to help the
needy). Feelings of concern or sadness for the needy or distressed ordi-
narily precede helping, sharing, or comforting, and pangs of guilt may
follow a failure to act prosocially. Prosocial behavior is not usually
based solely on logic, understanding, or reasoning; a number of emo-
tions are critical antecedents of altruism.

Empathy and sympathy

Empathy and sympathy have been defined in many ways (Eisenberg &
Strayer, 1987; Wispe, 1986). We define them primarily in affective or
emotional terms. Empathy is defined as a vicarious affective state that
stems from another's emotional state or condition and is congruent with
the other's state or condition. If a girl who observes a distressed peer
feels sad or distressed herself, she is empathizing. Although empathy is
an emotional response, it also involves cognitive skills such as the ability
to label another's emotional state or role taking (Feshbach, 1978;
Hoffman, 1982, 1984).
Sympathy is an emotional response to someone else's emotional state
or condition that consists of feelings of sorrow or concern for another,
but does not refer to feeling the same emotion as the other person.
Sympathy, which has often been called empathy, is the basis of much
prosocial responding. However, because the two terms have not been
differentiated in most research, we generally use the term "empathy" to
signify both types of reactions.

130

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 Emotional factors in prosocial behavior 131

Theories of empathy

Many writers consider empathy to be the chief motive for altruism

(Aronfreed, 1970; Batson, 1987; Hoffman, 1982, 1984). In Aronfreed's
detailed explanation, empathy is acquired early by conditioning or as-
sociation, that is, by repeated pairing of the child's feelings of pleasure
or pain with someone else's expression of those feelings (see chapter 2).
Some psychoanalytic theorists believe that empathy develops from
early infant - caretaker interactions as the caretaker's moods are com-
municated to the infant by touch, tone of voice, and facial expressions
(Ekstein, 1972; Olden, 1958; Sullivan, 1940). Others use the principles
of conditioning to explain the significance of the infant-caretaker
relationship in the genesis of empathy - especially if the infant is closely
attached to the caretaker:

When baby and mother face a common cause of distress, or when emotional
upset in the baby calls forth an empathic response in the mother, then from the
baby's point of view, perception of the mother's distress will become conditioned
to his own feelings of distress. As a consequence, the later perception of her
distress will evoke a similar response in him. We may presume that a good deal
of such conditioning goes on in the earlier years (and indeed throughout life).
But it is also most likely to occur within a close mother-child attachment, since
the devoted mother is constantly mirroring back to her child his own emotional
states in an amplified and enriched form. In her efforts to get into tune with him
she may well exaggerate her expressions of pleasure or distress in response to his
. . . . It is difficult to overestimate the importance of this continual and intimate
interaction of mother and child for the development of sensitive empathic re-
sponsiveness to others. [Wright, 1971, p. 135]

Martin Hoffman (1975b, 1982,1984) has proposed an intriguing theory

of growth and change in altruistic motivation in the early years that em-
phasizes both the cognitive and affective aspects of empathy:

The central idea of the theory... is that since a fully developed empathic reaction
is an internal response to cues about the affective states of someone else, the
empathic reaction must depend heavily on the actor's cognitive sense of the
other as distinct from himself which undergoes dramatic changes developmen-
tally. The development of a sense of the other. . .interacts with the individual's
early empathic responses to lay the basis for altruistic motivation. [Hoffman,
1975b, p. 610]

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 132 The roots of prosocial behavior in children

Empathic responses - particularly empathic distress, defined as "ex-

periencing another's painful emotional state" - develop early in infancy
as a consequence of either built-in (i.e., biologically determined) human
tendencies toward empathy or early classical conditioning:

. . . cues of pain or displeasure from another or from his situation evoke associ-
ations with the observer's own past pain, resulting in an empathic affective
reaction. A simple example is the child who cuts himself, feels the pain, and
cries. Later, on seeing another child cut himself and cry, the sight of the blood,
the sound of the cry, or any other distress cue or aspect of the situation having
elements in common with his own prior pain experience can now elicit the
unpleasant affect initially associated with that experience. [Hoffman, 1975b,
p. 613]

Hoffman believes that infants are capable of experiencing empathic

distress before they can differentiate themselves from others:

For most of the first year, witnessing another person in distress may result in
a global empathic distress response. Distress cues from the dimly perceived
"other" are confounded with unpleasant feelings empathically aroused in the
self. Since infants cannot differentiate themselves from the other, they may at
times act as though what happened to the other happened to themselves. An 11-
month-old girl, on seeing a child fall and cry, looked as if she were about to cry
herself and then put her thumb in her mouth and buried her head in her mother's
lap, which is what she does when she herself is hurt. [Hoffman, 1984, p. 285]

When babies cognitively differentiate themselves from others, at

about 1 year of age or a little later, they distinguish between their own
and others' distresses. Then the child's empathic distress "may be trans-
formed at least in part into a reciprocal concern for the victim. That is,
they may continue to respond in a purely empathic manner - to feel un-
comfortable and highly distressed themselves - but they also experience
a feeling of compassion for the victim, along with a conscious desire to
help because they feel sorry for the victim and not just to relieve their
own empathic distress" (Hoffman, 1984, p. 287).
However, for some time, although aware of others as separate indi-
viduals, toddlers see the world only from their own perspective and do
not realize that others have their own traits and feelings. Instead, tod-
dlers attribute their own feelings to others and may therefore use
inappropriate means in attempting to relieve another's distress. For
example, a 12-month-old brought his own mother to comfort a crying
friend, even though the friend's mother was equally available, and

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 Emotional factors in prosocial behavior 133

another toddler offered his own favorite doll to cheer up an adult who
looked sad.
At about the age of 2 or 3 years, children begin to consider others as
distinct physical entities with their own feelings, thoughts, and emotions.
Some children this age are capable of rudimentary role taking and are
more highly motivated to put themselves in another's place and find the
real source of another's distress. Hence, they can respond in ways that
may relieve the other's distress rather than their own. With the devel-
opment of language, which enables children to derive meaning from
symbolic cues of affect such as a written or verbal communication, "they
can empathize with an increasingly wide range of emotions including
complex ones like disappointment and feelings of betrayal" (Hoffman,
1984, p. 286).
Until they are between 6 and 9 years of age, children's empathic re-
sponses are restricted to another's immediate, transitory, and situation-
specific distress. With greater cognitive maturity and awareness of their
own and others' continuing existence, children begin reacting to general
conditions (including deprivation, oppression, illness, and incom-
petence) as well as to immediate distress. "With further cognitive devel-
opment the person may also be able to comprehend the plight not only
of an individual but also of an entire group or class of people - such as
the economically impoverished, politically oppressed, socially outcast,
victims of war, or mentally retarded" (Hoffman, 1975b, p. 617).
Hoffman's theory is appealing because it views empathy in a broad
developmental perspective, changing with increasing age, with advancing
cognitive capacities, and with maturation of affective processes. Some
aspects of the theory are supported by research findings, such as the shift
from helping behaviors based on an egocentric view of the world (e.g.,
confusion regarding helping that would make oneself, rather than an-
other, feel better) to those based on an understanding of others' needs
as well as one's own. However, some of the theory is speculative, and
some postulates are extremely difficult to test empirically.

Research findings

A major tenet of a number of theories is that empathy and sympathetic

concern are critical factors mediating prosocial actions. This idea has
been tested systematically, and there is ample evidence to support it.

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 134 The roots of prosocial behavior in children

The clearest findings with regard to the role of empathy in prosocial

behavior have come from studies with adults. One of the major investi-
gators in this area, Batson, has made an important conceptual differen-
tiation in his work - between empathy and personal distress. Personal
distress is an affective reaction to another's emotional state or condition
that is experienced as a personally aversive emotion such as anxiety,
alarm, or worry. According to Batson, empathy results in altruistic,
other-oriented motives (e.g., the desire to reduce another's distress),
whereas personal distress leads to the egoistic motivation to reduce
one's own distress. Thus, people who experience personal distress will
sometimes assist others, but primarily when helping is an easy way to
relieve their own distress. For example, individuals experiencing per-
sonal distress are likely to help if helping is an easy way to terminate
exposure to the aversive cues emanating from the needy other. How-
ever, if it is possible to escape easily from the aversive stimulus (e.g., the
needy other), they will tend to do so rather than assist.
In one study (Toi & Batson, 1982), college women were asked to
evaluate pilot radio shows; only one person would evaluate each pro-
gram. Half of the students were told to
imagine how the person in the news feels. Try to take the perspective of the
person who is being interviewed, imagining how he or she feels about what has
happened and how it has affected his or her life. Try not to concern yourself with
attending to all the information presented. [Toi & Batson, 1982, p. 285]

This manipulation was designed to induce role taking and empathy. The
remaining students were instructed in a manner designed to preclude
sympathizing:
Try to be as objective as possible, carefully attending to all the information pre-
sented about the situation and about the person who is being interviewed. Try
not to concern yourself with how the person being interviewed feels about what
has happened. [Toi & Batson, 1982, p. 285]

The first radio show was an affectively neutral newscast (containing

announcements of upcoming events at the university). After hearing it,
students described their emotional reactions to the newscast. Some of
the adjectives referred to feelings of personal distress (e.g., worried,
alarmed, anxious), whereas others assessed empathy (e.g., empathic,
soft-hearted). Then the students heard the next pilot radio show. It con-
cerned another student (Jane) who had been in an automobile accident

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 Emotional factors in prosocial behavior 135

and had broken her legs. She discussed the troubles she was having with
her psychology class and her fear that she would not graduate. After
hearing this tape, the students again reported their emotional reactions
after the experimenter left of the room.
When the experimenter returned, she told each subject that the pro-
fessor in charge of the study had asked her to give the subject two letters,
one from the injured student and one from the professor. The profes-
sor's letter said that the viewer might be able to help the injured student
by sharing notes from the psychology course lectures Jane had missed.
The second letter, ostensibly from Jane, described her need and said
that any assistance would be welcomed, but indicated that it was okay
not to help. For half the subjects, the letter indicated that Jane would
not be coming back to class (labeled the easy escape condition, because
subjects could escape from dealing with Jane); for the remainder, the
letter indicated that Jane would return to class, and so the subject might
have future contact with Jane (labeled the difficult escape condition).
The subject then was asked to indicate on a separate sheet of paper
how many hours she could assist and to put this slip of paper in a sealed
envelope to be mailed to Jane.
The results of this study (and similar studies) demonstrated that sub-
jects who were in the empathy-inducing condition were more sympathet-
ic than were the other subjects. Moreover, they helped approximately
the same amount in the easy and difficult escape situations. However,
persons instructed to remain objective helped less in the easy escape
condition than in the difficult escape condition. Batson suggested that
those instructed to listen objectively experienced personal distress be-
cause they were relatively unlikely to experience the needy other's
emotional state, and therefore they helped only when doing so was the
easiest way to relieve their feelings of personal distress. Batson also
found that the amount of reported sympathy correlated positively with
helping behavior in situations in which it was easy to escape from ex-
periencing the other's distress (i.e., to escape empathizing), whereas the
reported amount of personal distress was not correlated with helping
(Batson, 1987).
In another study involving adults (Krebs, 1975), each of the 60 partici-
pants (the average age of the participants was 20 years) was paired with
a confederate of the experimenter and watched him play a game of
roulette. Physiological responses associated with emotional arousal

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 136 The roots of prosocial behavior in children

(changes in galvanic skin response, blood pulse volume, and heart rate)
were recorded as the participants observed his fate. Perception of simi-
larities between oneself and someone else facilitates empathy arousal,
and so half of the subjects were told that the performer had values and
interests similar to their own; the other half were informed that they had
been paired with the performer because they were different from him.
In addition, half of the "similar" group and half of the "different" group
were informed that the performer received a reward (money) when the
roulette ball landed on an even number and a shock (punishment)
whenever it landed on an odd number. The other half of the participants
believed that they were observing the performer playing a game without
receiving any rewards or punishments. When the performer was re-
warded or punished, the psychophysiological reactions of those who
believed they were like him showed that they empathized with him.
After the first series of trials was completed, additional trials were
anounced. The participants were informed that for these trials they
would regulate both the amount of money the performer won and the
amount of shock he received. Furthermore, they themselves would win
money or receive shocks on these extra trials, the amounts depending
on what the performer received: The more favorable the outcome for
the performer, the less favorable the outcome for the participant. If the
performer received less money, the participant would receive more, and
if the former received less shock, the latter was given more. In short, the
participants had to choose between benefiting themselves at a cost to the
performer and helping him at a cost to themselves. The results were
clear-cut: The subjects who believed themselves to be similar to the per-
former (i.e., the group that was most empathic with the performer) were
highly altruistic toward him, most willing to help him even though this
entailed self-sacrifice.
Empathy and prosocial behavior also are correlated in real-life settings.
Rescuers of Jews from Nazis during World War II were interviewed
regarding their motives for their actions. More than half reported that
they helped primarily because they felt empathy and sympathy for the
victims they rescued (Oliner & Oliner, 1988).
The studies involving children have yielded somewhat less consistent
findings than those involving adults, but nevertheless indicate that
empathy is related to prosocial behavior. Some measures of empathy
are associated with children's prosocial behavior, whereas others are not

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 Emotional factors in prosocial behavior 137

(Eisenberg & Miller, 1987). Often, empathy has been assessed by tests
in which children are exposed to stories and/or pictures containing infor-
mation about another's affective state or situation (e.g., a child who lost
his dog). Then they are asked to report, either verbally or by means of
nonverbal responses (i.e., pointing to pictures of facial expressions),
what they themselves are feeling. Children are scored as empathizing if
they report experiencing the emotion the story protagonist would be
expected to feel (Eisenberg & Lennon, 1983; Feshbach & Roe, 1968;
Hoffman, 1982).
Such indices of empathy have not been positively related to prosocial
behavior (Eisenberg & Miller, 1987), perhaps because these indices are
not valid measures of emotional empathy. The stories used to evoke
emotion may be too short to elicit an emotional response, and, in ad-
dition, children may try to give the expected or desired response even
if they themselves feel nothing when they hear the stories (Eisenberg
& Lennon, 1983; Hoffman, 1982). Also, children may not be able to
shift their emotions as quickly as they are shifted from one scenario to
another.
When children's observed emotional responses are the measures of
empathy, the pattern of relations between measures of empathy and
prosocial behavior is quite different. For example, 1-year-olds who ex-
hibited sadness on their faces in response to viewing a sad adult were
more likely to make active attempts to comfort other persons they wit-
nessed crying by hugging, touching, or caressing them (according to
parental report) (Weston & Main, 1980). Moreover, young children
who exhibit emotional responsiveness to another's distress frequently
try to comfort the distressed person (Radke-Yarrow & Zahn-Waxler,
1984; Radke-Yarrow et al., 1983).
In one study, preschoolers viewed twofilmson two occasions, 1 week
apart (Lennon, Eisenberg, & Carroll, 1986). In each film, a boy and a
girl who were playing on local playgrounds fell and hurt themselves. The
actors in the two films were different pairs of children, and the children
were told that the events in the film were real. The preschoolers' facial
and gestural reactions to thefilmswere videotaped and coded for affects
such as sadness or distress. The preschoolers also were given an oppor-
tunity to assist the injured children in one film by helping to make a
game for them to play in the hospital, but at a cost to themselves (the
preschoolers could either help or play with attractive toys). Moreover,

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 138 The roots of prosocial behavior in children

on another day, the children were given the opportunity to donate at-
tractive stickers to needy children whom they did not know.
Among both boys and girls, facial and gestural indicators of empathy
(reactions such as sadness and distress) were associated with prosocial
behavior. Boys who exhibited more empathy were especially likely to
assist the hospitalized children; empathic girls donated more than did
less empathic girls. In a similar study (Eisenberg, McCreath, & Ahn,
1988), facial reactions indicating sadness or concern for injured others
were related to preschoolers' spontaneous sharing and helping in play
sessions with a peer. In addition, in a study in which second- and fifth-
graders viewed a film of injured, hospitalized children, there were some
positive correlations between facial evidence of sympathetic concern
and willingness to assist the children, as well as some negative correla-
tions between facial personal distress (anxiety in reaction to viewing the
films) and prosocial actions (Eisenberg, Fabes, Miller, Fultz, Shell,
Mathy, & Reno, in press). Thus, empathy and sympathy apparently
mediate at least some instances of prosocial responding.
Not all prosocial behavior is motivated by empathic or sympathetic
reactions. People may assist others for nonaltruistic motives, for exam-
ple, to attain rewards or social approval, or because of their desire to
adhere to internalized moral values (Bar-Tal, 1982; Eisenberg, 1986). In
some circumstances in which people could help or share with others,
there is no reason to feel empathy or sympathy because the potential
recipient of aid does not feel distressed, or there are no cues regarding
the recipient's emotional state. In such situations, people frequently
may assist for egoistic reasons (Schoenrade, Batson, Brandt, & Loud,
1986); for example, people often assist friends, even when they are not
distressed, because of the desire to maintain the friendship.

Guilt

Guilt refers to a "bad feeling one has about oneself because one is aware
of harming someone" (Hoffman, 1984, p. 289). Such feelings may also
mediate prosocial behavior (Hoffman, 1984). Children may first experi-
ence empathic distress when they observe someone who is physically
injured, and this empathy may be transformed to guilt if they believe
themselves responsible for the other's pain. Once children are aware
that others can have internal states, empathic reactions to another who

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010
 Emotional factors in prosocial behavior 139

is in pain, is unhappy, or is experiencing other difficulties also may be

transformed to guilt. Finally, when children develop the awareness that
other people have an identity and continuing existence beyond the im-
mediate situation, empathic responding to the other's general plight can
be transformed into guilt if the empathizer feels responsible for the
other's plight, or if the empathizer feels that he or she is in a relatively
advantaged situation (Hoffman, 1982, 1984).
Although signs of conscience or guilt have been observed in 1- to 2-
year-olds, there has been relatively little research on the role of guilt in
prosocial behavior. Some children not only try to alleviate the distress
they cause others but also exhibit guilty facial expressions or apologetic
reactions. Moreover, children who exhibit signs of guilt are more likely
than other children to assist when they observe another's distress, even
though they did not cause it (Zahn-Waxler, Radke-Yarrow, & King,
1983).
When resolving hypothetical moral dilemmas, children and ado-
lescents sometimes report that guilt reactions are important determinants
of prosocial behavior (Eisenberg-Berg, 1979; Karylowski, 1982a,b;
Thompson & Hoffman, 1980); and young children sometimes attribute
guilt to hypothetical story characters who hurt others (Chapman et al.,
1987). In experimental laboratory studies, adults who accidentally have
harmed others (e.g., have broken a valued object or dropped a box of
computer cards) are likely to provide assistance to the person they have
harmed (e.g., Freedman, 1970; Freedman, Wallington, & Bless, 1967;
Wallace & Sadalla, 1966). Such subsequent helping may represent an
attempt to make reparation because of feelings of guilt.
It is likely not only that guilt reactions motivate subsequent helping
but also that the anticipation of guilt motivates altruistic behavior
(Bandura, 1977). This issue is difficult to investigate because guilt, like
empathy, is an internal response. However, with the use of multimethod
approaches to assess these variables, including facial expressions and
physiological indices that are likely to give better assessments of internal
emotional reactions, our understanding of the dynamics of empathy and
guilt will increase (Eisenberg, Fabes, Bustamante, & Mathy, 1987;
Eisenberg, Fabes, Bustamante, Mathy, Miller, & Lindholm, 1988).

Downloaded from https://www.cambridge.org/core. University of Florida, on 09 Nov 2017 at 10:05:27, subject to the Cambridge Core terms
Cambridge Books Online
of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511571121.010