Professional Documents
Culture Documents
and Morality
Heidi L. Maibom
Title Pages
(p.iv)
Oxford University Press is a department of the University
of Oxford.
It furthers the Universitys objective of excellence in
research, scholarship,
and education by publishing worldwide.
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France
Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University
Press
in the UK and certain other countries.
Published in the United States of America by
Oxford University Press
198 Madison Avenue, New York, NY 10016
Oxford University Press 2014
All rights reserved. No part of this publication may be
reproduced,
stored in a retrieval system, or transmitted, in any form or
by any means,
without the prior permission in writing of Oxford
University Press,
or as expressly permitted by law, by license, or under terms
agreed with the
appropriate reproduction rights organization. Inquiries
concerning reproduction
outside the scope of the above should be sent to the Rights
Department,
Oxford University Press, at the address above.
You must not circulate this work in any other form
and you must impose this same condition on any acquirer.
CIP data is on file at the LOC
9780199969470
987654321
Printed in the United States of America on acid-free paper
Contents
Title Pages
List of Contributors
9 On Empathy
R. Peter Hobson and Jessica A. Hobson
2. Routes to Empathy
So far I have considered (affective) empathy as a reaction to either the
distressing situation or the distressed subject regardless of how the
situation or the (p.10) subject is represented. But the way in which the
person accesses the others situation or emotional state is hardly
irrelevant to the subsequent empathic reaction. For instance, evidence
from neuroscience indicates that when we imagine someone
experiencing an emotion, we activate fewer brain networks that overlap
with the activation when we experience that emotion for ourselves, than
when we observe someone experience that emotion (Lamm, Decety, &
Singer 2011).
There are 3 main routes to affective empathy: witnessing the person in
the situation (perceptual route), believing that the person is in a certain
situation or is experiencing a certain emotion (inferential route), or
imaginatively engaging with her point of view (imaginative route)
(Maibom 2007). The latter is often known as perspective taking.
Hoffman talks of more routes, but from a conceptual standpoint, these
three are the most important ones. The first route is often thought to be
the most basic one. It is the oldest phylogenetically and ontogenetically,
and the one most widely shared with nonhuman animals according to
Preston and de Waal. The perception-action model of empathy is the one
most compatible with this route. Preston and de Waal say that: (2002,
4):
A Perception-Action Model of empathy specifically states that
attended perception of the objects state automatically activates the
subjects representations of the state, situation, and object, and that
activation of these representations automatically primes or
generates the associated autonomic and somatic responses, unless
inhibited.
Alvin Goldman has something similar in mind when he talks of low-
level mindreading, and some neuroscientists conceptualize this process
as subserved by mirror neurons (e.g., Iacoboni 2008). What is essential
to this route to empathy is that the empathizer has perceptual access to
information about the subjects welfare or emotional state. This form of
empathic relating to others has an immediacy to it and is not well
understood as essentially involving complex inferences. Instead, it is
thought that automatic imitation with afferent feedback is responsible
for this type of empathy or that mirror neurons are directly responsible
(e.g., Goldman 2006).
The role of mirror neurons in understanding other minds is heavily
debated. Mirror neurons are neurons that fire both when the agent
performs an action, such as grasping, and when she sees another person
performing the same action, that is, grasping. It might therefore be
thought that mirror neurons contribute to seeing actions as being
actions, not just meaningless gestures, and actions of particular kinds.
Action individuation, however, recruits areas of the brain that do not
contain mirror neurons as such, but (p.11) whose activation are
probabilistically related to the firing of such neurons (Umilit et al.
2001). As a result, people often talk of the mirror neuron system as
including not only mirror neurons but also neurons whose firing is
probabilistically related to that of mirror neurons. More expansively,
people talk of mirroring processes if the same brain areas are activated
when observing someone experiencing an emotion and when
experiencing the emotion oneself (e.g. Iacoboni 2008).4
Some argue that mirroring gives us a direct experiential grasp of the
minds of others (Gallese, Keysers, & Rizzolatti 2004), others that it
causes certain types of mindreading (Goldman 2009), and yet others
that the best mirroring can possibly do is create representations of an
agents motor-intentions but not her intentions for performing the
action (Jacob 2008). Goldman thinks mirroring contributes to ascribing
emotions to others (cognitive empathy). Others, such as Iacoboni
(2008), believe mirroring plays an essential role in truly understanding
how others feel. As we have seen, a number of affective empathic
emotionsfor instance pain, disgust, fear, anger, and sadnessengage
areas of the brain similar to those that are activated when the subject
experiences these emotion nonempathically, that is, directly or
personally. It is important to keep in mind, however, if human mirror
neurons are localized mainly in similar parts of the brain to those of the
macaque, namely, those concerned with visual processing, then they are
unlikely to play a significant role in most empathic reactions. So-called
mirroring does not imply the operation of mirror neurons.
The second route to empathy is through belief or knowledge. The
prototype of such situations are ones in which we have no direct
perceptual access to what the person is feeling or their situation. This
knowledge might be acquired by the testimony of others or by simple
inference. If we know Bill has lost his job, we infer that he is upset about
it. It is quite plausible that this route is less potent than the other two.
That is, it may be less likely to lead people to experience empathy or it
may invoke less strong a reaction. That is not to say that knowing that
someone is in pain, say, does not typically make people feel bad for that
person. In perspective-taking studies of empathy, the group that is given
no instructions to relate one way or another to the story of a person in
need still experiences empathy or sympathy for that person (Batson
2011).
(p.12) The third route is what is sometimes known as perspective
taking. Perspective taking involves the imagination. When we take
someones perspective, we imagine what it is like for her or what we
would feel were we in her situation. This is also known as simulation or
high-level mindreading. In Goldmans offline simulation account, we
imagine being in someone elses situationadjusting for differences in
beliefs, desires, and so on when neededsee what we would think,
feel, want, or intend to do, then ascribe those psychological states to the
person we are trying to understand (Goldman 1989). The seeing part of
this operation is the interesting one, and is central to offline accounts of
simulation. The idea is that our psychological system is a machine of
sorts that can operate on actual stimuli or on pretend stimuli. Our
imagination feeds this system with pretend stimuli, and then the system
processes these stimuli as it would actual stimuli and produces the same
results, only in pretend mode, as it would were it operating on actual
stimuli. For instance, if I imagine missing my flight by 5 minutes, I
imagine being more annoyed than if I imagine missing it by 20 minutes,
just as I would be in real life. Other accounts of simulation dont have
the same interpretation of decision making as an almost automated
system and rely more on the idea of effortful reasoning. In all cases,
however, the idea is that we use the same system to make decisions
ourselves and to predict others decisions, thoughts, and so on.
There are reasons to be hesitant about offline accounts. The first is the
concern that this type of simulation is merely a glorified form of
projection. We see the other as a fragment, or reduplication, of
ourselves (per Lipps1903a). The often voiced concerns that we are bad at
adjusting our psychological interior so as to be able to produce accurate
simulations are extensions of this consideration. There is a fair amount
of evidence that we can, in fact, empathize with and understand others
going through experiences that we have not gone through, and who are
significantly different from us (Buccino et al. 2004; Lamm, Meltzoff, &
Decety 2009). Some people take this to be reason to reject the idea that
mirror neurons are responsible for us understanding others intentions
(Goldman 2006). The second worry concerns our ability to engage our
psychological system by means of the imagination to create the desired
results. There is much evidence that we are not very good at forecasting
our reactions to counterfactual scenarios (Maibom forthcoming). This is
particularly true when those reactions involve, or are strongly influenced
by, visceral and affective states. We underestimate the effects of hunger
and thirst on our preferences (Read & van Leeuwen 1998, Van Boven &
Loewenstein 2003), overestimate our ability to withstand pain (e.g.,
Christensen-Szalanski 1984), ignore the effects of embarrassment on
our actions (Van Boven, Loewenstein, Welch, & Dunning 2012), and
generally imagine being more morally (p.13) upright (Latan & Darley
1970, Milgram 1963) and braver than we actually are (Woodzicka &
LaFrance 2001).5 This suggests that the imagination cannot recreate the
situation in the sort of detail that is required for us to have similar
reactions to such representations as we do to the perceptual
representations of the situation. Most likely, the imagination works with
more abstract, naked representations of events, which are highly useful
for a variety of purposes. But it cannot recreate the situation in such a
way as to simply recreate in us reactions that we would have to actually
being in the situation. Our perceptions trigger implicit memories about
such situations and create visceral and affective reactions, which are
hard to predict on the basis of more naked and sparse cognitive
representations of the situation. This suggests that there is an important
difference between empathizing in the situation and empathizing in the
imagination (Maibom forthcoming).
It is in general a mistake to think of empathic reactions as being largely
rote and the result of an automatic process. There is a rapidly expanding
literature on empathy modulation by cognitions, in part from the social
neurosciences. Jean Decety, Tania Singer, and Claus Lamm all insist that
empathy is never automatic, but always modulated by the relationship
between the empathizer and the subject, what the empathizer knows
about the subject, and so on (Decety 2011, Singer & Lamm 2009).
Empathy is affected by attention, attitude toward the subject, knowledge
about the experiences of the other, and so on. For instance, if a person is
asked to count the fingers of a subject who experiences painful stimuli
to the hand, there is no activation of the insular and cingulate cortices,
which are typically activated during empathic pain (Gu & Han 2007).
And we already know from Stotland and Batsons work that asking a
subject to pay attention to factors other than the persons feelings often
reduces empathy (Stotland 1969, Batson 1991). Another interesting thing
to note is that Batsons imagine-self versus imagine-other instruction
set was used with subjects in some experiments and differential
activation was found. Imagine-self instructions activated more areas
with overlap with the emotion felt by the subject than the imagine-other
did. The latter activated more theory of mind related areas (Lamm,
Batson, & Decety 2007).
Empathy depends heavily on the relationship between the empathizer
and the person empathized with. This is unlikely to come as a shock to
anybody. Hume, himself, maintained that we feel more for those close to
us in affection, time, and place (Hume 1739/1978). But there is now
relatively good experimental support of the idea. Friends are empathized
with more than strangers (Meyer et al. 2013), ingroup members more
than outgroup members (Tarrant, Dazeley, & Cottom 2009; Gutsel &
Inslicht 2010), and (p.14) those who are deemed to be fair are
empathized with more than self-serving, unfair persons (Singer et al.
2006).
3. Individual Differences
Just as empathy depends on the situation of the person, her
relationships to the other, her attention, and so on, there are significant
interpersonal differences in empathy. Some mental disorders are
characterized by deficient empathy, such as autism, psychopathy, and
narcissism, but even within the population at large, there are significant
differences in the extent to which people experience empathy. Nancy
Eisenberg, for instance, has documented empathy differences,
particularly in children. Some children are more prone to experience
empathy than others, who are more prone to experience personal
distress (Eisenberg et al. 1988, Eisenberg & Fabes 1990, Eisenberg &
Spinrad this volume). These differences correlate with differences in
prosocial behavior, social adjustment, popularity, and so on.
Psychopaths are famous for lacking empathy. It is one of the diagnostic
criteria of The Psychopathy Checklist-Revised (Hare 2004).6 Criminal
psychopaths display a truly astounding disregard for the plight of their
victims, suggestive of impaired empathy (Hare 1993). Furthermore, their
justifications for why harmful actions are wrong involve welfare
justifications far below the norm (Blair 1995). A popular explanation of
their moral deficit is that it is due to their deficient (affective) empathy
or sympathy (Blair 1995, Nichols 2004).7
Physiological evidence has typically been taken to support the idea that
psychopaths lack empathy. Christopher Patrick and colleagues found
that psychopaths have reduced startle reflex to pictures of others in
distress compared to controls (Patrick, Cuthbert, & Lang, 1994). Startle
reflex, however, is a broad indicator of fear or anxiety more generally,
and thus is insufficient to show reduced empathy (and cannot, by its
nature, distinguish between empathy and personal distress). Patrick and
colleagues also found that psychopaths show the same abnormal startle
reflex in response to directly threatening images (Levenston, Patrick,
Bradley, & Lang 2000), which suggests that the deficit represents a
generalized disorder in the initiation of defensive action from an
orienting response. So psychopaths impaired negative reactions to
others distress may be due to the fact that they experience less fear,
anxiety, (p.15) or distress in response to seeing others in distress than
nonpsychopathic individuals. James Blair and colleagues found that
psychopaths have reduced palmar sweating to pictures of people in
distress compared to nonpsychopaths (Blair, Jones, Clark, & Smith
1997). But since skin conductance tests measure arousal only and not,
for instance, valence, the results can be interpreted in terms of deficient
empathy, sympathy, emotional contagion, compassion, pity, personal
distress, fear, shock, or stress. There is some reason to think that skin
conductance, like the startle reflex, captures personal distress and
anxiety/shock/stress more than empathy, sympathy, or their
compassionate cousins (e.g., Eisenberg, Fabes, Schaller, Miller, Carlo et
al. 1991). Many of the pictures used in skin conductance tests are
dramatic, picturing extreme distress (weeping, grieving) or scenes of
death and mutilation. The more distressed someone is, the more likely
witnesses are to feel personal distress, to the point where personal
distress is experienced more than empathic concern (Eisenberg et al.
1988, Hoffman 2000, Figley 2002).
By contrast to physiological tests, some more recent studies do not find
the expected negative correlation between affective empathy and
psychopathy. Although they did find that offenders were less empathic,
both cognitively and affectively, Gregor Domes and colleagues (2013)
found no difference between psychopathic and nonpsychopathic
inmates. David Lishner and colleagues report that people scoring high
on psychopathy do not differ from people scoring low in psychopathy in
emotional contagion or sympathy (empathic concern) (Lishner et al.
2012). If anything, high psychopathy scorers were more sympathetic
toward those in need. Another study with a college sample found that
people scoring high on coldheartedness were able to show perspective-
taking ability and show empathic concern as well as, or perhaps even
better than people who score low of PPI-SF-I8(Mullins-Nelson, Salekin,
& Leistico 2006, 139140). There are also important sex differences,
with female psychopaths appearing to have fewer affective empathy
deficits (Sutton, Vitale, & Newman 2002),9 even in studies that show
such empathy deficits in (p.16) children and adolescents with
psychopathic tendencies (Dadds et al. 2009, Isen, Raine, Baker, Dawson,
& Bezdjian 2010).10 Jean Decety, Laurie Skelly, and Kent Kiehl have
found mixed results when testing psychopaths on empathy measures
(Decety, Skelly, & Kiehl 2009). Though they did find decreased
activation in some brain areas in psychopaths compared to
nonpsychopaths (pariacqueductal gray, vmPFC, and OFC), they found
high activation in the anterior insular cortex (AIC), an area that has
most consistently been associated with emotional empathy. They
interpret the results as supporting the idea that AIC is more involved in
the cognitive aspect of empathy. However, Yawei Cheng, An-Yi Hung,
and Jean Decety (2012) show that adolescents high in callous-
unemotional traits process empathy-arousing stimuli abnormally,
particularly in the early stages, but that they nevertheless have intact
sensorimotor resonance, supporting the idea that that part of the
affective response to others in distress is intact in psychopaths.
Though psychopaths are assumed to have intact cognitive empathy,
some studies report distinct impairments, and a recent meta-analysis
shows very small but significant impairments in emotion recognition
(Wilson, Juodis & Porter 2011). Blair and colleagues found impairment
in recognition of facial expressions of fear and sadness in children with
psychopathic tendencies, but this finding has not been replicated with
adult psychopaths (Blair, Jones, Clark, & Smith 1997). Adult psychopaths
have difficulties only with the recognition of fear (Blair 2005, Iria &
Barbosa 2009, Marsh this volume). Some studies show sex offenders
also have deficits in the recognition of the facial expression of fear; they
often mistake it for surprise (Hudson et al. 1993). Empathy with fear
may be impaired in psychopaths due to their own fear deficit. Since they
experience diminished fear compared to nonpsychopaths, they often do
not experience much of a physiological reaction to imagining fearful
situations and their pain threshold is high, one would expect that their
understanding of, and ability to feel with and for people who are afraid,
would also be impaired. This certainly fits with the various accounts that
link perception of an emotion with the capacity to experience it
(Goldman & Sripada 2005, Preston & de Waal 2002), and is an option
explored further by Abigail Marsh (this volume). Of course, lack of fear
may itself cause a number of the deficits associated with psychopathy,
including the moral ones. It might, for instance, lead to an abnormal
pattern of decision-making.
(1) . I am going to ignore the origin of the term empathy here. I dont
think it helps much with the current debate. Suffice it to say that the
term was coined by Edward Titchener (1909) as a translation of the
German Einfhlung to denote an emotion that Theodore Lipps (1903)
famously associated with understanding works of art. For an excellent
historical overview of the uses of the term, see Karsten Stueber (2006).
(2) . This is one of Batons most used scenarios. Katie Banks is a senior
at the university (where the study is being conducted). Her parents and
sister have recently been killed in a car crash leaving her as the sole
caretaker of her surviving siblings: a brother and a sister. However, she
is struggling to finish her final year at university while taking care of two
children. She is very concerned that if she does not complete her degree,
she will be unable to find a job that pays enough for her to be able to
continue taking care of her siblings and that she will therefore have to
put them up for adoption. Katie needs someone to help her with her
studies to get her through the year (Batson, Early, & Salvarini 1997).
(3) . There are other findings of overlap, but in phenomena that may less
straightforwardly be called empathic emotions, such as in touch
(Blakemore, Bristow, Bird, Frith, & Ward. 2005, Keysers et al. 2004),
reward (Mobbs et al. 2009), and social exclusion (Masten, Morelli, &
Eisenberger 2011). See Boris Bernhardt and Tania Singer (2012) for an
authoritative review.
(4) . Goldman provides the following very helpful definition of
mirroring, as it is most commonly used: A neural process or event E is a
case of mirroring just in case E is the activation in an observer of a
neuron or neural system that (1) results from observing a targets
behavior or behavioral expression and (2) would, in a normalcase of
such behavior, match or replicate an activation in the target of a
corresponding neuron or neural system that the observed behavior
would manifest (Goldman 2009, 236).
(5) . For a fuller list of failures to forecast, see Maibom forthcoming.
(6) . One of the most used diagnostic tools of psychopathy.
(7) . Technically speaking, Shaun Nichols argues that psychopaths have
an impaired Concern Mechanism. He chooses that term to avoid
deciding whether empathy or sympathy or both play the role in
sociomoral regulation that he is concerned with.
(8) . The Psychopathic Personality Inventory-Short Form (PPI-SF) is
the short version of the Psychopathic Personality Inventory (PPI). Both
are self-report measures meant to be used with non-forensic
populations (Lilienfeld & Andrews 1996, Lilienfeld 2004). The PPI-SF
has two subcomponents. PPI-SF-I factor of psychopathy highlights
Stress Immunity, Social Potency, Fearlessness, and Coldheartedness,
and PPI-SF-II focuses on Impulsive Nonconformity, Blame
Externalization, Machiavellian Egocentricity, and Carefree
Nonplanfulness. This factor structure is meant to recapitulate Factor 1
vs. Factor 2 classification of thePsychopathy Checklist-Revised (PCL-R),
or that between primary vs. secondary psychopathy.
(9) . Steven Sutton, Jennifer Vitale, and Joseph Newman found that
highly anxious female psychopaths do not exhibit abnormal startle
reflex, suggesting that they have an intact stress response to distress in
others (Sutton, Vitale, & Newman 2002).
(10) . Dadds and colleagues use a measure of empathy, the Griffith
Empathy Measure, which does not differentiate between emotional
reactivity, contagion, personal distress, or empathy. Pajer, Leininger, and
Gardner (2010) report that girls with conduct disorder (correlated with
later diagnosis of Antisocial Personality Disorder) did not show
difficulties in identifying facial affect in others.
(11) . I use the term broadly to capture both high-performing people with
autism and people diagnosed with Aspergers Syndrome.
(12) . These results are mostly based on self-reports using the
Interpersonal Reactivity Index, Questionnaire Measure of Emotional
Empathy, or Hogans Empathy Scale.
(13) . However, since he was inspired by Hume and Smith, it is likely
that he really had empathy in mind.
(14) . Hrdys fascinating reconstruction of parental behavior has it that
human infants typically spend long periods of time with caregivers that
are not their parents. This puts pressure on the child to understand
others and how to affect them in such a way as to receive what she needs
or wants.
(15) . For more extensive overviews of empathic development, see
Eisenberg 2000; Eisenberg, Spinrad, and Sadovsky 2006; Spinrad and
Eisenberg this volume; Hoffman 2000.
(16) . Researchers often talk of primary and secondary psychopaths.
Primary psychopaths are the ones who have more pronounced
emotional-interpersonal deficits. Secondary psychopaths are high on
Factor 2 of PCL-R, that is, behavioral and lifestyle issues. But some
studies suggest that the latter are more violent (Hicks, Markon, Patrick,
& Krueger 2004). This suggests that violence correlates more with
emotional dysregulation than with absence of vicarious emotional
responses, in line with the importance someone like Eisenberg puts on
regulation (Spinrad & Eisenberg this volume).
(17) . Notice that in Batsons (1991) Katie Banks scenario, Katies
misfortune is not the result of anyones wrongful actions. So by relying
on this literature Nichols succeeds in basing his theory on quite solid
empirical grounds, but he also changes the topic in such a way that it is
no longer clear how he addresses the concern of central interest to both
him and Blair: judgments of harmful actions.
(18) . If, indeed, it is the oomph provided by the emotions that back
moral norms that lend them their categoricalness, then one might
wonder whether the oomph of a truly aversive emotional response like
empathic or personal distress might not lend a more powerful oomph to
harm norms than the relatively weaker and calmer sympathy
(empathetic concern) that is the operative emotion in most of the
studies Nichols relies on. This suspicion is reinforced by considering
that the exemplar emotion in his story of the genealogy of norms is
disgustalso a quite powerful aversive emotion.
(19) . He calls it empirical internalism about core moral judgment (cf.
Nichols 2004, 109115).
(20) . Mild stimulation of the module might not be sufficient to produce
the emotion itself, but merely the relevant intuition, for example, that
inflicting the relevant harm is wrong.
(21) . This is a slight rephrasing of Socrates question in 10a: The point
which I should first wish to understand is whether the pious or holy is
beloved by the gods because it is holy, or holy because it is beloved of
the gods.
(22) . That is to say, a number of people take the norm of loyalty to be as
much a moral norm as the norm that one should not hurt others
gratuitously. Whether we should care about purity, authority, and group
norms I leave to the side here, as do most currently working in moral
psychology.
Empathy-Induced Altruism and Morality
No Necessary Connection
C. Daniel Batson
The empathy-altruism hypothesis states that empathic concern
produces altruistic motivation (Batson, 1987,2011). To understand this
deceptively simple hypothesis, it is necessary to be clear about what is
meant by both empathic concern and altruistic motivation.
Empathic Concern
In the hypothesis, empathic concern refers to other-oriented emotion
elicited by and congruent with the perceived welfare of a person in need.
This other-oriented emotion has been named as a sourceif not
thesourceof altruism by Thomas Aquinas, David Hume, Adam Smith,
Charles Darwin, Herbert Spencer, and William McDougall, as well as by
several contemporary psychologists.
Four points may help clarify what this emotional state involves. First,
congruent refers not to the specific content of the emotion but to its
valence: (p.42) positive when the perceived welfare of the other is
positive, negative when the perceived welfare is negative. For example, it
would be congruent to feel sad or sorry for someone who is upset and
afraidor, like the Good Samaritan (Luke 10:33), to feel compassion for
the unconscious victim of a mugging. Second, although the term
empathy is broad enough to include feeling empathic joy at anothers
good fortune (Smith, Keating, & Stotland, 1989; Stotland, 1969), not all
empathic emotion is hypothesized to produce altruistic motivation, only
empathic concern felt when another is perceived to be in need. Without
need, there is no impetus for change.
Third, empathic concern is not a single, discrete emotion but includes a
whole constellation. It includes feelings of sympathy, compassion,
softheartedness, tenderness, sorrow, sadness, upset, distress, concern,
and grief. Fourth, empathic concern is other-oriented in the sense that it
involves feeling for the otherfeeling sympathy for, compassion for,
sorry for, distressed for, concerned for, and so on. Although feelings of
sympathy and compassion are inherently other-oriented, we can feel
sorrow, distress, or concern that is not oriented toward someone else, as
when something bad happens directly to us. Both other-oriented and
self-oriented versions of these emotions may be described as feeling
sorry or sad, upset or distressed, concerned or grieved. This breadth of
usage invites confusion. The relevant psychological distinction lies not
in what emotional label is used but in whose welfare is the focus of the
emotion. Is one feeling sad, distressed, concerned for the other, or
feeling this way as a result of what has befallen oneselfincluding the
experience of seeing another suffer?
In recent years, the term empathy has been applied to a range of
phenomena other than the feeling-for described above (see Batson,
2009, for an overview). It has been used by different researchers to
mean:
Knowing anothers thoughts and feelings.
Adopting the posture or matching the neural response of another.
Coming to feel as another feels.
Feeling distress at witnessing anothers suffering.
Imagining how one would think and feel in anothers place.
Imagining how another thinks and feels.
A general disposition (trait) to feel for others.
Each of these phenomena is distinct from other-oriented empathic
concern. The empathy-altruism hypothesis makes no claim that any of
these other phenomena produces altruistic motivation, except if and
when that phenomenon evokes empathic concern.
Orchestrating Motives
Empathy-induced altruism and moral motivation are distinct goal-
directed motives. Each has strengths, but each also has weaknesses. The
greatest good may come from strategies that orchestrate these motives
so that the strengths of one can overcome the weaknesses of the other.
Think once again about principles of fairness or of the greatest good.
These principles are universal and impartial, but motivation to uphold
them seems corruptiblevulnerable to oversight, rationalization, and
self-deception. Empathy-induced altruism is a potentially powerful
motive with a strong emotional base. But it is limited (p.56) in scope,
producing special concern for particular persons. Perhaps if people can
be led to feel empathy for the victims of injustice, or for those with
special needs, this will bring together the unique strengths of these two
motives. Desire to uphold standards, principles, and ideals may provide
perspective and reason; empathy-induced altruism may provide
emotional fire and a force directed specifically toward seeing the victims
suffering enda want to accompany the moral ought. The
combination may discourage oversight and rationalization (see
Solomon, 1990).
This orchestration of empathy-induced altruism and moral motivation
may seem to echo Martin Hoffmans idea of empathy-based morality,
especially when he says:
My hypothesis is that abstract moral principles, learned in cool
didactic contexts (lectures, sermons), lack motive force. Empathys
contribution to moral principles is to transform them into prosocial
hot cognitionscognitive representations charged with empathic
affect, thus giving them motive force. (Hoffman, 2000, p. 239)
In spite of sharing Hoffmans hypothesis, I believe that our views are
importantly different. He speaks of embedding empathy in morality
and of bonding empathic affect to a moral principle. Such language
suggests that empathy becomes inextricably linked to morality, that
rather than independent motives that may cooperate or conflict,
empathy-induced altruism and morality necessarily work in harmony.
The research reviewed above contradicts this optimistic assumption.
Although empathy-induced altruism can lead a person to act in a way
judged moral, it can also lead the person to violate his or her own moral
standards. Altruism and morality have no necessary connection. The
challenge is to orchestrate altruistic and moral motives so they
complement one another.
Examples
Let me close with a few concrete examples of orchestration of empathy-
induced altruism and moral motivation. A careful look at data collected
by Samuel and Pearl Oliner and their colleagues (Oliner & Oliner, 1988)
suggests that such orchestration occurred in the lives of a number of
rescuers of Jews in Nazi Europe. Involvement in rescue activity
frequently began with concern for a specific individual or individuals for
whom compassion was feltoften individuals known previously. This
initial involvement subsequently led to further contacts and rescue
activity, and to a concern to do right that extended (p.57) well beyond
the bounds of the initial empathic concern. In several cases, most
notably in the French village of Le Chambon, the result was dramatic
indeed.
Such orchestration also seems to have occurred at the time of the bus
boycott in Birmingham, Alabama, in the 1950s. The horrific sight on TV
news of a small Black child being rolled down the street by water from a
fire hose under the direction of local policeand the emotions this sight
evokedseemed to do more to arouse a concern for racial equality and
civil rights than hours of reasoned moral suasion.
In these two examples, orchestration was not planned; it occurred as a
result of unfolding events. At other times, the orchestra has a human
conductor. Intentionally creating confrontations designed to induce
empathic concern seems to lie at the heart of the nonviolent protest in
the face of entrenched oppression practiced by Mahatma Gandhi and by
Martin Luther King, Jr.
Such orchestration can also be found in the writing of Jonathan Kozol.
Deeply troubled by the savage inequalities in public education between
rich and poor communities in the United States, Kozol (1991) clearly
documents disparities, pointing out the injustice. But he does more. He
takes us into the lives of individual children. We come to care about
their welfare and, as a result, to care about setting things right. Kozols
goal is not simply to get us to feel; he wants to get us involved in action
to improve funding for schools in poor communities. He pursues this
goal by orchestrating empathy-induced altruism and moral motivation.
Harriet Beecher Stowe (1852/1929) used much the same orchestration
strategy to galvanize opposition to slavery inUncle Toms Cabin.
However difficult it may be in practice, coordinating altruism and moral
motivation by inducing empathy for victims of immorality is
theoretically straightforward. Yet this is not the only possible way to
combine these two motives. The story of wise King Solomon presents a
far more subtle example of the use of empathy-induced altruismand
the partiality it inducesin the service of doing right. Recall that two
women came before Solomon. One claimed that when the others infant
son died, the bereft mother switched her dead son for the first womans
live one. The other woman claimed that the dead son was the first
womans and the live son hers.
So the king [Solomon] said, Bring me a sword, and they brought a
sword before the king. The king said, Divide the living boy in two;
then give half to the one, and half to the other. But the woman
whose son was alive said to the kingbecause compassion for her
son burned within herPlease, my lord, give her the living boy;
certainly do not kill him! The other said, It shall be neither mine
nor yours; divide it. Then the king responded: Give the first
woman the living boy; do not kill him. She is his mother. (1 Kings
3:2427 NRSV)
(p.58) Thus did Solomon execute justice (1 Kings 3:28). It is hard to
imagine a more successful orchestration of empathy-induced altruism
and moral motivation.
Orchestrating these motives is a promising strategy for promoting action
on behalf of both those suffering immoral treatment and society at large.
It appears capable of producing dramatic results. Yet it is rarely even
considered. The assumption that empathy-induced altruism is
necessarily moral has prevented us from recognizing the importance of
such a strategy. With this assumption no longer tenable, new challenges
and possibilitiesarise.
Empathy and Morality
A Developmental Psychology Perspective
Tracy L. Spinrad
Nancy Eisenberg
The constructs of empathy and sympathy have been considered
important emotional aspects of morality, particularly with regard to how
such emotions contribute to moral values and moral behavior (Batson,
1991; Eisenberg, Fabes & Spinrad, 2006; Hoffman, 2000).
Understanding these aspects of morality in children may have important
implications for childrens sense of responsibility, compassion, and later
humanitarian conduct. Thus, it is important to take a developmental
approach to understanding moral behavior in children; understanding
the development, correlates, and origins of such behavior are critical in
order to develop ways to promote such characteristics.
In this chapter, definitional issues with respect to empathy and
empathy-related responding and the links between such responses and
moral behavior are discussed first. Next, the development of empathy
and prosocial behavior is considered. We then briefly cover some
correlates of empathy, sympathy, and prosocial behavior. We conclude
this chapter with a discussion of the origins (p.60) of empathy and
prosocial behavior, focusing on the role of temperamental differences in
emotion and regulation, and we suggest areas for further study.
Conclusions
Research indicates that prosocial behaviors and empathy/sympathy
generally increases with age until mid-childhood or adolescence and that
individual differences are related to temperament. Prosocial behaviors
and empathy-related responding are predictive of childrens social
competence and low problem behaviors. There is still much to explore
with relation to morality from a developmental perspective.
In thinking about the origins of childrens prosocial behavior and
empathy-related responding, researchers rightly assume that such
behaviors are affected by heredity as well as their relationships with
others. Due to space constraints, these issues have not been considered
in this chapter. A future direction for research is to better understand
the genetic contributions to childrens empathy and prosocial
responding (see Knafo & Israel,2010). The ways that parents can (p.70)
encourage caring behaviors in their children also has received
considerable attention (see Eisenberg, Fabes, & Spinrad, 2006);
however, more research focusing on how parenting and/or the
environment moderates the contribution of genetics is needed
(Bakermans-Kranenburg & Van IJzendoorn, 2011; Knafo, Israel &
Ebstein, 2011).
Another area for further study is to focus research on the role of culture
in childrens empathy-related responding and prosocial behavior. It is
important to understand whether the development, associations, and
origins of caring generalize to minority populations. For example, the
relations of parenting and childrens regulatory abilities to Indonesian
childrens prosocial behaviors have been studied, and they appear to be
somewhat similar to those with children in Western societies
(Eisenberg, Liew & Pidada, 2001; 2004). Further, childrens prosocial
behavior and empathy toward members of disadvantaged groups should
be examined. In one recent investigation, European-American children
(ages 513) felt more positive about helping an unfamiliar child from a
racial in-group versus a racial out-group (African-American) as well as a
greater obligation to help a child from the racial in-group (Makariev &
Lagattuta, in press). In addition, interactions between minority and
majority children may have complex effects on childrens prosocial
behavior/sympathy: In Indonesia, 7th graders from a minority group
with at least one friend from the majority culture were more
sympathetic and prosocial (Eisenberg et al., 2009). The factors that are
responsible for childrens empathy toward members of out-groups (e.g.,
socialization strategies) and ways that across-group relationships affect
prosocial development have clear implications for promoting moral
action in our society.
Finally, it is important to understand whether the tendency for
individuals to care and help others may be explained by a moral
personality trait. Very little is known about the role of empathy-related
responding in emergence of a moral self (Kochanska, Koenig, Barry,
Kim, & Yoon, 2010). However, there is evidence of some consistency in
childrens moral motivations and sympathy (see Malti, Gummerum,
Keller, & Buchmann,2009; Malti & Krettenauer, in press). Moreover,
higher levels of childrens and adolescents moral reasoning about
prosocial moral dilemmas, as (i.e., hypothetical dilemmas about whether
to help another at a cost to the self), as well as other-oriented prosocial
moral reasoning, tend to be related to higher levels of empathy and
especially sympathy (Eisenberg, Carlo, Murphy, & Van Court, 1995;
Eisenberg et al., 1987; Eisenberg, Zhou, & Koller, 2001). Future research
should explore this possibility and the ways to foster these traits in
young children.
Empathy-Arousal Modes
There appear to be five empathy-arousing modes. Threemimicry,
conditioning, direct associationare automatic and preverbal. Two
mediated association and perspective-takingare typically voluntary and
involve language and cognition.
(p.73) Mimicry was intuitively understood two and a half centuries ago
by Adam Smith (1759/1976):The mob, when they are gazing at a dancer
on the slack rope, naturally writhe and twist and balance their own
bodies as they see him do (pp. 4,10). It was more precisely defined 150
years later by Lipps (1903b) as an innate, involuntary, isomorphic
response to anothers expression of emotion that occurs in two steps
operating in close sequence: one automatically, rapidly, and without
conscious awareness changes ones facial expression, voice, and posture
in synchrony with slightest changes in the models facial, vocal, postural
expressions of feeling (Dimberg, Thunberg, & Elmehed, 2000); the
resulting muscle movements trigger afferent feedback to the brain,
producing feelings in observers that match the models.
Conditioning. Empathic distress becomes a conditioned response when
ones actual distress is paired with anothers expression of distress. This
happens in infancy when a mothers distress stiffens her body and is
transferred to the infant in the course of physical handling. The
mothers facial and verbal expressions of distress accompanying her
body stiffening then become conditioned stimuli that can evoke distress
in the child even in the absence of physical contact; and they can
generalize to the facial and verbal expressions of distress by anyone,
which can subsequently arouse distress in the infant.
Direct association makes the connection between a victims expression
of distress or cues in the victims situation and the observers own
painful past experience, without requiring conditioning. Having
experienced separation oneself may be all it takes to empathize with
someone in the midst a distressing separation. Likewise, only if you
have experienced hangover can you empathize with those who wake up
in terrible agony because of having drunk too much; only if you have
had children can you adopt a mothers perspective; and only if you have
slept outside can you understand a homeless person (Hakansson,
2003).
Empathy aroused by these modes is passive, involuntary, based on
surface cues, and requires little if any cognitive processing and
awareness that the source of ones distress is anothers pain, not ones
own. I describe them here in detail because they have been neglected in
the empathy literature. Though limited to empathy with simple
emotions and victims who are visibly present, they are important
because they allow infants to have an empathic response whether or not
they know the source of their pain is anothers distress. More important
here is that all three modes continue to operate and give empathy an
involuntary dimension throughout life. Most of their limitations are
gradually overcome by language and cognitive development, especially
selfother differentiation, which support the cognitively advanced
modes and enable empathy with subtle emotions such as sadness, guilt,
disappointment in oneself, and victims who are absent.
(p.74) Verbally mediated association. Here the victims distress is
communicated and connected to ones own painful past experience
through language, which makes it possible to empathize with someone
who is absent (a letter from the victim, someone elses description of his
plight, a newspaper article).
Perspective-taking. David Hume wrote that because people are
constituted similarly and have similar life experiences, imagining
oneself in anothers place converts the others situation into mental
images that evoke the same feeling in oneself (1751/1957). Adam Smith
went more deeply into the experience of empathy: By the imagination
we place ourselves in the others situation, we conceive ourselves
enduring all the same torments, enter, as it were, into his body, and
become in some measure the same person with him and thence form
some idea of the sensations and even feel something which, though
weaker in degree, is not altogether unlike them (1759/1976, p. 261).
Modern research, begun in the 1950s, reveals three types of perspective-
taking: (a) self-focusedimagining how one would feel if the stimuli
impinging on the victim were impinging on oneself (Humes and Smiths
type), especially when enhanced by association with similar painful
events in ones own past, and to some extent independent of changes in
the victims facial expression, voice, or posture. A widely televised
example is the ex-wife of New Jerseys former disgraced governor
describing how she felt watching the televised scene of the stoic-looking
wife standing off to the side as her husband, the governor of New York,
apologized for a sexual indiscretion (Celizic, 2008): My heart ached for
her when I was watching her. I could see the pain in her face, and I
certainly know what that feels like. Shes there physically, but Im sure
shes not absorbing anything thats going on. Self-focused perspective-
taking can produce as much, at times even more, pain than that felt by
the victim (see discussion below of vicarious trauma in therapists). The
second type of perspective-taking (b) other-focusedfocusing ones
attention on the victims feelings, life condition, or behavior in similar
situations, sometimes enhanced by categorical information
(stereotypes) and theories of how people behave in similar situations
seems to arouse less intense empathic distress than self-focused
(Batson, Early, & Salvarini, 1997). This may be because it is more
cognitively demanding and focused more on the others experience and
less on evoking associations with ones own painful past. In the third
type (c) combined self/other focus, one focuses on both the victim and
oneself, simultaneously or sequentially. This may be the most frequent
mode because people find it difficult to confine their focus on the other
without drifting into their own thoughts and feelings. It is especially
important for empathys contribution in the real world because it
benefits from both the emotional intensity of self-focused and the
enlarged scope and sustained attention to the victim afforded by other-
focused. It is instructive to note that although most people (p.75) can
take anothers perspective, actually doing so is culture dependent (Wu &
Keysar, 2007).
Just as the preverbal modes enable an automatic primitive empathy, the
higher order modes enable mature empathy to develop, beginning with a
veridical empathy that more or less matches the victims feelings in the
immediate situation and progressing to empathy that takes account of
his life condition, situation, personal history, and most important for
present purposes, empathy with a victim group or category.
To summarize, the five arousal modes can operate alone or in any
combination. Together they enable empathy with whatever distress cues
are available: a victims facial, vocal, or postural cues can be picked up
through any or all modes if one is nearby with a clear view of the victim;
situational cues can be picked up through conditioning and association
even if one cannot see or hear the victim; distress expressed verbally, in
writing, or by someone else can arouse empathy through the more
cognitive modes. Multiple modes not only enable but often compel one
to respond to anothers distress empathicallyinstantly, automatically,
with little or no awareness. Even the cognitive modes, often drawn out,
voluntarily controlled and involving reflection, can kick in immediately
if one attends closely to the victim. This multi-determined quality makes
empathic distress a reliable response and may explain why it has
consistently been found to motivate helping others, even strangers, in
distress (Eisenberg & Miller, 1987)though not competitors or people
one envies or actively dislikes, in which case one may blame the victim
or feel pleasure in his distress (Hareli & Weiner, 2002). It may also
explain why empathy may have survived natural selection (de Waal,
2012; Hoffman, 1981) and has a hereditary component (Zahn-Waxler,
Robinson, & Emde, 1992).
A note about mirror neurons. In the mid-nineties when a team of Italian
researchers noticed that certain brain cells were activated both when a
monkey performed an action and when that monkey watched another
monkey perform that same action, mirror neurons were discovered.
Since that time mirror neurons have also been found to operate in
humans, not only for motor acts and intended motor acts, but also by
communicating with the brains limbic system they operate in relation
to certain emotions: similar neural pathways are activated when one
feels disgust and sees the facial grimace of disgust on someone else
(Wicker, Keysers et al., 2003). Some researchers assume the same is
true for other motor-expressive (facial expression, voice, posture)
emotions like pain and anger. If true this would show mirror neurons to
be the neural substrate of mimicry. And some assume future research
will show mirror neurons underlie empathy with any emotion.
For now, there are things about empathy in the real world, where causes
of anothers state are often outside the situation or ambiguous, that
mirror (p.76) neurons cant easily explain: empathic anger toward an
aggressor on behalf of a victim who doesnt feel anger, empathy with a
group like the quotation below from my students empathy with
American slaves, responses like this from Oliner and Oliners (1988)
interviews of Germans who rescued Jews from the Nazis: I think there
was a double feeling: a feeling of compassion for Jews and anger toward
the Germans (p. 118). Mirror neurons also cant explain empathy with
someone whose feelings belie his life condition (a child happily playing,
unaware he is terminally ill). Mirror neuron activation might actually
interfere with processing life-condition information about the victim
and thinking of absent or future victims, as in criminal court trials (see
below). Finally, what we knew about empathy before mirror neurons
discoveryits developmental stages, arousal modes, shaping by casual
attribution, biases, and other limitationswould still apply, as would the
external events that arouse emotions in others that we empathize with.
Mirror neurons may be part of empathic responses but not an
explanation or cause.
What then do mirror neurons buy us? Rochat (2001) criticizes
psychologists who seem to gain intellectual comfort by increasingly
seeking neurosciences molecular and mechanistic high tech stamp of
approval. But there is a good side to this: the idea that all humans share
mirror neurons and mirror neurons cause empathy has helped publicize
empathy beyond the university. As Rochat also notes, knowing monkeys
and humans share mirror neurons helps substantiate evolutionary
continuity in empathic feeling. Some neuroscientists suggest that
spreading the word about mirror neurons may help break down barriers
between peoples and foster world peace (Iacaboni, 2007). So, though we
always assumed empathy has a solid neural base and mirror neurons
may not do much more than tell us what that base is, knowing there is a
base and what it is might ultimately contribute to positive social change.
It might also have medical value, for example, helping find a cure for
autism and other empathy-deficit disorders.
Empathic Over-Arousal
Empathic distress increases with the intensity of the victims actual
distress but can become so aversiveempathic over-arousalthat
bystanders shift attention to their own distress, leave or blame the
victim, or think of other things to turn off the image of the victim
(Hoffman, 1978). Strayer (1993) showed 513-year-olds film clips of
distressed children (unjustly punished; forcibly separated from family;
disabled child climbing stairs). The subjects empathic distress and
attention to the victim increased with intensity of the victims distress
until it reached the level of the victims distress, after which the
subjects (p.77) focus shifted to themselves. Bandura and Rosenthal
(1966) gave adults watching someone being given electric shocks a drug
that intensified their empathic distress, which they reduced with
distracting thoughts and attending to lab details. People are especially
vulnerable to empathic over-arousal when they feel unable to reduce
victims distress or keep their empathic distress within a tolerable range.
Nursing trainees new to hospital wards were so empathically over-
aroused by terminally ill patients that they tried to avoid them, but
changed when they found they could improve patients quality of life
(Williams, 1989). Similarly, children who exert emotional control and
are taught coping strategies for handling anxiety at home are less
vulnerable to empathic over-arousal, can keep empathic distress within
a tolerable range, and maintain their focus on the victims distress
(Eisenberg, Fabes, Schaller, Carlo, & Miller, 1991; Valiente, Eisenberg, et
al. 2004).
Empathic over-arousal and vicarious traumatization. I originally
advanced the empathic over-arousal concept to explain bystanders
turning away from victims. This doesnt apply to people who are highly
committed (witnesses, see below) or whose role requires staying and
helping (clinicians, nurses, rescue workers). There is a substantial
literature on trauma clinicians compassion fatigue or vicarious
trauma (Figley, 1995; Pearlman & Saakvitne, 1995), which I suggest
may be due to empathic over-arousal (Hoffman, 2000, 2002). I asked
125 clinicians how they felt and coped in their most recent therapy
session with a trauma patient.1 They reported lots of empathic over-
arousal with cognitive disruptions and horrible images, nightmares, and
physical symptoms that were often hard to shake off afterward ( . . . I
felt the sadness re-surge and envelope me . . . almost impossible to
concentrate and attend properly; . . . My neck felt strained, tired,
stomach ached, dizzy). Proximal causes of the empathic over-arousal
were patients facial, vocal, and postural expressions of pain (tears,
description of childhood events, crying why dont they understand how
what they do affects me?); vivid trauma narratives that evoked painful
images (I still see the picture I saw as she spoke of the man who hurt
her, he looks so malevolent . . . I imagined her small body size with a
grown adult, her grimace of pain); a calm demeanor that masks intense
suffering (Oh my God hes speaking as though he were describing the
weather [drunken father threatened him and mother with gun] . . . hes
fully dissociated from the feelings, been wounded so badly). In some
cases the clinicians empathic distress is more intense than the patients
distress that evoked it, though of course not the patients original
trauma.
Patients traumas sometimes evoke images and feelings of distress
associated with therapists own past traumas (counter-transference),
which can divert (p.78) their attention from the patient. This can be
harmful, but it can also be helpful when the therapist uses his own
trauma constructively, as in these instances:
His pain/suffering [car sideswiped and crashed in wall] have
deprived him of hope. I have felt hopeless in life so I can relate to
this. I have overcome hopelessness and I believe he can too. He
knows this and in effect he is both asking me to heal him and teach
him how. Because of my experience I feel I can help him do both.
Same thing happened to me [witnessed car-crash death of young
daughter]. Had image of her car crash scene, her smiling child, my
crash scene, my smiling daughter. Patients experience as a whole
evokes my memories, contributes to feelings/images that continue
to resonate forever, but its also cathartic, enlightening, and has
positive outcomes for people who have had the trauma. Burnout
occurs with people who have not had trauma: they fear it, so keep
distance from those who have.
Told very vividly [repeated abuse in childhood by alcoholic father
and recent beating by husband], was very disturbing. History of
physical violence in my own family, so these images keep
reverberating with me and Im trying to use them to get deep
understanding of her feelings in her current situation.
The coping strategies clinicians use to keep empathic over-arousal under
control vary: gaining distance by imagining a patients trauma narrative
is just a movie; splitting ones focus so that one is partly an objective
observer; taking time out by pushing trauma images aside, thinking
about other things, then regrouping; reminding oneself of past successes
with patients. They also use breathing and other relaxation techniques,
consult with colleagues and supervisors, talk it over with their own
therapists, spouses, friends (If I cant get the terrible images out of my
mind I seek co-workers for debriefing), join or start self-help groups, do
volunteer community service, start a strenuous exercise routine. One
clinician (a witness?see below) took to political action on behalf of
people with her patients problem (I pictured that terrified child who
had been physically and sexually abused, neglected, being threatened
with removal from her foster/adoptive family. I had thoughts of lashing
out at the system, and as result of many such cases, I did work to change
it).
Trauma clinicians empathic over-arousal is part of an ongoing process
of interaction between intense empathic distress and attempts to control
it in order to maintain ones professional focus, keep going personally,
and help the patient. Something similar though less dramatic may
happen with any clinician whose patients describe painful though not
quite traumatic events. Being (p.79) a clinician is thus a risky business
and can be hazardous to ones health (several in our sample called
patient traumas toxic, infectious, like a virus passed from listener to
listener entrenched in mind like a foreign body.). That most clinicians
stay with their patients despite bouts of intense empathic pain suggests
clinical interviews are a type of prosocial moral encounter (see
Hoffman2000 for other types), and being a clinician is a moral
profession. The same may be said for other health professionals and
caretakers who treat people in pain on a regular basis.
Empathic over-arousal and vicarious trauma may be experienced by
anyone daily exposed to descriptive news reports of people in pain
throughout the world. This fits the bystander category where one can
easily turn away, but it extends the scope of intense empathic distress
and near-empathic-over-arousal experience to include most Americans
and others with access to the media. This points up the enormous reach
of empathy beyond the laboratory. Empathys actual impact on society,
mediated by laws, is discussed later.
Empathys Limitations
I have long studied empathys s limitations (Hoffman, 1984, 2000),
which I consider serious. To begin, it should be clear from the foregoing
that empathy is limited by its fragility, by its being influenced and biased
in favor of ones similarity to, familiarity with and relationship to
victims, and by its dependence on the salience of distress cues.
1. Empathys fragility allows it to be trumped by egoistic motives like
fear and personal ambition, as in Lyndon Johnsons backtracking when
he ran for president. Its fragility also allows empathy to become so
aversive that bystanders may shift attention to their own personal
distress, leave or blame the victim, or think of other things to turn off
the image of the victims pain. This self-limiting feature of empathy is
not a total loss: it protects the observer and enables him to go on. And as
I noted it may give trauma clinicians temporary relief, enabling them to
regroup and return to the patient.
2. Though people may empathize with anyone in distress, they
empathize more with people like themselves. Recent research makes it
clear, at the neural as well as behavioral level, that people empathize
more with victims who share their race (Avenanti, Sirigu, & Aglioti,
2010) and ethnic (p.94) group (Xu, Zuo, Wang, & Han, 2009) as well as
gender, and with kin and friends. They empathize more with distressed
members of their own team than distressed competitive rivals (Smith,
Powell, Combs, & Schurtz, 2009), unless the latter are viewed as having
lower status (Cikara & Fiske, 2012), perhaps because that signifies added
distress. Finally, in a longitudinal study of AIDS volunteers, empathic
subjects helped in-group members more than out-group members
(Sturmer, Snyder, & Omoto, 2005). For a review and discussion of in-
group bias studies see Cikara, Bruneau, and Saxe (2011).
This in-group or familiarity bias may not be a serious problem in small
homogeneous groups except where there are multiple victims and one
must make a choice. It can be serious in the courtroom in criminal cases
where plaintiff, defendant, judge, and jury are from different class or
ethnic groups, in determining guilt, innocence, and severity of
punishment. It can be very serious in complex multiethnic or class-
stratified societies when scarce resources foster intergroup rivalry,
where it can contribute to violence and empathic anger toward anyone
seen as a threat to members of ones own group. This can add fuel to the
rivalry and make people willing to fight, even sacrifice themselves in
intergroup warfare. Mass murder can seem legitimate to people who feel
their community is under threat and they are merely doing what is
necessary to save the lives of their nearest and dearest. So I didnt shoot
anyone and I really feel bad about it, really bad, because now he will
throw stones at another soldier and could injure him (Israeli soldier in
first intifadaElizur & Yishay-Krien, 2009). It is ironic, and unfortunate
for human survival, that in-group empathy can lead to out-group killing.
Clearly reducing in-group bias is a worthy research objective.
3. Empathys dependence on the salience of distress cues makes people
more likely to empathize with victims who are present than those who
are absent, which I call empathys here-and-now bias (Hoffman, 1984).
This bias is probably due to preverbal empathy-arousing modes being
activated only when victims are present, whether or not cognitive modes
are also involved. It has been shown experimentally: college students
who were induced to empathize with a girl suffering from a fatal disease
decided to move her up the waiting list for a new life-saving drug, at the
expense of children who were more in need and legitimately higher on
the list (Batson, Klein, Highberger, & Shaw 1995).
Empathys here-and-now bias is especially important in law. Posner
(1999) views it as a manifestation of the availability heuristic in the
courtroom when judges give too much weight to vivid immediate
impressions and hence pay too much attention to the feelings, interest,
and humanity of the parties (p.95) in the courtroom and too little to
absent persons likely to be affected by the decision . . . You dont need
much empathy to be moved by a well represented litigant pleading
before you. The challenge to the empathic imagination is to be moved by
thinking or reading about the consequences of the litigation for absent
often completely unknown or even unbornothers who will be affected
by your decision. Posner coined the term judicial empathy, which I
like because it stresses the positive. Posner might advocate prohibiting
or strongly discouraging empathy narratives and victim impact
statements in the courtroom, but he seems to assume that even without
them here-and-now empathic bias is inevitable and it is up to judges to
make every effort to counterbalance it. How? By taking the perspective
of and empathizing with victims and potential victims who are not
present in the courtroom.
An example of how here-and-now bias can interact with the media
(media-enhanced here-and-now bias), and affect society as well as the
courtroom is the highly publicized British nanny trial (New York Times,
November 11, 1997, pp. A1). When the 8-month-old child in the nannys
care was shaken to death, there was widespread media condemnation of
the nanny and sympathy for the parents. After her trial and conviction of
second-degree murder the empathic tide shifted in her favor (empathy
can be fickle). She became the victim and recipient of widespread
empathic distress partly because of the severity of her sentence. A retrial
was denied and the judge, after reviewing the trials history, found fault
with the sentence, described her action empathically as characterized by
confusion, inexperience, frustration, and anger, not malice in the legal
sense. He reduced the jurys verdict to manslaughter and time served
(297 days) and let her return to her U.K. home, reportedly saying, In my
judgment it is time to bring this sad scenario to a compassionate
conclusioncompassion for the nanny; the absent victims were
seemingly forgotten. The extent to which empathic here-and-now bias
affected the judges decision is not clear; what is clear is its impact on
the media and the cases larger social context. But this is just one case.
The larger generalization to be drawn is that empathizing with the
immediate victim can do a serious injustice to absent victims, real or
potential.
Conclusion
It seems clear that an individuals empathy contributed to emancipating
slaves, desegregation, civil rights, and abortion laws, all of which
significantly changed our social, political, economic, and cultural life. Of
course emancipation wasnt due to Stowe alone, but it wouldnt have
happened as soon as it did without her, and the same for Johnson and
civil rights legislation. Empathy narratives contributed to Supreme
Court desegregation, abortion, and Miranda decisions despite the laws
traditional view of empathy as anathema (p.96) to justice. These laws
had a substantial effect on society, to which we can add empathys
helping prepare the culture, society, and public policy changes that made
them possible in the first place. Not all laws relate to empathy, but those
pertaining to civil liberties, civil rights, and societys disadvantaged do.
They reflect Pinkers better angels of our nature and fit Rawlss
definition of a moral society as one that promotes the welfare of the
disadvantaged (although Rawls denies the role of empathy.)
Not all empathys contributions to society are positive. Empathic bias is
a serious problem, particularly in-group bias, which fosters intergroup
hostility and can keep nations from getting together to solve human-
survival problems like climate change. It may have survived natural
selection and be part of human nature, making empathy itself an
obstacle to Rifkins empathic civilization. There is also the current
human condition, with crises worldwide involving victims described by
the media in vivid detail every day, making everyone vulnerable to
empty empathy, which serves no constructive purpose and may waste
ones empathic resources and reduce the likelihood of empathizing
when needed in the future. Some may respond with too much empathy
and to avoid the unbearable pain put up defenses that immunize them
against empathizing with future victims.
Putting it all together, I still view empathy as the bedrock of morality,
the glue of society, and an important factor in changing laws and society
in a prosocial and projustice direction. Its not a panacea and has flaws.
It cant override powerful economic and political forces, ethnic divisions,
natural disasters, and personal ambition and has the other flaws I noted.
But its not just a sideshow, and in the absence of a viable alternative it
may be the only available human resource for keeping the diverse parts
of society together. If so, it is necessary to do whatever it takes to keep
empathy alive. This means recognizing and finding ways to overcome its
limitations and harness its power to serve communities larger than
ones own group. How? Perhaps, with help from the media, people could
be taught (a) to think about potential and absent victims of their actions,
the actions of others and of certain institutions (laws, taxes, tax cuts) by
imagining for example that the victims, even enemies from outside ones
group, are ones own loved ones; (b) to know that everyone regardless of
culture and including ones enemies share the same basic hopes, fears,
pain of loss, anxiety over climate change and human survival; (c) to pass
laws that require the kinds of interaction between groups and cultures
that reveal these commonalities and promote working together for
common objectives. Research wont solve it but can help, for example
research on the value of taking the others perspective (Galinsky &
Moskowitz, 2000) and on directly intervening to reduce intergroup
hostility and promote positive interaction (Cikara, Bruneau, & Saxe,
2011).
I
Suppose your young teenage child deeply desires to go to a certain
excellent film (with their friends) instead of studying for a final exam
that they have very early the next day. If you know just what your child
desires and feels and you had similar experiences as a youngster, and
this helps to explain why you know what it is like to be in their shoes,
and all of this results in your likewise desiring that your child have this
same wonderful experience (of going to the film with their friends), then
this would be a case of empathyagain, as we plan to use this term.
Consider a second scenario. Suppose that, in addition to having empathy
toward the child in their situation, you start reflecting on your childs
future, on the fact that your daughter or son is academically very
talented and earnestly interested in becoming a doctor, and you reflect
further on how extremely competitive getting into a medical school can
be. Moreover, you realize that the final exam the next day could be
critical in their effort to get into medical school. Its for an advanced
placement course in biological science. These additional thoughts
concerning your childs desires and future frustrations and upset in
failing to get into a medical school could result in your having empathy
for your childs future and thereby desiring them to take the test
tomorrow in order to avoid future suffering and distress. In this future-
orientated way, your empathy for your childs desire to go to the film
with their friends would be in conflict with your empathy for the childs
future happiness. Furthermore, further reflection may lead you to see
that it would be better for the child from the childs own standpoint and
likely greatest happiness to forgo the present anticipation of pleasure
rather than to suffer as a result of career aspirations that may be
permanently blocked. So, out of empathy for your childs future desires
and future happiness, you may reflectively control your empathetic
impulse toward the childs current desire to go to the film with friends.
(p.125) We can even suppose that there is a third scenario. In this case,
after your child has accepted your decision that they give up the film,
you learn that the same test will be given after the weekend on Monday
by the same teacher. So, you call the teacher and explain the situation
regarding your child and you ask the teacher if your child might not take
the test on Monday instead of Friday (the next day). The teacher is
accommodating and kindly grants your request. In this case, your child
can see the film, take the test Monday, and have the whole weekend to
prepare for it. So, we can imagine how still further reflection plus some
empirical investigation changes your decision once again. All along, your
empathy for or imaginative identification with your child is the drive
behind your actions.
These three possible scenarios, as homely as they are, teach us that the
marriage between empathy and reflection/investigation is not only a
possible marriage, but may be very good marriage for everyone involved.
It may help to promote the best interest of the person toward which one
has an empathetic reaction.
What we now need to do is to characterize empathy in a form that is
distinctively human and reflective. This is the kind of empathy
illustrated above. What makes such empathy distinctively human and
reflective is that only a creature with a descriptive or propositional
language can reflect on the other person (and they on themselves) in the
manner above. After all, there is no evidence whatever that nonhuman
animals can describe the world, themselves, or others or discriminate
between descriptions that are true and descriptions that are false.
Everything distinctively human depends upon such a language: religion,
politics, law, science, technology, philosophy, and art all presuppose an
ability to describe both what is real (true) or what might be real (true)
and what is entirely imaginary or fictional (false).
Having what we here are calling a descriptive language is precisely what
permits us to tell others how we feel, what we consciously think, what
we consciously experience and, thereforewhat it is like to be us from
the inside. This is how the child (in the story above), we assume, is able
to communicate to the parent what they was experiencing internally.
Without the childs words or talk to go on (with only facial expressions,
bodily postures, behaviors and the like) a parent would have little or no
clue about what is going on internally in the child, little or no clue that
all the psychological commotion was about going to the movies with
their friends. Indeed, we persons could not even so much as have certain
feelings, certain experiences, beliefs, values, or thoughts if we did not
possess a descriptive language. We could not for example have feelings
about what is happening in China, about landing on the moon, about the
president of United Statesor about attending medical school. Nor could
we describe to ourselves the various nonverbal feelings, thoughts, or
desires (p.126) that we have without a descriptive language. So, having
a descriptive language has been utterly transformative for humankind
and, in particular, it has radically transformed the nature and complexity
of our capacity for empathy with other persons.
What then is empathy of the distinctively human and reflective kind?
We may characterize empathy of the distinctively human and reflective
kind as follows: When we are empathetic in the distinctively human way
we not only imaginatively share anothers inner life and emotions, but
we do so in a fully descriptive and appreciative way, and we will therein
have the very same desires for them that they have for themselves and
thereby a very powerful inclination to act on their behalf. So, there is
nothing to prevent us from merging empathy or empathetic behavior
with the analysis, investigation, or rational assessment of our conduct.
To be fully human, our capacity for empathy, in fact, needs to be infused
with reflection. Stripped of our descriptive powers, we may share in the
sorts of non-descriptive empathetic capacities of which certain species of
nonhuman animals seem capable. Yet, when infused with such
descriptive powers we can, as it were, imagine being our child happily
and successfully pursuing our career as a doctor and we can equally
imagine being our child and being permanently blocked from our career
as a doctor, with all of the consequent frustration and misery. That said
we now turn to examine the role of empathy in our moral life.
II
In order to examine the role of empathy in our moral life, and the
reasons why we believe that empathy should be at the center of our
moral life, we need first to consider the concept of what truly is best for
or in the best interest of a person. This is the concept whose application
helped the parent in the example above to realize that the child may
need to forgo going to the movies with their friends.
Let us begin by noting that it follows analytically from the very concept
of what is truly best for some person that nothing better can be so much
as conceived. If something truly is best, its the best, period. Nothing else
is better.
To be sure, it may seem that the application of such a concept of the best
for us is beyond human knowledge or any kind of certainty. However,
further reflection reveals that this concept, nonetheless, has an
important role to play in guiding our ordinary everyday actions, and we
often understand that our daily practices assume that the concept is
applicable. Consider the following simple example: Barry is all alone
sleeping on the couch in his home, when suddenly he wakes up and
becomes aware that there is a raging fire all around him. Under the
circumstances, we can plainly see that what is truly best for (p.127)
Barry, at that particular time, is simply to get out of the house as fast as
he can. Although we can conceive of some conduct it might have been
better for Barry to perform: For example, we can conceive of a possible
world in which he could just push a button and the fire would go out and
we can imagine other scenarios in which getting out of the house as
soon as possible would not be what is truly best for him. Nonetheless,
under the present circumstances it is hard to imagine anything better
that he might do.
The example of Barry and others like it are quite compatible with the
fact that there will be situations where we are completely at a loss to
know or even guess what might be the truly best thing for us to do. But,
and this is crucial, this is not always or even usually the case. There are
many cases where we can be extremely certain or at least quite certain
enough that we are doing what is truly best for ourselves or another. We
indeed make this assumption when were making careful, well-thought-
out decisions about medical care for ourselves or for loved ones,
decisions about buying a home, decisions about our childrens schooling,
and much, much more.
We can, moreover, improve our ability to determine what is truly best
for ourselves by becoming more rational in examining and assessing
ourselves and our circumstances. By being rational here we mean
reasoning in ways that are likely to lead to the truth or insights about
self and circumstance, and irrational would simply mean to reason in
ways that are likely to lead to falsehood. Thus, the more informed we are
about our situation and a particular problem at hand and the more
knowledge we have about the world, about ourselves and about other
persons, the better off we will be at determining what is truly best for
ourselves or for another. Again, in many cases, we can discuss or consult
experts on the matter. And we can become more skilled at distinguishing
good reasoning (i.e., arguments that are very likely to lead to the truth)
from bad reasoning (i.e., arguments that are very likely to lead to
falsehood). We can cultivate habits of action and mind that will
eliminate our intellectual vices, for instance our judgmental egotism,
our lack of epistemic humility, our out-of-control passions. And we can
likewise cultivate intellectual virtues or habits of action and mind that
will enhance our performance at good reasoning, for example listening
very carefully to others, being open to completely new ways of seeing
things, valuing truth as a supreme good, and having the courage to admit
that we are wrong (when this is the case). In these and various related
ways, we can get better and better at determining what is truly best for
ourselves as well as what is truly best for other people.
Thus aiming for what is truly best for ourselves or for others is not only
something we do and should always try to do, it is something we can
improve at doing. So, aiming at what is truly best for ourselves and for
others is entirely realistic and desirable.
(p.128) III
In order to better understand the concept of what is truly best, we now
need to consider the kind of things to which the concept applies. For
example, a mechanic may say The truly best oil for your car is such and
such oil. However, what the mechanic means is simply that such and
such oil is least expensive, lasts longer, is most environmentally
friendly, will enable your car to perform at its best, and so on. They dont
really mean it is truly best for your car, but rather that it is best for you
and/or other people. If, of course, we imagine that your car is conscious
(that there is something it is like to be your car) and our car could suffer
or feel pain or undergo pleasure or joy, then we could sensibly talk about
what was truly good or bad, better or best or worst for your car from
your cars standpoint. So, we can talk about what is truly best for your
car or anything, only so far as your car or anything else is something for
which we can have empathy, that is, something that has an inner life or
subjective perspective. This is hardly surprising if we consider that in
both cases (in the case of empathy and in the case of what is truly best
for something), we are dealing with a being that is conscious, a being
that can suffer or experience pain and that can experience joy or
pleasure.
This means that we can talk about what is truly best for something only
so far as we have access to that things inner life or subjective
perspective through empathy. Empathy is the doorway to understanding
what is truly best for any creature or person.
But what has all this to do with empathy being at the center of our moral
lives? To help to answer this question, we need to consider what it
means for human beings to live in various forms of union and ideally in
interpersonal harmony.
IV
There are different ways in which two or more people or even a whole
village, society, or nation may cohere or form various kinds of union.
There is, for example, such a thing as honor among thieves, where
mutual dependency can hold all kinds of people together, people who
care nothing about each other and have not the slightest empathy for
one another. In such cases, people are simply useful instruments for one
another. There are other ways in which people may be bound together.
For example, a culture may speak of peoples moral obligation to help
one another and to avoid harming one another. And we can further
imagine that such a culture or community may teach that what is
morally good is good in and of itself apart from any considerations
(p.129) of empathy. So, we can certainly imagine that societies are
variously bound together by some sort of quasi-moral glue that has
nothing whatsoever to do with empathy.
At the same time, to the extent that we may empathize with other
persons, we will not harm them and, indeed, we will try to help them.
Consequently, it would seem that a family, tribe, society, nation, or
world that is bound together by empathy will turn out to be the best of
social worlds. A world in which the glue, as it were, that unites people
and holds them in union does so in a manner infused by peoples
appreciation of and reflection upon what it is like to be other persons
to possess anothers desires, beliefs, emotions, and so on.
However, there is a social complication here in the very possibility of
union among persons, namely the phenomena of genuine conflicts
between what is truly best for one person, group, family, or nation and
what is truly best for some other person, group, family, or nation. In
order to make this problem or complication entirely clear, we need to
consider the difference between a conflict that is genuine or real and one
that is artificial or fictitious.
Suppose a woman or man wishes to become a senator and has an
opportunity to do so. The husband or wife of the person in question may
fear that this will negatively impact upon their relationship with one
another, their lives in general, and upon the lives of their children. So,
much marital discussion ensues and each partner, failing to understand
the other, may, at times, become angry and alienated from their spouse.
However, suppose the couple goes to an excellent therapist who helps
them work through the problem carefully, thoughtfully, and
empathetically. And suppose that after extensive sessions, the husband
or wife (whichever one was fearful) is subsequently willing to take a
chance and go along with his/her spouses becoming a senator. In this
case, let us assume that as things turn out, the fears of the husband or
wife were quite unfounded and that in fact the spouses becoming a
senator positively impacts upon all of their lives. In retrospect, they all
come to realize that there was no genuine conflict of interests, only an
apparent conflict of interests.
Note how empathy is an important key here. Without empathizing with
one another, each spouse would only be able to really understand their
own personal feelings and desires and may have viewed the other as
being quite irrational or unreasonable with the result that no resolution
would have been reached. Indeed, we may go so far as to say that if each
party has enough empathy for the other person or persons in any
situation, then with sufficient patience and talking there is an extremely
good chance of resolving the conflict in question. We may even say that
if there was enough empathy in the world at large, then people would
never deliberately harm one another and would always be willing to try
to help one another when the conflict was merely apparent. This would
be a tremendous step forward for all of us, since many (p.130) conflicts
that arise between people are, like the above example, only apparent and
not genuine or real.
Unfortunately, things are not quite so simple. Genuine conflicts of
interest are not conflicts that will go away if only the parties caught up
in the conflict have sufficient skills, if only they have their passions and
emotions under control, or if only they are open and willing and able to
listen empathetically to one another. These are the kind of prima facie
intractable conflicts that we must ultimately consider in order to see just
how far empathy can take us.
Lets start with a simple example. Suppose two friends are camping out
in the woods after a long hike and that they are at least 5 miles from any
drugstore, hospital, or physician. Unfortunately, they unknowingly settle
down in an area that is infested with rattlesnakes. Consequently, the two
of them no sooner fall asleep, when they both awake from fatal
rattlesnake bites. Even worse, they discover that they only have enough
anti-venom with them to save one of them. If they try to save both, by
sharing the anti-venom, it is very clear that both of them will die. So, at
best, only one of them can be saved. Here we have a genuine conflict of
interest, a conflict of interest that will not go away or be dissolved by all
of the talking and empathy in the world. In the end, at least one of them
will die.
Many would argue that this kind of an example shows the moral limits
of empathy. In one sense this is true, but in another sense it is far from
the truth. It is true that all the empathy in the world will not eliminate
such genuine conflicts of interest. Genuine conflicts of interest are not
dissolvable through mere empathy alone. Moreover, such genuine
conflicts are a permanent feature of human existence. However, as we
shall see, genuine conflicts of interest are resolvable, resolvable in ways
that make for the best world that humans can imagine constructing or
bringing about.
V
Suppose in our actual world we possessed a special sort of empathy. This
special sort of empathy would be empathy of a type that is no mere
empathy but that reflectively takes account of what is truly best for each
and every person. What we are imagining here is empathy that is
tantamount to assuming the equal worth of persons. Let us call such
empathy morally sufficient empathy or simply sufficient empathy.
What exactly is meant by the words sufficient empathy? Sufficient to
accomplish what is the question? And the answer to the question is:
Sufficient to produce a world in which there is maximum harmony
among persons.
(p.131) Suppose that you had so much empathy for another person that
they became, to you or in your own mind, the moral and spiritual
equivalent of yourself. Return to the two friends bitten by rattlesnakes
deep in the woods with only enough anti-venom to save one. Suppose
that each friend had so much empathy for the other that each friend
cared about their friend as much as they cared about themselves. Thus,
for example, the death of the friend would be as horrible to each of them
as their own death. Under such conditions, we can easily imagine that
they would resolve their problem by according an equal chance for
survival to each one of them, that is, they would let random chance
decide their fates. Since, for each one, their own death and the death of
their friend would be equally horrible, by, say, flipping a coin they could
accept their fate, either way, with grace and peace. This is sufficient
empathy at work. In a world in where each person had sufficient
empathy for every other person, this would amount to a world in which
each and every human life has an equal worth for everyone.
So, lets imaginatively extend our concept of sufficient empathy to the
entire world and consider the consequences. There are all kinds of
genuine conflicts of interest throughout the globe, genuine conflicts of
interest between and within individuals, between and within families,
groups, organizations, political parties, religious groups, townships,
states, and nations. So, the world is full of strife as a result. Nonetheless,
we can at least imagine a world in which human beings have evolved or
developed to a point where each and every human being has sufficient
empathy for every other human being. Put another way, the worth of
every human being for every other human being would be equal to their
own worth. Since equals are equal to equals, it follows that if every other
human being in the world had the same value to you as you have to
yourself, then everybody in the whole world would necessarily have an
equal worth. No one in the whole world would have more (or less) worth
for you than anyone else. So, we are trying to imagine a world in which
everyone has sufficient empathy for everyone else, a world where
everyone has an equal worth for everyone else.
The notion of such a world is immensely complicated of course, and
there are terms in its description that need careful analysis of a sort that
we cannot provide here. But let us proceed to outline or sketch what we
have in mind and consider the idealistic possibility of sufficient
empathy.
In such a world, no one would deliberately harm anyone else (any more
than they would deliberately harm themselves). Nor would anyone
deliberately fail to help any other person (any more than they would fail
to help themselves). So, in such a world there would be what may be
called maximum harmony of everyones best interest. However, is such
a world even close to being empirically (p.132) possible, even in the
remote future? Can we expect humans to evolve to such a high moral
state?
There are reasons to believe that we might. In the first place, it would be
the best sort of world to live in, since everybody would gain from a world
in which there is maximum harmony. After all, in such a world, conflicts
of interest would either be dissolvable (in most cases of apparent
conflicts) or at least resolvable (in most cases of genuine conflicts). So,
there would be no war, no strife, no hostility, and no harm deliberately
done to one another. Moreover, everybody would help everybody else as
far as possible. Each person living in such a world would surely benefit
immensely and be far happier than the world in which we currently live.
It would be a world of win-win or gain-gain for everyone.
However, and here is the rub, how do we get from the world we actually
live in, a world filled with greed, vanity, hostility, bigotry, hatred, and,
above all else, selfishness to the world that we are imagining? How is
such a world possible? After all, as good as such a world might be for
everyone, the question remains: Who is going to take the first step?
Who is going to cultivate such incredible empathy so as to have such
empathy for each and every person in the world? Who may make such a
supreme sacrifice, if, indeed, such a thing as sufficient empathy for each
and every person in the entire world is at all possible and not just some
delicious pie-in-the-sky ideal that philosophers, such as the two of us,
have dreamed up? Would not such a sacrificing person (or persons) be
trampled by those around them who are anxiously looking out for
themselves?
Indeed, it would seem to be a fruitless and impossible sacrifice to try and
have sufficient empathy for the entire world, if what is required in order
to make the sacrifice is the expectable prospect of receiving rewards
from other people. In our woefully morally imperfect world, the reward
for moral behavior often is not the reciprocal and appreciative behavior
from others. So, what then could be the reward?
We claim that the reward is a special kind of personal happiness. Not the
happiness of a happy or pleasurable but transient feeling, but the
happiness that comes from passing positive judgment on [ones] life; . .
. on the living of it and on whether [one] is [deeply] satisfied with its
course and character (Graham 1998: 193; see 192197 for discussion).
This is the happiness that many religious persons have spoken of when
they have described the experience of finding the greatest happiness
through helping and behaving morally toward others and therein giving
happiness to others.
Religions often clothe this claim about human happiness in various
muddled theological and confusing metaphysical garments. These outer
garments, though often perhaps harmless in and of themselves,
sometimes blind many (p.133) proponents of different faiths to the
true virtues of other faiths and consequently have divided people of
different faiths, breeding endless controversy, intolerance, hatred, and
war. What we need today is to see beyond these outer metaphysical
garments and embrace the real core of a claim that can unite us, the
proposition that those who spread happiness enjoy a far deeper and
lasting happiness than other people. Whatever the ultimate nature of
the universe (which is beyond human knowing), what stands fast and
endures above all other truths is the fact, or so we claim is a fact, that
those who have sufficient empathy, those who understand that all
human beings have equal worth, and those who have gotten beyond a
narrow, self-centered existence enjoy the greatest happiness humanly
possible.
Granted this essential truth, our primary goal in life should be to
empathetically value others as if they were ourselves. Such empathetic
valuations will not only produce the greatest happiness for our own
person, but for others as well. Indeed, it quite clearly follows that the
very highest and best gift that we can give any human being is to help
them to have sufficient empathy for each and every other person on the
globe thereby becoming a person who likewise seeks the happiness of
others. In doing this, we not only help them to achieve the greatest
happiness possible, but set them on a path that will help them to help
others to do the same, thereby spreading happiness throughout the
globe.
VI
A grand ideal? Indeed. But there are those who would concede that all of
the consequences adduced above really would occur only if it were
possible to have the kind of sufficient empathy described above. But
they deny that having sufficient empathy (for everyone in the world) is
even close to being humanly possible. We just are not built by nature, by
the circumstances of human culture, or by the demanding character of
life itself to have such sufficient empathy.
In its most extreme form, this objection would say that we are incapable
of having sufficient empathy for even one single other person. But this
extreme form of the objection is clearly mistaken. For many people who
are loving parents or grandparents deeply understand that such empathy
is possible, since they have experienced it firsthand with respect to their
own children or grandchildren.
In a less radical form, the objection may be expressed by saying that our
empathetic concern for individuals can blind us to the common good
and to what we would otherwise take to be morally right. As a
consequence of an over-concern or solicitation for certain specific
individuals, such empathy would incline us to put the wellbeing of some
individuals ahead of others (p.134) by exaggerating the importance of
the individual when compared with the common good.
However, some people can and do come to empathize with whole groups
of persons. This is illustrated by the profound impact of Harriet Beecher
Stowes mid-19th-century novel Uncle Toms Cabin, which was written
prior to the Civil War in United States. This wonderful book enabled
white Americans to understand what it was like to be a slave and thereby
helped people to empathize with an entire group of people. The effect
was so powerful that this single book helped to inspire the movement to
abolish slavery in America.
Equally important is the fact that what we are describing as sufficient
empathy, above all else, is what prompts us to seek, again above all else,
what is truly best for each and every person on each and every occasion,
which is to do all that we can to help them in turn come to have
sufficient empathy for each and every other person. Given this
condition, there is no way that the common good and the good of the
individual can be in competition. Quite to the contrary, to the extent that
we succeed in helping others to have sufficient empathy for each and
every person, we thereby serve the common good throughout the globe.
A final version of the objection to the possibility of having sufficient
empathy for all human beings was posed by David Hume, who famously
pointed out that we have far greater empathy for those close to us in
space and time, or for those, whom he puts it, to which the mind from
long custom has become familiarized, than for those removed from us
(see citation in Morton 2011: 325). What makes Humes observation
appear compelling is that it is in fact quite true of many (or possibly
most) people. After all, it makes good evolutionary sense to say that we
very likely do have a genetic predisposition to be most empathetic to
those closest to us. Moreover, it is factually true that many of our
various cultures reinforce this tendency to be more empathetic to those
belonging to our group or tribe and, in fact, if anything, reinforce our
tendency to be hostile or at least less empathetic to outsiders.
However, the crucial question is not whether in fact many or most
humans are much more empathetic to those who are closer to them in
time and space, but whether or not they have the potential to be equally
empathetic to those who are removed from them in time and space.
It is, indeed, necessary and desirable to have special persons in our lives
such as family and friends. Most people are raised, fed, housed, trained,
loved, and educated by their family, friends, and teachers. All of these
people, close to us in time and space, made extraordinarily important
contributions to our development as persons. We could not have become
the persons we are without such special persons or significant others in
our lives. Indeed, humans would never have evolved into creatures with
a descriptive or propositional language (p.135) had they not lived in
families or small tribes. Nor could they, therefore, have developed
sufficient empathy for anyone else or have developed into moral beings
without family and friends (and others close in time and space) in their
lives. We might call such special persons our primary tribe (see Garrett
1989 for related discussion).
Over time, we come to join other tribes, such as political parties,
religious groups, and the like. And, moreover, we automatically become
part of still other tribes simply as a result of where we live, our
neighborhood, our town or city, our county or state and our nation, and
we can even speak of the international tribe, the tribe which includes the
entire world. Technology has made that world tribe more and more
possible, because as technology develops (especially in the realm of
communication and transportation) time and space shrink and we find
ourselves affected by and interacting with people who live at enormous
distances from us. As this happens, we move more and more toward
both the possibility and the necessity of becoming a single tribe.
Tribes are a good thing. The problem is not with tribes, with
communities that bond, nurture, and connect us, but with tribalism,
with the very tendency of which Hume speaks, namely the tendency to
favor those who are closer to us in time and space or, we might add, in
race, religion, political ideology, and all the rest. Tribalism views
members of other tribes with suspicion, bigotry, hatred and, frequently,
as nonpersons. Tribalism is quite real and it is the biggest threat to the
world today, precisely because it is the biggest threat to a world in which
each person has an equal worth, a world in which there is maximum
harmony.
However, there is a remedy for the disease we might call tribalism. It is
the development of sufficient empathy for each and every person in the
world.
History has shown us again and again that it is indeed quite possible to
go against our natural inclination toward being attached to our own tribe
when doing so is morally unhealthy or contrary to one or another moral
imperative. The very first thing that we need to do is come to see that
those belonging to tribes other than ours (and even engaged in struggles
with our tribes) as human beings just like ourselves. This is the teaching
of the parable of the Good Samaritan in the New Testament. According
to the Gospel of Luke (10: 2937) a traveler (who may or may not be
Jewish) is beaten, robbed, and left half dead along the road. First a priest
and then a Levite come by, but both avoid the man. Finally, a Samaritan
arrives on the scene. Samaritans and Jews generally despised each other,
but the Samaritan helps the injured man. Jesus is described as telling
the parable in response to a question regarding the identity of the
neighbor who Leviticus (19:18) says should be loved. Quite literally,
the meaning here is that anyone who can help or harm us and whom we
in turn can help or harm is our neighbor. The point is made in this way
in order (p.136) to illustrate to the Jewish tribe that those belonging to
the alien Samaritan tribe can be just as loving and kind as can Jews.
The second thing that we must do to overcome tribalism is to practice
sufficient empathy (i.e., regard all others as having an equal worth) with
respect to everyone, no matter who they are or what tribe they belong to.
Even in cases where we dont immediately feel sufficient empathy, we
should act as we would if we had sufficient empathy for the person or
tribe in question.
Something else must also be recognized. Cultivating sufficient empathy
for everyone in the world is a lifetime pursuit. Gautama Buddhas
Eightfold Path is a deep analysis of what humans must do in order to
transform themselves into beings that have sufficient empathy for each
and every conscious living being. We have to understand the truth that
without such sufficient empathy, life is suffering for ourselves and for
others, and we have to understand that the cost of our suffering and the
suffering of others is due to the lack of sufficient empathy. We need to
see how attachment to our fears, desires, hatreds, prejudices, passions in
general, and especially our attachment to ourselves (to the ego) cause
our blindness and our inability to cultivate sufficient empathy. And we
need to see how such passions frequently lurk beneath the surface of
consciousness and need to be brought forth and made part of our daily
awareness, so that we can control them, instead of being controlled by
them (see Garrett 2011 for related discussion).
If we understand that various religious and other communities actually
exist and actually bring about such radical transformations in persons,
then here above all we see that humans really do have the potential to
have sufficient empathy for each and every person in the universe, for
each and every person whose life they can affect.
Thus, we can agree with Hume when he says that most people in fact do
not have sufficient empathy for their neighbor (i.e., for those whose
lives they can affect). But this in no way contradicts the empirically
relevant point that we have the potential for practicing empathy for
every living person in the world, for those whose lives we can affect. And
we have the potential therein for respecting the equal worth of persons.
Conclusion
Jesse Prinz describes his paper on empathy, cited at the beginning of our
paper, as a kind of campaign against empathy, that is, against
empathys occupying a central role in our moral lives (Prinz 2011a: 212).
He also claims empathy is a response directed at individuals, and many
of the most urgent moral events involve large numbers of people (229).
What we really need is an intellectual recognition of our common
humanity (229).
(p.137) One way of reading our chapter is that we agree with Prinz that
we persons need both to recognize our common humanity and to
respond to urgent moral crises facing large numbers of people, but we
disagree that this means that empathy should not play a central role in
our moral lives or behavior. Our capacity for empathy is not necessarily
restricted to our tribe. It is not confined to feeling only for the primary
persons in our life. It is a power of identification with others in a
manner that can respect our equal worth as persons as well as what is in
the best interest of each and every person affected by our actions. It
should not occupy an exclusionary or non-complemented/partner-less
center in our moral lives (for down that path lies tribalism). It should
occupy a center with complementary components, such as commitment
to the equal worth of persons and to our own and other peoples best
interests. Morally sufficient empathy, so understood, is an ideal center
each step toward which may make it more empirically real and socially
encompassing.
Empathy and Moral Deficits in
Psychopathy
Abigail A. Marsh
In the disaster he brings about he cannot estimate the affective
reactions of others which are the substance of the disaster . . . the
real psychopath seems to lack understanding of the nature and
quality of the hurt and sorrow he brings to others.
Hervey Cleckley, The Mask of Sanity
If there is a single psychiatric condition that is defined in terms of
morality, it is psychopathy. Modern conceptions of psychopathy
emerged from 19th-century observations that the primary affliction of a
subset of criminal and mentally ill populations was a breakdown of the
moral faculties. Benjamin Rush described individuals afflicted by
apparent perversion of the moral faculties (Rush, 1812), and James
Cowles Prichard created a diagnostic category, which he termed moral
insanity, that was marked by moral or emotional madness in the
absence of hallucinations or delusions (Prichard, 1837). The
contemporary definition of psychopathy reflects these origins,
incorporating both deficiencies in moral emotions like empathic concern
and guilt and persistent immoral behaviors like deceit, conning, theft,
and interpersonal violence (Hare, 1991). Although other psychological
conditions, such as borderline personality disorder or post-traumatic
stress disorder, are associated with increases in immoral behavior, there
is no other disorder for which immorality is such a central feature.
Because of this, psychopathy is an important phenomenon for (p.139)
better understanding the nature of human morality. In learning about
individuals in whom moral emotions and behavior are consistently
impaired (in the absence of other major cognitive impairments) we may
derive important information about the neurocognitive systems that
support morality.
Since the development of reliable scales for measuring psychopathy
(e.g., Hare, 1991; Forth, Kosson, & Hare,2003; Lilienfeld & Andrews,
1996), research on psychopathy has burgeoned, driven in part by the
practical importance of understanding individuals whose
disproportionately violent and criminal behavior is costly to society
(Rutter, 2012; Hare, 1993). Understanding the nature of moral deficits in
psychopathy is therefore both a pragmatically and a theoretically useful
endeavor. This chapter will review the accumulating empirical research
that supports the roots of specific moral deficits in psychopathy. In
particular, it will explore the evidence that moral deficits in psychopathy
emerge from fundamental deficiencies in the capacity for certain forms
of empathy and will consider the possible neural basis for these deficits.
These findings may illuminate the role of empathy in moral judgments
and behavior more broadly.
Psychopathy
The earliest criteria for assessing psychopathy were formulated by the
psychiatrist Hervey Cleckley, whose book The Mask of Sanity was first
published in 1941 and has since become a touchstone for psychopathy
researchers in the modern era (Cleckley, 1982). Cleckley noted that a
subset of patients in the mental institutions where he worked were set
apart by characteristic features. These features included an absence of
afflictions typical in institutionalized patients, such as delusions,
irrational behavior, suicidality, and nervousness or neuroses. These
items were included in his original 16 criteria for psychopathy, in
addition to items related both to moral emotions (such as a lack of
remorse or shame) and to immoral behavior (including untruthfulness
and inadequately motivated antisocial behavior). Of note, Cleckley also
included the failure to learn by experience as symptomatic of
psychopathy, as many of the psychopathic individuals he observed
persistently engaged in deviant behaviors, seemingly undeterred by the
prospect of negative consequences like future incarceration.
More recently, Hare and colleagues applied psychometric techniques to
create a reliable instrument for researchers to assess psychopathy in
institutionalized populations, called the Psychopathy Checklist (now in
its revised form, PCL-R) (Hare, 1991). The PCL-R is a 20-item scale with
a maximum score of 40 that was originally created and standardized in
male prison (p.140) populations rather than patients in mental
institutions (Hare,1991). The selection of items on the scale reflects this
fact. It features more items specifically assessing criminal behavior
juvenile delinquency, criminal versatility, revocation of conditional
releasein addition to items featured in Cleckleys criteria, for example,
lack of remorse and pathological lying. Absent from the PCL-R, however,
are the items that Cleckley included to distinguish psychopaths from
other mentally ill populations: the absence of nervousness, absence of
delusions or irrational thinking, and infrequent suicidality. The ubiquity
of the PCL-R and its variants in forensic psychiatry and related
disciplines has led to this instrument being described as the gold
standard for the measurement of psychopathysometimes to the
extent that the instrument is considered synonymous with the
psychopathy construct (Skeem & Cooke, 2010; Skeem, Polaschek,
Patrick, & Lilienfeld, 2011; Ermer, Kahn, Salovey, & Kiehl, 2012). That
said, drawbacks of this instrument include the requirement that file data
or other background information be used when scoring it (meaning that
it cannot easily be used in noninstitutionalized samples); its heavy
reliance on items assessing criminal behavior (Skeem & Cooke, 2010);
and the exclusion of items that assess fear or anxiety, which has led
some investigators to supplement the scale with anxiety measures or
clinical assessments of anxiety disorders (e.g., Koenigs, Kruepke, Zeier,
& Newman, 2012; Marsh et al., 2008).
A variety of self-report measures of psychopathy are also available,
which are generally reliably correlated with the PCL-R (Malterer,
Lilienfeld, Neumann, & Newman, 2010; Poythress et al., 2010) and
obviate the need for file data, permitting psychopathy to be assessed in
noninstitutionalized community samples. The use of self-report
measures in community samples is consistent with the idea that, like
most other psychological disorders (Markon, Chmielewski, & Miller,
2011), psychopathic traits are continuously distributed in the population
(rather than being taxonomic in structure) such that information about
psychopathy can usefully be drawn from both clinically diagnosed
samples and community samples (Edens, Marcus, Lilienfeld, &
Poythress, 2006; Guay, Ruscio, Knight, & Hare, 2007; Malterer,
Lilienfeld, Neumann, & Newman, 2010).
Views on the factor structure of psychopathy vary (Skeem, Polaschek,
Patrick, & Lilienfeld, 2011; Jones, Cauffman, Miller, & Mulvey, 2006),
but the classic division of psychopathic traits is a two-factor solution
incorporating socio-affective traits termed callous-unemotional traits
that include lack of guilt or remorse and shallow affect; and antisocial
and under-controlled behaviors, like irresponsibility, impulsivity, and
poor anger control. Antisocial behaviors observed in psychopathy may
also be observed in other deviant populations, but callous-unemotional
traits set psychopaths apart and are often referred to as the core
features of the disorder (Sylvers, Brennan, & (p.141) Lilienfeld, 2011).
Assessments of children typically focus only on these traits, and a
callous-emotional traits specifier has been proposed for children
diagnosed with Conduct Disorder using the forthcoming Diagnostic and
Statistical Manual V (DSM-V) (Frick & Moffitt, 2010). The two factors
that compose psychopathy are strongly positively related, such that
higher levels of callous-unemotional traits predispose an individual to
increased antisocial behaviors, particularly antisocial behavior that
serves an instrumental goal, such as bullying, sexual violence, or assault
during the course of a robbery (Blair, 2001; Woodworth & Porter,2002;
Kahn, Byrd, & Pardini, 2012; Viding, Frick, & Plomin, 2007; Dadds,
Fraser, Frost, & Hawes, 2005).
(p.154) Conclusions
The case of psychopathy presents a strong case that some forms of
moral reasoning rely on intact empathic responses to victims distress,
particularly fear, and therefore are reliant on basic emotional processes.
There are many compelling reasons to focus on the rational basis of
moral judgments (Cima, Tonnaer, & Hauser,2010), but interpreting
psychopaths moral reasoning deficits as primarily rooted in rationality
presents several difficulties. For one, as Nichols has argued (Nichols,
2002a), it is difficult to identify a rational defect that is present in
psychopaths but that is absent in populations (e.g., very young children,
autistic adults) that reliably draw the moral/conventional distinction.
For another, the evidence seems to suggest that psychopathic deficits in
moral judgments are more likely to emerge the more the moral
reasoning task requires the consideration of victims distress,
particularly fear. This phenomenon can be observed both across tasks
and within tasks (e.g., Aharoni. Sinnott-Armstrong, & Kiehl, 2012; Marsh
& Cardinale, 2012). Deficits in responding to others fear in moral
judgment tasks closely parallels findings that the fear system appears to
be generally defective across a variety of neurocognitive paradigms in
psychopaths. Finally, recent neuroimaging research suggests that
psychopaths deficits in both fear processing and moral reasoning are
linked to dysfunction in evolutionarily ancient subcortical structures
like the amygdala, the function of which is primarily affective. This
suggests that the empathic deficits that lead to moral reasoning deficits
in psychopathy emerge from basic affective processes.
These points are among the accumulating evidence that supports the
presence of circumscribed deficits in moral reasoning in psychopathy. In
better understanding the nature of these deficits, including their
neurodevelopmental origins, we may gain an improved understanding
not only of the nature of psychopathy, but of the nature of human
morality.
Are Empathy and Morality Linked?
Insights from Moral Psychology, Social and Decision Neuroscience, and
Philosophy
Giuseppe Ugazio
Jasminka Majdandi
Claus Lamm
Introduction
Empathy is commonly viewed as a necessary condition for moral
behavior in most of the treaties proposed in the history of moral
philosophy (Aristotle/Roger, 2000; Hume, 1777/1960; Smith, 1853),
which has resulted in the widespread belief that empathy and morality
are intimately related. Defining the relationship between empathy and
morality, however, has proven to be difficult for two main reasons. First,
empathy has been defined in many different ways, which makes it hard
to differentiate it from other socio-emotional states, such as compassion
or sympathy (e.g., Batson, 2009b, this volume). Second, evidence on the
causal role of empathy, and of emotions in general, in morality is mixed.
Some scholars indeed maintain that emotions play no role in morality
(Hauser, 2006), while others claim that emotions play a dominant role
in moral judgments (Prinz,2004). Addressing these two issues will allow
us to gain a clearer view of the relationship between empathy and
morality.
(p.156) In this chapter, we will therefore summarize the most
important philosophical approaches to defining morality. We will then
propose a definition of empathy that differentiates it from the emotional
states with which it is often confused. Having laid a theoretical
foundation in the first two sections, in the third section we will discuss,
in light of the existing literature, what role emotions, and more
specifically empathy, most likely play in morality. We will explain that
empathy plays a very important role in morality in two respects. First,
empathy allows humans to understand how others are emotionally
affected by a given action, which can directly inform moral decisions and
actions. In addition, by means of the link between empathy and
compassion (or sympathy), empathy can motivate people to behave in
accordance with moral principlessuch as maximizing the wellbeing of
as many people as possible (Bentham, 1789/1996) or not inflicting harm
or using a person as a means to an end (Kant, 1785/1965). However, we
will argue that, although empathy is an important source of information,
the knowledge acquired via empathy does not directly translate into
moral decisions as, under some circumstances, the morally appropriate
option may be different from the option following from ones empathic
response. For instance, previous results have shown that empathic
responses can reduce the frequency of utilitarian judgments, such as
when one decides to refrain from sacrificing the life of an innocent
person with whom one strongly empathizes in order to save a larger
number of innocent people (Crockett, Clark, Hauser, & Robbins, 2010;
Gleichgerrcht & Young, 2013; Majdandi et al., 2012). This might be
viewed as at odds with the moral judgment prescribed by the utilitarian
school (see Figure 1 below for a schematic illustration of this point).
Furthermore, empathic responses can lead people to express
contradictory judgments depending on whether their decisions regard
ingroup or outgroup members (Cikara, Farnsworth, Harris, & Fiske,
2010). Third, the knowledge acquired through empathy may sometimes
be used to motivate immoral behavior, such as in the case of torture.
1 . What Is Morality?
If we want to find out whether and, if so, how empathy informs
morality, we first need to define morality. The roots of the English
noun morality evolved from the Latin noun morlia and lie in the
Latin mores, which can be literally translated into habits, customs,
or traditions. These are also the nouns that are closest in meaning to
the Ancient Greek thos from which the English ethics originated.
Although these words can be considered synonyms, we should note that
over the centuries ethics has been used to denote the study of the
principles which should be used to establish the (p.157) appropriate
habits for a given community, social group, or professional activity;
morality, instead, has been mostly used in its adjective form, that is, as
a synonym of ethical, denoting the habits that are in accordance with
the principles identified by ethics. In the context of the present chapter,
we will refer to morality as the expression of judgments classifying
behaviors into good/right and bad/wrong ones. Two perspectives can be
taken when studying morality: (a) a normative perspective that
establishes the principles that should be used to decide which behaviors
are good and which are bad and (b) a descriptive point of view that
studies how we decide whether a given behavior is good or bad.
Normative moral theories thus inform us about how we ought to behave
and how we ought to decide which behaviors are right or wrong. More
specifically, normative ethics provides us with the means to discriminate
between right and wrong. However, singling out which behaviors are
right and wrong is a task of practical ethics, a branch of ethics that we
will not discuss in this chapter. From the numerous theories proposed in
the normative moral philosophical literature, two have particular
relevance in the contemporary moral debate: consequentialism and
deontology (Tobler, Kalis, & Kalenscher, 2008). These theories differ
mainly in what they focus on in order to identify the normative
principles. While the principles proposed by consequentialism focus on
the foreseeable consequences of behaviors, deontological principles
specify the requirements that the behaviors need to meet.
More specifically, consequentialism holds that the outcomes
(consequences) of our actions ought to be as good as possible (Scheffler,
1988; Singer, 1974). Consequentialist theories are further distinguished
between act consequentialism and rule consequentialism. According to
the former, the outcome of individual actions ought to be as good as
possible. On the other hand, given that the consequences of individual
actions are sometimes difficult to predict, the latter holds that the
consequences of established action-guiding rules ought to be as good as
possible. Actions are thus evaluated with respect to these rules (see also
Heinzelmann, Ugazio, & Tobler,2012). For example, one of the most
relevant consequentialist theories is utilitarianism: One ought to do
what maximizes the wellbeing of the greatest number of people (or
minimizes their unhappiness).
Deontological theories assign a special role to duties (deontology
refers to the study or science of duty, from the Ancient Greek deon =
duty). Duties are actions that follow one or more principled rules. From
this perspective, the rightness or wrongness of an action is not so much
determined by the goodness or badness of its consequences, but rather
by whether the action itself fulfills the established requirements. For
instance, one of the most popular requirements can be found in Kants
(1785/1965) moral theory, in which the (p.158) author states that one
may never treat other human beings as a means to an end, but only as
an end in themselves (p. 30).
In contrast to normative moral theories, descriptive moral theories seek
to elucidate how a person decides whether a given behavior is right or
wrong. Following David Humes Treatise on Human Nature
(1739/1978), which is one of the most complete attempts to provide a
scientific description of moral judgments, moral philosophers have
diverged into two groups: those who believe that morality is driven
solely by rational considerations (Clarke, 1738; Cudworth, 1996; Kant,
1785/1993) and those who propose that morality is of an emotional
nature (Hume, 1739/1978; Hutcheson, 2002; Prinz, 2004; Cooper, 1999).
Briefly, those who consider morality to be of an emotional nature
suggest that, in order to evaluate the moral appropriateness of an event,
one must base ones judgment on the gut feeling provoked by the
event. If the gut feeling is a pleasant one, then the event is morally
appropriate, but if the gut feeling is unpleasant the event is morally
inappropriate (Hume, 1739/1978). In other words, by paying attention to
ones own feelings, a person can infer whether something is morally
appropriate. In contrast, those who believe that morality is solely a
matter of reasoning claim that evaluating the appropriateness of an
event requires a deliberative reasoning process based on a moral
principle that is purely based on practical reason, in other words, a
principle that rational agents would all agree on (e.g., never treat
humans as means to an end): If an event violates such a principle, it is
morally inappropriate; if an event is in accordance with such a principle,
it is morally appropriateirrespective of the emotions accompanying the
decision (Kant, 1785/1993).
It has recently become problematic to maintain that morality is solely of
either an emotional or a cognitive nature. On the one hand, while we
normally think of moral judgments as beliefs, they are characteristically
motivating, and as Hume notes, judgments of facts alone do not have
the power to move us (Schwartz, 2005, pp. 12). In other words, if
morality were only of a cognitive nature, then moral judgments alone
would lack the motivational aspect that induces a person to act
according to his/her judgments. Prinz (2011a) proposed an interesting
thought example that captures such motivational aspects: Consider the
following two rules that pupils are frequently taught at school: (a) a
conventional rule stating that pupils should raise their hand and wait for
the teacher to call on them before speaking and (b) a moral rule stating
that pupils should not harm other pupils. If a schoolteacher told the
pupils that they could speak whenever they wanted to and no longer
needed to raise their hand and wait to be called on, most of them would
conform to the new norm, speaking whenever they wanted to. However,
if a teacher told the pupils that they could hurt each other, very few of
them would actually do so as moral norms have (p.159) intrinsic
motivational power and do not need an external element (such as an
authority) to be obeyed. Furthermore, a purely rational view of morality
is inconsistent with the recent body of evidence that moral judgments
typically involve strong immediate subjective feelings (Greene,
Sommerville, Nystrom, Darley, & Cohen, 2001; Haidt, 2001; Moll, De
Oliveira-Souza, & Zahn, 2008; Ugazio, Lamm, & Singer,2012).
On the other hand, considering morality to be solely of an emotional
nature would result in denying the ubiquitous emergence and
consolidation of moral knowledge (i.e., sets of moral principles) in
human societies (Schwartz, 2005). Indeed, in order for such moral
regulations to emerge, it is necessary that a group of people reach an
agreement on the moral appropriateness of a given behavior based on
grounds that exceed the level of individual feelings. Founding moral
criteria on formal rules of logic seems to constitute a more widely
accepted common ground than basing them on more erratic, individual
emotions. Due to the conclusion that morality requires both emotional
and rational components, scholars who argued that emotion and
rationality mutually exclude each other in moral judgments ran into the
logical impossibility of maintaining at the same time that moral
knowledge exists and that morality is of a solely emotional nature (R.
Campbell, 2005).
Thus, as R. Campbell (2005) proposes, morality is best considered to
have elements of both reason (or belief) and emotion (and desire)that
is, it can be considered to be a besire (Altham, 1986). From this
perspective, then, moral judgments are considered to be a combination
of beliefs, emotions, and motivations, but sometimes they can also be
solely rational or solely emotional responses to events. In sum,
according to this moral descriptive view, the emotional component of
morality is mainly associated with its motivational aspect, that is, the
force that morality has to motivate a person to act in a certain way, while
the rational component is linked to the capacity of acquiring moral
knowledge, that is, a set of norms that guide our moral judgments (R.
Campbell, 2005).
The dichotomy of philosophical views on morality, that is, whether it is
of an emotional or rational nature, has also been reflected in the mixed
results of scientific attempts to clarify the nature of morality. On the one
hand, some scholars claim that, given the obtained data, morality is
motivated by emotions: Schnall and colleagues found that induced
disgust leads people to express more severe judgments of condemnation
toward certain moral violations (such as incest) than people in a neutral
emotional state (Schnall, Haidt, & Clore, 2008); a similar disgust-
induction effect was found by Wheatley and Haidt (2005) on the same
types of moral scenarios. On the other hand, otherscholars who mostly
analyzed the motivations for the moral considerations expressed by
people claimed the opposite, that is, that morality is of a (p.160) purely
rational nature (Kohlberg, 1976; Piaget, 1932; Rest, Narvaez, Bebeau, &
Thoma, 1999). As we have argued in previous work (Ugazio, Lamm, &
Singer, 2012), it is likely that the origins of the many contradictory
findings lie in the heterogeneity of experimental procedures that have
been adopted. At present, the views that considered morality to be solely
driven by emotional or rational forces are losing strength, as most
scholars now agree that both reason and emotions play an important
role in morality (Cushman, Young, & Greene, 2010; Moll, Zahn, de
Oliveria-Souza, Krueger, & Grafman, 2005; Moll, De Oliveira-Souza, &
Zahn 2008; Ugazio, Lamm, & Singer, 2012). In line with R. Campbells
(2005) dual view of morality, the evidence proposed by moral
psychologists seems to support the theoretical view that moral
judgments result, depending on the circumstances, from a combination
of rational deliberations and emotional responses to an event.
In light of the literature discussed so far, we propose that emotions may
play a crucial role in morality. Being a social emotion, therefore,
empathy may serve as a crucial source of information for a person to
judge which behaviors are morally right or wrong (Bartels, 2008; Haidt,
2001; Nichols, 2002b; Ugazio, Lamm, & Singer,2012). Through this role,
empathy can then trigger associated emotional states, which have the
potential to move people to act in accordance with morally prescribed
behaviors. Since it is important to distinguish between empathy and
other constructs that are also associated with emotional responses (such
as emotional contagion, sympathy, or compassion), the next section
focuses on defining empathy and related terms.
2 . What Is Empathy?
The Anglo-Saxon linguistic roots of the word empathy lie in the
Ancient Greek empatheia (passion), which is composed of en (in) and
pathos (feeling). The term was originally coined by the German
philosopher Theodor Lipps, who used the term Einfhlung (of which the
English word empathy seems to be a direct translation) to describe the
process of understanding works of art. At a basic phenomenological
level, empathy denotes an affective response to the directly perceived,
imagined, or inferred emotional state of another being. To our own
understanding, empathy requires the affective sharing or resonating of
an observer with another persons (the target) affect in an isomorphic
manner. In addition, the observer has to be aware at any point in time
that the source of his or her feelings is the target. This stresses the
central importance of the capacity for self/other distinction, which is the
ability to distinguish between mental and bodily representations related
to the self and to the other (de Vignemont & Singer, 2006; Decety &
Lamm, 2006; Singer & Lamm, 2009). (p.161) Empathy can therefore be
described as a mechanism enabling a (usually impartial) copy (feeling
with) of the targets emotional state by the observer, with full
awareness of which parts are copied and which parts originate in the
observer him- or herself.
This definition, which stresses the role of empathy in gaining
information about the internal affective representations of others,
deviates from the predominant folk psychological definition, namely,
that empathy is an other-oriented or even moral social emotion. In
order to avoid confusion with such definitions, some conceptual
clarifications are needed (see also Batson, 2009b). At least five key
concepts are related to empathy, ranging from motor mimicry to
prosocial or altruistic behavior.
Motor mimicry describes our tendency to automatically synchronize our
movements with those of another person. For instance, considerable
evidence suggests that perceiving a targets affective facial expressions
activates the corresponding facial muscles in the observer (for a review,
see Dimberg & Oehman, 1996), and the strength of such mimicry
responses correlates with self-report questionnaire measures of
empathic skills (Sonnby-Borgstrom, 2002). Notably, though, this
correlation is rather weak, indicating that such bottom-up resonance
mechanisms are only one aspect of empathy. In addition, recent
accounts contest the automaticity of human mimicry and propose that it
acts as a social signal (Hess & Fischer, 2013). We propose that motor
mimicry might subserve both resonant and signal functions and support
a virtuous circle sustaining smooth social interactions (Heyes,
forthcoming).
Emotional contagion is another phenomenon that is strongly relevant to
yet clearly distinct from empathy (Hatfield, Cacioppo, & Rapson, 1994).
It denotes the tendency to catch other peoples emotions and has also
been labeled primitive empathy (Hatfield, Rapson, & Le, 2008) or
affective empathy (de Waal, 2008). Notably, a few days after birth,
human newborns already start crying in response to the distress calls of
other babies. To turn this contagious response into a full-blown
empathic response requires the development of a sense of self, however,
since experiencing empathy requires the awareness that the source of
the feeling state is the other, not the self. This sense emerges around the
age of about 12 months (Hoffman, 2000). Taken together, motor
mimicry and emotional contagion may in many instances be important
antecedents of empathy, but in general should neither be regarded as
necessary nor as sufficient processes for the experience of empathy.
With respect to the consequences of vicarious emotional responses,
empathy as defined here needs to be separated from sympathy, empathic
concern, and compassion. While all four terms include affective changes
in an observer in response to the affective state of another person, only
the experience of (p.162) empathy entails vicarious responses that are
not modified by the observer (in the sense of the copied state or
feeling with referred to above). In contrast, sympathy, empathic
concern, and compassion carry additional feeling for processes
attributed to the observer. For example, in the case of empathy,
observing the sadness of another person is associated with a partial
feeling of sadness in the observer. Sympathy, empathic concern, and
compassion, however, are characterized by additional feelings, such as
concern about the targets welfare or the wish to alleviate his or her
suffering. These processes are the outcome of the interaction between
observer and target, but go beyond what the target is actually feeling.
The main distinction between empathy and phenomena like sympathy,
empathic concern, and compassion is therefore whether the observers
emotions are inherently other-oriented (feeling for; compassion,
sympathy, empathic concern) or whether they reflect affective sharing in
the sense of feeling with (empathy) the other person.
Finally, many accounts of empathy, broadly defined (Batson, 1991; de
Waal, 2008), relate its occurrence to prosocial, other-oriented
motivations (i.e., a motivation to increase the other persons wellbeing
or welfare or to forego selfish, self-related benefits for the benefit of
others). This is not necessarily the case when empathy is defined as
feeling with another person. Empathy as understood this way simply
enables us to feel as accurately as possible what others are feeling,
without any sort of valuation attached to these feelings. Whether this
then has prosocial, antisocial, or neutral consequences is the result of
other variables, including other social emotions (such as envy or guilt),
as well as acquired behavioral tendencies, moral values, or the personal
relationship between observer and target (which if competitive can even
result in counter-empathy; e.g., Lanzetta & Englis,1989; Yamada, Lamm,
& Decety, 2011). Notably, while consistent evidence for the link between
feeling for (empathic concern, compassion) and prosocial behavior
exists (e.g., Batson, 1991; Eisenberg, 2000b; Eisenberg et al., 1989a), a
clear-cut empirical demonstration of a link between empathy as feeling
with and prosocial or moral decision-making is still missing.
In terms of the neural foundations, which have received increased
attention in the last few years, the definition of empathy as a shared
feeling state has received widespread support. For example, two recent
meta-analyses of functional neuroimaging studies unequivocally
demonstrated that witnessing others suffering engages a neural network
indicating that the observer is in an emotional state him- or herself
(Fan, Duncan, de Greck, Northoff,2011; Lamm, Decety, & Singer et al.,
2011). This network includes the anterior insular cortex and the medial
cingulate cortex (MCC), two brain areas that constitute an intrinsically
linked network involved in emotional(p.163) awareness and the
homeostatic regulation of physiological bodily responses associated with
emotional responses (Lamm & Singer, 2010). Of note for the present
discussion, by means of its connections to output centers of the brain,
the MCC in particular is able to provide a rather direct link between
emotional and behavioral responses, which enables the organism to
maintain homeostasis.
In addition, it is important to point out the distinction between self- and
other-related emotional responses resulting from the observation of
distress in others (e.g., Batson, Fultz, & Schoenrade, 1997; Lamm,
Batson, & Decety, 2007). Witnessing negative emotions in others can
result in empathic concern, which is an other-related vicarious response.
This might in turn increase altruistic motivation toward the person
empathized with and thus motivate prosocial behavior (Batson, Early, &
Salvarini, 1997). In contrast, harm to others might also evoke personal
distress, which is a self- rather than other-focused vicarious emotion.
Rather than motivating a prosocial act, personal distress might increase
ones tendency to escape the distressful situation. Alternatively, if a
prosocial act is performed, it might result from essentially selfish
motives, that is, be an attempt to reduce ones own distress by
eliminating the source of the distress (i.e., the other persons suffering;
e.g., Cialdini et al.,1987; Maner et al., 2002). Hence, decisions to help
others that are motivated by vicarious responding can stem from other-
related or from self-related motives. Whether such decisions can be
considered moral decisions strongly depends on the relevance one
assigns to the motives for the decision: On the one hand, if motives are
not among the primary determinants of whether a decision is moral or
not (e.g., if we only care about the consequences of a decision), then
decisions motivated by self-related motives can also be considered
moral. However, if motives or intentions play a central role in
determining whether a decision is moral or not (e.g., as for Kant,
1785/1965), then selfish motives can never be moral motives. Hence,
decisions motivated by self-related motives would not be considered
moral. From Kants perspective, decisions resulting from selfish motives
should not be considered moral as the given motives cannot be
universalized and ultimately come to constitute a moral principle (Kant,
1785/1965). For example, imagine a bystander who decides to risk his
life to save a drowning person because he could not bear the distress of
witnessing the death of a person (or, due to a more trivial selfish motive,
such as because the drowning person owed him a large sum of money).
From the two perspectives discussed previously, if we only care about
consequences, this decision has to be considered a moral decision; from
the second point of view, however, this decision can be considered a
good one, but not a moral one.
Conclusion
In this chapter, we have tried to shed light on the relationship between
empathy and morality. In the first two sections, we defined and
contextualized morality and empathy, respectively, in order to identify
some of the potential connections between the two. In the resulting
theoretical framework, we identified an epistemological and a
motivational role of empathy in morality, but also pointed out that
empathy cannot be considered a necessary condition for morality.
Neither the epistemological nor the motivational aspects of empathy
align themselves specifically with judgments or motivations that are
morally right. We propose that empathy contributes to moral judgments
by providing information about the emotional reactions of people
affected by an action and by motivating a person to act in a certain way.
Whether these decisions are in accordance with moral principles
depends on the contextual circumstances in which an agent finds
himself or herself. In sum, these views point to a much more complex
link between empathy and morality than the one suggested by the
widely held folk belief that empathy is closely and directly linked to all
aspects of morality.
On Empathy
A Perspective from Developmental Psychopathology
R. Peter Hobson
Jessica A. Hobson
Introduction
Over the past two decades, and across the disciplines of philosophy,
developmental psychology, and cognitive neuroscience, there has been a
resurgence of interest in the nature of human beings psychological
connectedness with each other, and alongside this, debate over the basis
for young childrens understanding of peoples minds. For many of those
caught up in the intellectual maelstrom, as well as for many more who
catch news from afar that something momentous is being deliberated,
there appears to be a relatively clear-cut option: either people are
connected with, and understand, other individuals by simulating their
mental states, or people need to theorize about minds. True, there are
so-called hybrid theories that encompass certain features of each
approach, but these inherit the intellectual restrictions and
preoccupationsand in particular, a way of thinking about the gulf
between one persons mind and that of anotherthat characterize a
stand-off between two dominant schools of thought.
At what point does empathy enter the fray? Empathy is an especially
interesting case for the study of how young children develop
interpersonal relations (p.173) and come to understand the mind. One
reason is that to many people it seems obvious that if an individual is to
have an empathic emotional response to someone elses suffering (for
example), then that individual must already have a sophisticated
conceptual understanding of what it means to be a self with a subjective
orientation. After all, the one person needs to figure out or imagine what
the other is feeling, perhaps on the model of what the observer is prone
to feel under similar circumstances. Opposed to such a view are
arguments that one could not acquire an understanding of other people
with minds unless one could already relate to them in a manner that is
empathic (Hamlyn, 1974; Hobson, 1991). If one were to apprehend
people (or more accurately, peoples bodies) like things, for example,
there would be little to justify the ascription of subjective states to these
bodies, even if, implausibly, one could conceptualize such states all by
oneself in order to do so (see Hobson, 1991, for further arguments
against a simulationist view). The resolution of these conflicting
perspectives over empathy might have far-reaching consequences for
our view of the development of social cognition and morality.
There is a second reason why the study of empathy could help us clarify
the nature of interpersonal understanding and what it means to hold a
moral stance in relation to others. Let us take it that our having empathy
for someone else reflects our grasp that the person has a subjective
orientation of his/her own. Even if one acknowledges that this grasp is
partly intellectual/cognitive in nature, clearly it is not simply cognitive,
because we have feelings about and/or in relation to that persons state
of mind. It matters to us when we witness a person suffering. Not only
this, but we are inclined to do something about the state of affairs. To be
sure, what we are inclined to do varies from case to caseconsider the
friend who seeks to comfort, the surgeon who prepares his team to
operate, and the torturer who racks up the painbut in each case, the
other persons suffering motivates us to act, and to act in relation to the
state of suffering. Does this mean we need a developmental account that
traces how empathy is constructed out of cognitive (thought), affective
(feeling), and conative (will) components? Or to the contrary, should the
nature of empathy prompt us to rethink the justification for dividing up
the psychological domain in this way? Here we offer an account that
eschews the division of empathy into cognitive and affective
varieties, and entails a reorientation toward empathy as an evolving
mode of interpersonal relatedness with cognitive and affective aspects.
In addressing these issues, we shall adopt the stance of developmental
psychopathology. We shall consider early human development, in order
to reflect upon the structure of interpersonal experience that empathy
entails. We shall compare and contrast typical development with a case
of atypical development, namely that of early childhood autism, in order
to give an extra dimension to these considerations. Through the study of
a condition where empathy (p.174) is seriously compromised, we may
acquire insight into the nature of empathy itself. More than this, we
might discover that what is often portrayed as a rather cool
understanding of disembodied minds is actually founded upon
dynamic and affectively charged relations between embodied persons,
relations that are vividly exemplified by empathy.
A final feature of our approach is that we shall consider empathy
alongside other modes of human relatedness and thinking. Our aim here
is to remind us how empathy is but one among diverse forms of social
engagement that implicate human-specific modes of self-other
connectedness and differentiation. Partly for this reason, we do not
consider it worthwhile to dwell on the distinction between, say, empathy
and sympathy. Suffice it to say that we are taking empathy to encompass
a class of personal relations that provide a basis for human beings to
experience persons as persons and that establish the foundations for
what will become conceptual understandings of people with minds.
Having sketched out some of the theoretical issues, we shall dive
straight in to some empirical research on autism.
Studies in Autism
Autism is a syndrome. A syndrome is simply a cluster of clinical features
that tend to occur together. Children with autism have profound
impairments in social relatedness and both nonverbal and verbal
communication, as well as a tendency toward rigid and repetitive forms
of activity and thinking.
We begin with evidence that autism involves a severe restriction in the
childrens empathic relations toward other people. Our intention is not
so much to illustrate that autism involves abnormality in this respect.
Rather, we want to see if we can begin to specify in what the restriction
consists, in order that this might inform our view of what empathy
entails.
We have not space to dwell on clinical observations on autism, although
Kanner (1943) gives fine descriptions of what he called the childrens
impaired affective contact with others. However, we shall provide a brief
illustration of what such impairment might mean for affected
individuals experience of other people. Here is what an intelligent
young autistic adult said when interviewed by the psychiatrist Donald
Cohen (1980). This man described how the first years of his life were
devoid of people:
I really didnt know there were people until I was seven years old. I
then suddenly realised that there were people. But not like you do. I
still have to remind myself that there are people . . . I never could
have a friend. I really dont know what to do with other people,
really.
(p.175) Now let us turn to a classic quasi-experimental study by the
UCLA team of Sigman, Kasari, Kwon, and Yirmiya (Sigman, Kasari,
Kwon, & Yirmiya, 1992; also Charman, Swettenham, Baron-Cohen, Cox,
Baird, & Drew, 1997). These researchers tested 30 young autistic
children with a mean age of under four years, along with closely
matched children without autism. The approach was to code these
childrens behavior when an adult pretended to hurt herself by hitting
her finger with a hammer, simulated fear toward a remote-controlled
robot, and pretended to be ill by lying down on a couch for a minute,
feigning discomfort.
In each of these situations, children with autism were unusual in rarely
looking at or relating to the adult. When the adult pretended to be hurt,
for example, children with autism often appeared unconcerned and
continued to play with toys. When a small remote-controlled robot
moved toward the child and stopped about four feet away, the parent
and the experimenter, who were both seated nearby, made fearful facial
expressions, gestures and vocalizations for 30 seconds. Almost all the
children without autism looked at an adult at some point during this
period, but fewer than half the children with autism did so, and then
only briefly. The children with autism were not only less hesitant than
the mentally retarded children in playing with the robot, but they also
played with it for substantially longer periods of time.
These observations extend beyond person-with-person responsiveness,
insofar as the children with autism were less influenced by the fearful
attitudes of those around them when it came to their behavior toward
the robot. They appeared to be relatively unengaged not only in their
one-to-one interpersonal-affective transactions, but also in relation to
another persons emotional attitudes toward objects and events in the
environment. They were not gripped by the others plight, nor moved to
adopt the others affective stance toward a shared world. They were less
drawn toward the stance of the other personor as we have expressed
this elsewhere (Hobson, Chidambi, Lee, & Meyer, 2006), less other-
person-centeredthan were the children without autism.
From the starting-point of this quasi-experimental study, let us move in
two directions. First, we turn to a tightly controlled experiment, because
experiments are especially useful in determining the specificity of
abnormalities such as that of childrens unresponsiveness to expressions
of feeling. Moore, Hobson, and Lee (1997) tested children and
adolescents with and without autism, matched for age and verbal ability,
and showed them videotape sequences of peoples moving bodies
depicted merely by dots of light attached to the trunk and limbs. First we
presented separate five-second sequences of the point-light person
enacting in turn the gestures of surprise, sadness, fear, anger, and
happiness (each of which could be recognized with very high
(p.176) reliability by nave adult raters). In the surprise sequence, for
example, the person walked forward and suddenly checked his stride and
jerked backward with his arms thrown out to the side; in the sad
sequence, the person walked forward with a stooped posture, paused,
and sighed. The children were told: Youre going to see some bits of
film of a person moving. I want you to tell me about this person. Tell me
whats happening.
In response to this request, all but one of the children without autism
made a spontaneous comment about the persons emotional state for at
least one out of the five presentations, and most referred to emotions on
two or more of the sequences. In contrast, 10 of the 13 children with
autism never referred to emotional states, whether correctly or
incorrectly. In the case of the children and adolescents with autism, it
was the persons movements and actions rather than feelings that were
reported. For example, the sad figure was described as walking and
sitting down on a chair, walking and flapping arms and bent down,
and walking and waving his arms and kneeling down . . . hands to face.
Almost none of these responses were wrong, but very few referred to the
depicted persons feelings.
It was not that the children with autism were unable to interpret what
they saw. They did so very well and conveyed this in complex
psychological terms that captured the peoples actions. They were
distinctive in failing to report on the subjective experience of the
depicted figures.
Now let us move in a second direction, toward real-life descriptions of
how children with autism relate to others. Colleagues and ourselves
(Hobson, Chidambi, Lee, & Meyer, 2006) conducted semi-structured
interviews with parents of children with autism, and children without
autism of similar age (613 years) and verbal mental age (3.59 years).
Most of the questions concerned whether the children showed social
emotions such as jealousy, guilt, and concern. We enquired after specific
instances of each emotion. For example, the question about jealousy
was: Have you observed jealousy in your childthat is, resenting the
attention you or someone else is giving to other individuals?
Parents of both groups of children reported that their offspring showed
feelings such as happiness, distress, and anger (although we did not
enquire closely on the person-directedness of the anger). They also
reported that their children were affected by the moods of other people,
and here it was clear that the children with autism were not globally
unresponsive. Nor was it the case that all forms of differentiated
relatedness were absent. In particular, the groups were almost identical
insofar as the majority of children with as well as without autism
showed clear signs of jealousy. Indeed, of the only two parents who
thought that their children with autism did not show jealousy, one was
our only poor respondent, and the other was far from confident about
(p.177) the matter. Here is an example of what one parent said about
her child with autism:
I:
He doesnt like S (partner) and me hugging or holding hands
sometimes . . . When he was very tiny, like two, he was very
jealous of us I think. He didnt like us sitting next to each other
or hugging. I remember one occasion when he actually led you
[to partner] to the door and shut the door.
On the other hand, when parents were asked about their childrens
emotions of pity, concern, and guilt, there were marked group
differences. A majority of children without autism were said to show
clear manifestations of these feelings. In the case of the children with
autism, by contrast, a majority showed possible or atypical signs of pity
and concern, but only one was reported to show clear instances of these
feelings. For instance, here are two parents describing their children
with autism:
PARENT 1:
When it comes to concern for feelings of others, if he was told
we were upset, perhaps hed be concerned. Im not sure he
would be able to pick it up very easily. He might actually find it
quite hard to deal with as he finds it hard when people are
upset. He might actually insist we stop. I dont imagine he
would like it. He might be worried but he doesnt have that
empathy sort of concernhe doesnt show that at all . . .
Empathetic sadness isnt there.
PARENT 2:
When Im sad, it disturbs him, he doesnt quite know what to
do and then he just looks and if I dont say anything, he just
moves away. A normal child would ask or say what is
happening, he wouldnt.
Again, it is not the case that the children were unresponsive. Rather, it
was in the organization of their behavioral and expressive relatedness
and some theorists might become exercised over whether this is in the
organization of their thinking, feeling, or motivationthat the children
with autism were said to be atypical. To repeat: few parents were able to
report that their children with autism showed clear instances of other-
person-centered emotions such as guilt, pity, or empathic concern for
someone else, nor shame or embarrassment before another person.
We trust these observations illustrate the vital links between feeling for
others and moral attitudes such as guilt and concern, as well as behavior
that expresses such an orientation to other human beings. These reports
from parents are complemented by what may be gleaned from self-
reports given by verbally fluent children and adolescents with autism.
For example, Kasari, Chamberlain, and Bauminger (2001) described how
high-IQ children with (p.178) autism reported feeling guilt, but only
14% participants with autism (versus 42% of those with typical
development) spoke of guilt over physical harm to others, and none
referred to emotional harm such as hurting someones feelings. Instead
they were more likely (73% of instances) to describe situations of rule-
breaking, disruptiveness, or property damage. In the case of
embarrassment, fewer participants with autism explicitly mentioned an
audience (also Capps, Yirmiya, & Sigman, 1992).
Now if children with autism tend to show less concern than children
without autism, then does this amount to more than a failure to perceive
and/or respond to expressions of emotion? Or is there a more far-
reaching limitation in the childrens propensity to experience and
orientate to other persons as centres of subjectivity?
Consider the following study (J.A. Hobson, Harris, Garca-Prez, & R.P.
Hobson, 2009). Sixteen school-age children with autism and 16 children
without autism of similar age and verbal ability took part. The children
were between the ages of eight and sixteen years, with a mean verbal
mental age of about seven years. There were two adult testers who sat
around a table with a participant and played a game in which they each
drew an animal of their choice. Then in a standardized, slow-paced
procedure, one tester proceeded to tear up the drawing of the other
tester. The tester whose drawing was torn did not show any overt
emotional reaction to the event, although she did witness its occurrence.
Therefore it could not be the case that an observable emotional display
played a role in triggering participants responses. In a control condition,
a blank piece of paper was torn instead of a picture.
Videotapes of the episodes were given to two raters who were asked to
find each look to the tester whose drawing was torn and then evaluate
which of those looks expressed concern. These were looks in which the
child appeared to become involved with the tester whose drawing was
torn, apparently taking on her psychological stance (becoming upset on
her behalf), experiencing concern for her feelings, or showing a sense of
discomfort about her position (e.g., through nervous laughter). The
raters had excellent agreement on the quality of such looks.
The results were that when the blank index card was torn, the children
rarely looked at the tester seated across the table. When it was the
testers drawing that was torn, however, some of the children with
autism, but especially those without autism, looked at her during or
immediately after the event. More importantly, while on the blank
drawing condition only one child (a child without autism) ever showed
a concerned lookand only onceon the tear drawing condition, ten
out of sixteen children without autism showed between one and six
concerned looks, while only three out of (p.179) sixteen children with
autism ever showed a concerned look. We should add that this group
difference was not confined to differences in quality of looks, because
other expressions of concern were relatively lacking among the children
with autism.
Any interpretation of the results needs to account for the speed with
which, as well as the feeling with which, participants without autism
looked to the tester whose drawing was torn. One might also take into
account how charged an atmosphere was generated by the procedure.
Empathy can be very powerfulalbeit not, it seemed, for most of these
participants with autism.
Beyond Empathy
Now we turn to some research that may help us to see what it is that is
missing in the ill-organized and diminished empathic responsiveness of
children with autism. We are hoping that the studies we shall report give
substance to the claim that the children are limited in the propensity to
identify with the attitudes of other peoplea capacity we take to be
critical in the development of a moral sense toward other feeling human
beings.
To introduce this idea, let us return to a theoretical point: empathy
means responding to the other persons feelings as the others feelings.
The feelings involved in an empathic response are both ones own and
experienced in relation to the subjective state of the other. In what sense
is the others subjective state felt?
One way to approach this question is to consider what it means to
identify with someone else. The important thing about identification is
precisely that one feels in accordance with the other, but one does not
entirely become the other. The other persons feelings-as-experienced
are part of ones own complex response, yet these are still partitioned
off, as it were, within that response. One implication is that the other-
person-anchored part of the experience can be relived. It can become a
part of ones own repertoire, both in relation to the world and in relation
to oneself.
Now if human forms of empathy entail identification, in the sense of a
specially powerful connectedness through involvement with the others
actions and attitudes, then perhaps we should revisit autism to see if
there is evidence for weakness in the propensity to identify with others.
This issue becomes even more pressing once one appreciates that it is in
being moved to the emotional stance of others, and therefore in
adopting alternative perspectives through others, that children with
autism are especially handicapped. We have already seen some
implications for social referencing (from the study of reactions to
(p.180) a toy robot), and there are further repercussions for joint
attention, symbolic functioning, and language (e.g., Hobson, 2002/4).
We shall cite three research studies, very briefly, to illustrate pertinent
findings.
(a) Imitation
Hobson and Lee (1999) tested matched groups of children with and
without autism for their ability to imitate a person demonstrating four
novel goal-directed actions on objects in two contrasting styles, which
in most cases meant executing the actions either harshly or gently. The
children with autism copied the goal-directed aspect of the actions, but
showed marked divergence from the control group insofar as very few
adopted the demonstrators style of acting upon the objects involved.
We believe that this reveals a distinction between childrens ability to
observe and copy actions per se, relatively intact in autism (and here you
may recall autistic childrens ability to recognize actions but not
attitudes in videotaped point-light displays of humans gestures), and
the propensity to identify with and thereby imitate apersons expressive
mode of relating to the world, something that is relatively lacking in
autism.
There was a further finding from this study. In one condition, the
investigator demonstrated strumming a stick against a pipe-rack held
against his own shoulder. What happened when the children without
autism copied this action is that a substantial majority identified with
the demonstrator and positioned the pipe-rack against their own
shoulders before they strummed it with the stick. By contrast, few of the
children with autism made this adjustment: most positioned the pipe-
rack on the table directly in front of them. Therefore not only with
respect to style, but also with respect to self-orientation, the children
with autism did not assume the manner with which the other person
executed actions, even though they copied the actions per se.
In our view, these results were not merely an index of imitative styles
that followed perception; rather, they rendered explicit what the
perception entailed in terms of registering and assimilating the stance of
the person demonstrating the actions. Indeed, we interpreted the
findings as reflecting how children with autism have a relative ability to
copy (as well as perceive) simple goal-directed actions on objects, but a
reduced propensity toidentify with the person whose actions those were.
As in the dots of light study, children with autism seemed to view the
actions from the outside, rather than getting beneath the skin of the
person they observed.
(p.181) We have conducted more recent studies that have confirmed
group differences in the imitation of style and self-orientation (J. A.
Hobson & Hobson, 2007; Hobson & Hobson, 2008; Hobson, Lee, &
Hobson, 2007; Meyer & Hobson, 2004). It is intriguing that here, in
contexts where the emotional quality of the task appears to be minimal,
but where self/other role reversals appear to play a critical role in
determining participants responses, children with autism are
distinctive. This raises the possibility that in typical development, the
self/other structure of empathic engagement derives from something
more general, namely the organisation of identifying-with.
(c) Conversation
One of the things that happens in a conversation is that each speaker
tends to pick up features of the other persons language when framing
their subsequent responses. This is more than a kind of echoing,
because each conversational partner builds upon what he or she adopts
(and often adapts) from the other person. Here is an example from a
child with autism talking to an interviewer:
I:
What are you good at?
P:
I am good at, eh, science.
In a collaborative study (Du Bois, Hobson, & Hobson, 2014; Hobson,
Hobson, Garca-Prez, & Du Bois, 2012), we studied conversational
linkage between an (p.182) adult and matched participants with and
without autism. Our principal prediction was that failures to build on
what is adopted from the conversational partner would be different in
the two groups. It turned out that although children with autism picked
up linguistic forms from their conversational partner, they were (as
predicted) significantly more likely to follow this with incoherent,
truncated, vague, or nonresponsive elaboration. Here is one example:
i: And tell me things about yourself that you dont like.
p: That I dont like.
Why did we predict that such abnormalities would occur more often in
the conversations involving children with autism? Because here, in the
linguistic domain, is something closely akin to identifying-with the
stance of the other that we have already described in the domains of
empathy, imitation, and self-other coordination of nonverbal
communication. Critically, as Freud (1955/1921) remarked,
identification involves not just imitation but assimilation, a making of
ones own such that what is assimilated can become foundational for
what follows. Our results confirmed that in conversation, children with
autism are less drawn into adopting the stance of the other (as
linguistically expressed), and to construct their succeeding utterances
upon this basis.
Back to Theory
From developmental and epistemological viewpoints, the crux of the
matter is this: Does having and showing empathy require conceptual
understanding of the nature of other people (and perhaps the self) as
beings with subjective experiences and minds of their own? Or is such
conceptual understanding a developmental achievement founded upon
earlier and more basic forms of affective responsiveness in which
humans register the otherness of the other (Hobson, 1993a, b; Hobson
et al., 2006; Hoffman, 1984b)?
Within the literature on typically developing young children, as
Thompson (1987) points out, the dominance of cognitive-developmental
perspectives on emotional awareness and responsiveness, coupled with
an emphasis on relatively detached and intellectually demanding
methods to assess empathy, may have underestimated infants
capacities for feeling toward others who are apprehended, but not
conceptualized, as separate beings. Although someone who empathizes
has the ability to register and sense self/other differentiation, this does
not necessarily entail that he or she conceptualizes this distinction, nor
that imaginative role-taking (or so-called cognitive empathy) is
required (p.183) for such responsiveness. The appropriate way to
characterize empathy is up for grabs.
How much might cognitive perspective taking accomplish by way of
empathic concern, if the perspective-taker did not bring to the situation
a background of affective engagement with others? Consider the
philosophical account of Goldie (2000), which may be taken as an
exemplifying an approach that distills some of the potential value of a
cognitive theory while avoiding many of its pitfalls. Goldie himself
stresses how there is a nonderived intentional element to feelings
(feeling towards), and criticizes the over-intellectualization of
emotion in contemporary philosophy (p. 11). Despite this, Goldie
considers that our abilities to empathize, to imagine ourselves in
another persons shoes, and to sympathize themselves presuppose some
degree of understanding (pp. 177178). Not only this, but in order to
understand and explain another persons emotion, Goldie claims, it
does not require that any emotion be felt by the interpreter (p. 181,
Goldies italics).
Here Goldie is writing about imaginative empathy among adults and is
not concerned with developmental issues. Yet from developmental and
epistemological standpoints, it is critical whether or not there needs to
be abackground of feeling for other people if one is to understand and
explain another persons emotion. If Martians were to have no feelings
for other Martians nor anyone else, how far would they understand what
it is for someone to have an emotion or be in an emotional state?
Probably, not far at all. First, they would not understand what it is to be
a someone, because understanding persons is grounded in the kinds of
relations we experience with persons, and those relations are based on
feelings (Hamlyn, 1974); second, they would not understand what it is to
have a person-centered subjective orientation (Hobson, 1993a); and
third, they would not have the kinds of socially derived cognitive
architecture to understand in the relevant manner (Hobson,1993b).
The quotation from Goldie includes the claim that sympathy also
presupposes understanding. He considers that sympathy is distinct from
the imaginative processes of empathy: It is, I think, best understood as
a sort of emotion, involving thought about and feelings towards the
difficulties of another, motivations to alleviate those difficulties where
possible, and characteristic facial expressions and expressive actions
(Goldie, 2000, p. 9). Working up a head of steam, Goldie moves toward a
climax: It is entirely mistaken to assume that in addition to this
recognition of, feeling towards, and response to anothers difficulties,
sympathy also involves undergoing difficulties and having feelings of the
same sort as the other persons . . . your feelings involve caring about
the others suffering, not sharing them (p. 214, Goldies italics). Good
point, and nicely expressed. We shall return to sorts of emotion in due
course.
(p.184) Meanwhile, one might balance Goldies account with that of
someone who defends simulationist views, but whose writings seem to
relinquish central tenets of simulationism. Stueber (2006) specifically
rejects a detached conception of simulation and the invocation of
analogical inferences from self to other. Instead, Stueber suggests the
following: Mechanisms of basic empathy have to be understood as
mechanisms that underlie our theoretically unmediated quasi-
perceptual ability to recognize other creatures directly as minded
creatures and to recognize them implicitly as creatures that are
fundamentally like us (p. 20). Not only this, but Stueber insists that one
does not start from a position of detecting or interpreting others as like
me, but rather, I understand my subjectivity as a moment of
interpersonal intersubjectivity (p. 143).
One wonders, then, how Stueber places himself among empathy
theorists who claim that in learning of other minds we proceed
essentially in an egocentric manner . . . My finding out about another
persons mind depends on using myself and my own mind as a standard
or model for the other persons mind. In particular, proponents of the
empathy view claim that I gain knowledge of other minds primarily
because I can simulate or imitate others mental processes in my own
mind (pp. 34). Perhaps the reason is that Stueber is most concerned
with a specific mode of re-enactive empathy, a personal-level process
central to our understanding of others as agents. His thesis is as follows:
Only insofar as I treat her thoughts as thoughts that could be my own . .
. can I grasp them as her thoughts and as thoughts that constitute her
reasons for her action (p. 165). Here we find vestiges of a simulationist
stance, yet much of the original theory appears to have been jettisoned.
At this point we can turn to phenomenological perspectives.
Phenomenology gives us conceptual tools to loosen the shackles of
prejudice that bind us to a very questionable view of the human
condition, namely that we need to theorize or analogize if we are to
understand the minds of our fellow human beings. Writers such as
Scheler (1954) and Merleau-Ponty (1964) point out how it is simply not
so, because we perceive feelings in the bodily expressions and behavior
of other people. More than this, our mode of person perception is such
that we become engaged with the persons whom we perceive. Merleau-
Ponty (1964, p. 146) suggests that Sympathy . . . is the simple fact that I
live in the facial expressions of the other, as I feel him living in mine.
Elaborating further, he writes:
In perceiving the other, my body and his are coupled, resulting a
sort of action which pairs them [action deux]. This conduct which
I am able only to see, I live somehow from a distance. I make it
mine; I recover [reprendre] it or comprehend it . . . Mimesis is the
ensnaring of me by the other, the invasion of me by the other; it is
that attitude whereby I assume the gestures, (p.185) the conducts,
the favorite words, the ways of doing things of those whom I
confront . . . [it] is the power of assuming conducts or facial
expressions as my own. . . . (Merleau-Ponty, 1964, pp. 118 and 145)
It is notable how Merleau-Ponty uses the words live and living in
these quotations. One persons perception of another entails a rich form
of intersubjective involvement. From this theoretical starting-point, we
encounter no mystery when addressing how one persons non-
conceptually-mediated relations with others provide the basis for that
person coming to understand the nature of people-with-minds. As long
as we can explain how a child comes to acquire the conceptual
equipment to think about embodied persons as having mental states, it
will be natural for those states-as-conceptualized to be ascribed to
appropriate targets, namely persons or person-like creatures (or
occasionally, things). This will be the case even when we are considering
mental states that do not necessarily find overt expression, such as a
persons beliefs.
This does not mean that all the philosophical problems are solved, of
course. But the challenges facing a developmental account are no longer
those explaining how the gap between one persons experiences and
those of another are bridged. Rather, the tasks become those of
explicating the structure of self-other relations, and then explaining how
on the basis of such relations, children develop (and adults fluently
apply) concepts of mind.
On the question of empathy, there are subtle controversies within as
well as beyond the domain of phenomenology. Zahavi (2010) has
discussed points of agreement and disagreement among Scheler, Stein,
and Husserl. For each of these thinkers, empathy is a basic, irreducible
form of intentionality that is directed toward the experiences of others.
They reject the view that imitation, emotional contagion, or mimicry is
the paradigm of empathy. Instead, empathy is like perception in being
direct, unmediated, and noninferential. Yet Zahavi also contrasts the
views of Scheler, who argues that emotional states are given in
expressive phenomena so that we are directly acquainted with anothers
feelings, with those of Stein and Husserl, who stress that another
persons experiences cannot be given to me in the same way as my own
experiences and that the empathized experience is located in the other.
As Zahavi argues, however, these differences may not be so substantial,
if there are ways of experiencing (rather than imagining, simulating, or
theorizing) anothers subjectivity that is not the same as having first-
person experience, but is no less valid as a primary mode of experience.
And this is critical.
Phenomenological views on the direct and unmediated quality of
interpersonal experience does not (of course) entail that
phenomenologists eschew attempts to determine subpersonal
mechanisms that underlie such experience, whether at a psychological
or neurological level. The claim is that whatever (p.186) form these
mechanisms might take, at root they do not involve processes such as
inferring mental states from the perception of mindless bodies, nor the
projection of feelings based upon ones own experiences.
I want to highlight two features of phenomenological discussions that
may be especially worth bearing in mind.
First, in an interesting review of alternative phenomenological
perspectives on empathy, Zahavi (2001, p 163) elaborates on Merleau-
Ponty thus:
Since intersubjectivity is in fact possible, there must exist a bridge
between my self-acquaintance and my acquaintance with others; my
experience of my own subjectivity must contain an anticipation of
the other, must contain the seeds of alterity . . . Thus, Merleau-
Ponty can describe embodied self-awareness as a presentiment of
the other.
What this means is that, for all the attention we should give to
someones experience in actual face-to-face interpersonal encounters,
we should also consider what the individualand from a developmental
perspective, the individual in question may be an infantbrings to such
encounters to give self-other structure to such experience (also Brten,
1998, on the virtual other).
Second, there is the issue of role-taking (broadly conceived), as this
features in many aspects of interpersonal relatedness. Merleau-Ponty
(1964) cites the psychoanalytic notion of identification in a passage
where he reflects on the emergence of language in what he refers to as
the childs affective environment. He describes how a child assimilates
the attitudes of his mother and continues thus: To learn to speak is to
learn to play a series of roles, to assume a series of conducts or linguistic
gestures (p. 109, Merleau-Pontys italics). Not only in language but also
in other spheres of communication, there is an intimate relation
between connecting with others and being moved into new orientations
and stances vis--vis the world.
Identifying-with Revisited
How does an account invoking a biologically based process of identifying
with the attitudes of others square with these considerations?
We have seen that research in autism has yielded evidence that, in
situations that range from empathy to imitation and from nonverbal
communication to conversation, affected children seem restricted in the
organization of their social behavior and experience. This restriction is
of a specific kind. In particular, children with autism are not gripped by
the expressions of other people, they are not so powerfully moved to
adopt the stance of the other in (p.187) relation to a shared world, and
their communication is often lacking flexibility and role-responsiveness.
In concert with all this, by the way, they also have specific limitations in
self-awareness (see Hobson et al., 2006). In short: individuals with
autism have a weakened propensity to identify with the attitudes of
others.
I hope that the studies of autism I have described might help to anchor
what may otherwise seem a rather abstract characterization of
identifying-with in what follows.
The notion of identifying-with comes from psychoanalysis. The
definition provided by Laplanche and Pontalis (1973, p. 205) is a good
place to start: Psychological process whereby the subject assimilates an
aspect, property or attribute of the other and is transformed, wholly or
partially, after the model the other provides. Through the process of
identifying with others bodily expressed emotional attitudes, for
example, an individual perceives and assimilates the attitudes in such a
way that they become possibilities for the persons own relations with
the world, including the individuals relations toward him- or herself.
Children who are beaten may come not only to beat others, but also to
have punitive attitudes toward themselves.
There is a complication here, namely that the very nature of
identification changes with development. Although Freud (1955/1921)
illustrated his notion with a cognitively elaborated instance, namely a
boys wish to be like his father, he concluded a brief essay on
identification with a footnote in which he made the following claim: A
path leads from identification by way of imitation to empathy, that is, to
the comprehension of the mechanism by means of which we are enabled
to take up any attitude at all towards another mental life (p. 110).
Clearly this refers to a much more basic level of identifying-with, but
one for which it remains true that . . . identification is not simple
imitation but assimilation(Freud, 1953/1900, p. 150, Freuds italics).
Identifying-with is a process that links individuals without merging their
identities. In the act of connecting with someone else through
identification, one retains a distinction between self and other as a basis
for oneself assuming what one experiences as the other persons stance.
To repeat: what is experienced as other-centered within ones own
experience can become a feature of ones self-centered repertoire of
feeling and action.
Therefore the notion of identification that we employ does not
correspond with the concept of self-other merging justifiably criticized
by Batson (2011, pp. 14560). The idea is not that oneself and the other
become indistinguishable, nor that one experiences what someone else
experiences, nor that there is a confusion between self and other. Like
Batson, we have emphasized that at the core of empathy is the capacity
and propensity to feel for others, for their own sakes. Where Batson
(2011, p. 11) refers to other-oriented emotions, (p.188) we have
written of person-centered qualities of relational self-awareness
(Hobson, Chidambi, Lee, & Meyer, 2006, p. vii), where the critical
feature is that the other person is encompassed within a feeling state.
In the case of responding to attitudes, there appears to be a dissociation
between being affected by expressions of feeling in others in a rather ill-
focussed manner, and being affected (through identification) by the
other persons feelings as the feelings of another self with whom one is
engaged (Hobson, Chidambi, Lee, Meyer,2006, p. 135).
In order to think about this further, consider the case of sharing
experiences. If I have an experience of sharing feelings with someone
else, this is indeed my own experience. On the other hand, it is only (felt
to be) sharing insofar as my experience encompasses the other as
participating, with me, in that experience. In other instances, ones
interpersonal experience may encompass a registration of the others
attitude, for example of anger, alongside a complementary feeling of a
different kind, say of fear. In yet other cases, where one might identify
with attitudes directed elsewhere than toward oneself, the others
attitude may be registered with relative equanimity. In the case of
empathy, ones responsiveness to someone elses suffering encompasses
both what one registers as a result of seeing or imagining the person
suffering and a relation toward the person. In each instance, there is a
special structure and phenomenology to such experiences, one that
entails that one registers the distinctiveness of whatever is experienced
as originating in the other.
Vital aspects of psychological development occur through the
interiorization of interpersonal transactions, as long argued by
psychoanalysts and developmentalists in the tradition of Vygotsky
(1978). The conundrum is that the individual needs to construct the
socialor if you like, experience the social as socialwhere what is
social then makes a pivotal contribution to the individuals development.
In virtue of the structure of the process of identifying-with, certain
emotions entail an expectation and/or experience of otherness (a central
tenet of post-Freudian psychoanalytic thinking). This is foundational for
social experience and at the same time sets the stage for social
experience to shape the self. Identifying-with serves as a mechanism for
the enrichment and development of the individual through his/her
engagements with other people. What results is a mind that is composed
of parts of the self in various states of relatedness to each other and to
various degrees felt to be central to or alienated from the self.
Identifying-with has cognitive, affective, and motivational aspects
(Hobson, 2008). To identify with someone else is to be engaged with
them affectively, it is to be motivated to feel and behave in certain ways,
and it is to apply cognitive categories of relevant kinds. The earliest
forms of identification (or one might say, the structures of self-other
relatedness that provide the basis for (p.189) identification proper)
appear early in the first year of life. The manifestations are the uniquely
human forms of sharing evident in typically developing infants from
around two months of age. Typically developing two-month-old infants
appear not only to share affectively charged exchanges with their
caregivers (Trevarthen, 1979), but also to be upset by disruption in the
interactions. A prime example is when a caregiver adopts a still face
(Tronick, Als, Adamson, Wise, & Brazelton,1978)and we have seen
two-month-olds trying to reinstate a pleasureable to-and-fro with their
still-faced mothers. On the other hand, of course, we should not suppose
that the infants experience another person in the way that we do.
Rather, it is the case for infants as well as adults that certain states of
mind entail that one registers an embodied other who plays an integral
role in making that state of mind what it is. When we witness two-
month-olds sharing pleasure in face-to-face interaction with their
mothers and then either intermittently averting their gaze or making
bids for re-engagement when the mothers assume a blank face (see
Hobson,2002/4 for further details), the other is essential to a
description of what the infants are experiencing.
Then more explicit forms of identifying-with become apparent later in
the first year, when infants are moved to adopt the attitudes of another
person toward a shared world. Examples are instances of social
referencing, when infants attitudes to objects and events may change in
accordance with their perception of other peoples attitudes to those
same objects and events (e.g., Sorce, Emde, Campos, & Klinnert, 1985),
or certain varieties of joint attention and imitation (Trevarthen &
Hubley, 1978, for vivid illustrations, and Zahn-Waxler, Radke-Yarrow,
Wagner, & Chapman, 1992, for subsequent developments in empathy
over the second year). It is possible, albeit hazardous, to see phenomena
such as joint attention and social referencing as amounting to a form of
understanding, but it is important not to suppose one can speak of the
infant knowing that someone else has such-and-such a take on the
world. Rather, as John Campbell (2005, p. 288) has expressed the
matter, On a relational view, joint attention is a primitive phenomenon
of consciousness. Just as the object you see can be a constituent of your
experience, so too it can be a constituent of your experience that the
other person is, with you, jointly attending to the object. In other
words, we are still in the realm of what is structured in the givenness of
experience, rather than what is built up out of component
understandings. Or to put this differently: a part of what goes into
childrens (and our) understanding of persons is what they already
experience as sharing with persons, including what they experience as
sharing in relation to a world out there. Finally, there are further
versions of identifying-with that develop later in life, for instance when a
child identifies with a parent or when someone identifies with a
religious group.
(p.190) Two observations, one from developmental psychopathology
and one from neuroscience, are pertinent here. First, it is commonplace
to find young children acting out scenarios in their symbolic play. When
the scenarios involve pretend people, the figures are often given roles in
interaction with each other. Not infrequently, the roles have personal
significance for the child. For instance, if the child has been told off by a
parentand what one might have seen at the actual telling-off was an
upset child (although more accurately, a child-upset-by-being-told-off-
by-a-parent)what one may observe subsequently is the child re-
enacting boththe figure of the scolding parent and the figure of the
upset child. The self-other structure of the original experience becomes
unpacked, as it were, in the replaying of events. Such patterns of
transgenerational identification can be a potent source of
psychopathology, for example when abused children become abusers.
Then from the field of neuroscience there is evidence for something like
resonance of action-readiness and/or feelings between one individual
and another (e.g., as considered by Decety & Chaminade, 2003)and
also provisional evidence that such neurofunctional mirroring is
relatively absent among individuals with autism and/or Asperger
syndrome (e.g., Dapretto, Davies, Pfeifer, Scott, & Sigman, 2006;
Oberman, Hubbard, McCleery, Altschuler, Ramachandran, & Pineda,
2005; but see Southgate & Hamilton, 2008). Although such evidence
supports the notion that there is transmission or communication of at
least certain features of psychological states from one person and
another, one needs to be circumspect about using such terms as
mirroring or simulating, when mirroring is so inexact a metaphor for
the complex processes of identifying-with.
Wrap-Up
It may be appropriate to conclude with some final reflections on where
the present account is situated within current philosophical debate on
the nature of and basis for interpersonal understanding. It diverges in
very many respects from theory theory attempts to explain basic
mechanisms of interpersonal understanding (e.g., Hobson, 1993b). This
is notwithstanding that mental concepts and self-reflective forms of
role-taking feature in developmentally elaborated forms of mind-
reading. Perhaps the most important point is that quintessentially
(although not exclusively), early forms of identifying-with are
perceptually grounded emotional processes that are necessary for, rather
than dependent upon, the acquisition of concepts of persons-with-
minds.
The present approach also eschews ideas that appear in versions of
simulation theory. For example, identifying-with does not work from an
egocentric (p.191) stance. Rather, it is a process that structures
interpersonal engagement. Self/other-awareness and understanding are
constructed on the basis of emotionally configured intersubjective
experience. Therefore it is not the case that in the early phases of life, a
child uses him- or herself as a model for understanding someone else.
Simulationist accounts tend to underestimate how much development
in self-awareness and conceptual ability needs to have taken place
before a child could use him-/herself as a model for anything. Second,
identifying-with does not depend upon imagination, in any of the usual
senses of that term. On the contrary, imaginative role-taking becomes
possible on the basis of infants experiencing specific forms of
interpersonally grounded shift in attitude toward the world though their
affectively configured perception of and alignment with the attitudes of
others.
We should add that an account in terms of identifying-with alters the
prominence given to a range of self-other and person-world relations in
our explanation of what it means to understand oneself and others, and
how we come to such understanding. For example, consider a very
young childs ability to perceive and respond to what one might call a
possessive or acquisitive attitude in someone else. Around the second
birthday, a child adopts the adults expression Mine! by identifying
with the attitude this expresses when used by an adult or older sibling
(Charney, 1980). Is not this, too, of great significance for our grasp of
what it means to be a self with a set of person-anchored desires toward
and beliefs about the world?
There are very many sorts of emotion, a fact that is obscured only by
our conventional habits of thinking in terms of abstracted feelings such
as those of happiness, sadness, anger, fear, surprise, and disgust (to
mention only the supposedly basic emotions). To be happy-with-
someone-who-is-talking-with-oneself is not the same mental state as to
be happy-with-someone-who-is-greeting-oneself, never mind the same
as happy-with-someone-who-is-outside-ones-perceptual-field, or even
happy-with-a-good-meal. To be angry with someone who has just
expressed an insult is not the same mental state as to be angry with
someone who has forgotten ones birthday, never mind angry with
someone who has abandoned one, or even angry with a broken-down
car. And so on, ad infinitum. Such affective states are not only relational,
they also have specificity in relation to the identity and state of the
object of the relation (that is, how the object of the intentional state is
experienced). Again to cite Goldie (2000, p. 191), when one is in a
confrontational interaction, one may very clearly recognize anothers
emotion, yet it would be absurd to presume that one has the same
emotion as the other person. If one feels rage-in-relation-to-the-
others dismissiveness, then ones mental state is not the same as that
of the other, but it does encompass ones own version of the others
mental state. This is even true for sharing, when it is only in a (p.192)
special sense that one should speak of ones own state being the same as
that of the other.
In conclusion, then, we see empathy as one among a range of states
structured by the process of identifying-with the attitudes of another
person. The case of autism helps us to see what happens when this
structuring of social experience is limited or absent. In particular, a
limited quality of other-person-centeredness is revealed not only in a
restriction in the organization of affected individuals feelings, thoughts,
and actions toward others when it comes to such states as empathy,
concern, and guilt, but also in diminished role-taking and flexibility in
thinking, language, and imagination. This is not to say that individuals
with autism are without any capacity to identify-with. Nor is it to claim
that without such a process, all manifestations of empathy, guilt, and so
on, never mind flexibility in thinking, are impossible. Our claim is that
identifying-with gives a special quality and depth to these modes of
psychological functioning and to the moral stance to which they make a
central contribution.
Empathy in Other Apes
Kristin Andrews
Lori Gruen
1. Introduction
Aldrin was a sickly little fellow and didnt play with the others very
much. In fact, he usually didnt do much besides sit next to his
babysitter and hug her leg. But one day a terrifying turtle appeared, and
he was motivated to climb high in a tree to escape the horror. Later that
day when it was time to head back to camp, the babysitters realized that
Aldrin wasnt with them. They never saw him come down from the tree.
Then the babysitters noticed that Ceceb, the leader of the group of
youngsters, wasnt around either. When they went back to where the
turtle had been, they found Aldrin and Ceceb perched high in different
trees. Cecebs tree was closest to the path, and he looked back at Aldrin,
caught his eye, and then moved on to the next tree. Aldrin followed
Ceceb from tree to tree until they reached the path back to camp.
Though Ceceb had been looking back at Aldrin from time to time, when
he got down to the ground he just scampered away, joining the rest of
the group, with Aldrin following.
Hearing this story, one might be inclined to talk about Aldrins fear,
Cecebs understanding of Aldrins emotional state, and his desire to help.
It would not be unusual to think that Ceceb was responding
sympathetically to Aldrin, understanding that he was afraid and trying to
calm him. Perhaps one might suspect that Cecebs sympathetic response
was caused by an empathic reaction (p.194) to Aldrins plight. Further,
one might dramatize the story by describing Ceceb as playing the role of
the policeman who is trying to keep the peace and make sure everyone is
doing OK. As juvenile rehabilitant orangutans, however, Ceceb isnt the
kind of creature to whom these ideas are generally applied. If he were
human, there may be little protest. However, the cognitive requirements
for empathy, sympathy, and grasping social norms are not generally
thought to be possessed by nonhuman animals.
A number of scholars have offered behavioral and physiological
arguments in favor of the existence of empathy in other species (see
Bekoff & Pierce 2009, Flack & de Waal 2000, Plutchik 1987). While the
evidence is compelling, claims about empathy in nonhuman apes face
two different challenges. The first challenge comes from a set of
empirical findings that suggest great apes are not able to think about
others beliefs. The argument here is based on a view that empathy is
associated with folk psychological understanding of others mental
states, or mindreading, and the existence of mindreading among the
other apes is a matter of some dispute. The second worry comes from a
host of recent experiments suggesting that nonhuman great ape
communities lack certain social norms that we might expect empathic
creatures to have, namely cooperation norms, norms of fairness, and
punishment in response to violations of norms (especially third-party
punishment). If apes are empathetic, yet they do not use this capacity to
help or punish, what is the role of empathy? We think that both these
challenges can be answered by getting clearer about what empathy is
and how it functions as well as by considering the nature of empathic
societies. We also believe that this analysis will clarify the relationship
between being empathetic and being ethical.
2. Varieties of Empathy
Both the concept of empathy and the phenomenon have been
understood in many different, often contradictory, ways, and this makes
it particularly tricky to determine what is being claimed when someone
says that other apes are or are not empathetic. In everyday use, empathy
is usually thought to be connected with ethical perceptions and
behavior. An empathetic person is a good person, someone with
qualities and virtues that are to be praised. One reason why there is
skepticism about whether apes or other animals engage in empathy is
because it is hard to understand the idea that animals have morality.
There is a growing acceptance of the idea that they may have what de
Waal has called the building blocks of morality, which includes
empathy as well as reciprocity, conflict resolution, a sense of fairness,
and cooperation, but perhaps not full blown ethical agency (de Waal,
2006).
(p.195) In the psychological literature, empathy is alternatively used to
mean a state of feeling what another person or being is feeling (an
affective state that may or may not require cognition), knowing what
another person or being is feeling (an epistemic state that involves
mindreading or metacognition), or responding compassionately to
anothers distress (perception/action state that is often associated with
ethical engagement) (Levenson and Ruef, 1992, 234). The
phenomenological and affective states do not necessarily require
cognition. The epistemic state requires both phenomenological
experience and other affective mental states. Responding
compassionately, caring for and about, or engaging in what one of us
calls entangled empathy (Gruen 2012, 2013) requires both cognitive and
affective states, but not necessarily the same sorts of states that are
associated with the other types of empathy. But all of these types of
empathic experiences involve the transfer of emotion, and this transfer
occurs in a variety of ways.
The most basic form of empathy, usually called emotional contagion or
affective resonance, involves a spontaneous response to the emotions of
another. Anyone who has lived with dogs will be familiar with this
phenomenon. Dogs are emotional spongesthey often become stressed
when their person is stressed, sad when their person is sad, joyful when
their person is joyful. Infants and small children also regularly engage in
these spontaneous reactions. Emotional contagion or affective
resonance is a kind of mimicry of the individual(s) in ones immediate
environment and does not require any developed cognitive capacities.
This very basic type of empathy involves the direct perception of the
emotions of others and automatically triggers or activates the same
emotion in the perceiver, without any intervening labeling, associative,
or cognitive perspective-taking processes. (Lipps 1903b) And in the
majority of such cases, this initial response seems unavoidable.
With the more automatic forms of empathy, the empathizer isnt
distinguishing his or her own feelings or mental states more generally
from those of another. In fact, an awareness of this distinction in agency
may not yet have developed and perhaps never will. And in cases where
such awareness already exists, occurrent recognition of the others
individuality may interfere with the emotional sharing, as between a
mother and infant or in a freshly declared love relationship. Some
theorists have limited their understanding of empathy to just these sorts
of experiences and from that point of view it is difficult to see what role,
if any, empathy that involves fellow feeling in which the agent loses
herself in the emotions of another should play in an account of ethical
engagement.
There are also types of empathy that rely on more complex cognitive
capacities. One sort of cognitive empathy involves taking the perspective
of another in order to understand what that other is experiencing and
making decisions about what to do in light of what the other is
experiencing. This sort (p.196) of empathy generally requires
mindreading or metacognition, and is what the psychologist William
Ickes (1993) calls empathic accuracy. We will discus the conflicting
evidence about metacognition in apes in Section 4 below. The other sort
of empathy is entangled empathy that involves being able to understand
and respond to anothers needs, interests, desires, vulnerabilities, and
perspectives not as if they are or should be the same as ones own. It
involves a reaction to anothers experience and a judgment to act in
response. This latter form of empathy has a clearer connection to ethics,
although here too there is disagreement (Prinz, 2011a, 2011b).
3. Functions of Empathy
Among empathys functions is to better understand the individuals in
ones societyto know what they want and why they want it. There has
been great interest in the evolution of this ability as an explanation for
cognitive differences across species that led to the development of the
Social Intelligence Hypothesis (SIH). Most generally, this hypothesis
suggests that social animals evolved a greater cognitive complexity
because of the need to interact with a great number of autonomous
agents. Apes and monkeys (as well as the social carnivores, birds such as
the corvids, and marine mammals such as the bottlenose dolphin) live in
intricate social groups that require substantial cognitive commitment;
they must be able to recognize individuals (visually, aurally, and perhaps
via other modalities as well), they must keep track of kin relations
(especially in matrilineal species such as baboons), they must keep track
of dominance relations and alliances, plus they must be sensitive to
possible defections. They must be able to remember who did what to
whom when, and who should care about it. In addition, they must decide
what to do in the face of such actions and make judgments about
whether they should, for example, challenge a dominant, join a coup, or
court the dominants mate. They must decide when to let others know
they have found food and when to keep it for themselves. The SIH is
premised on the theory that sophisticated cognition must be adaptive
given the high costs associated with developing a large brain. Evolution
does not optimize, and creatures certainly shouldnt be expected to be
cleverer than they need to be. From this it follows that primates
developed sophisticated cognitive abilities for some function.
There are two approaches to this hypothesis. According to Machiavellian
versions of the hypothesis (Humphrey1976, 1978; Byrne & Whiten
1988), the ability to understand other minds arose in order to come out
on top in a cutthroat environment of scarce resources. By understanding
what others believe and what they want, and by being able to manipulate
others beliefs or change (p.197) their desires, one can steer
competitors away. The Machiavellian perspective emphasizes the
importance of making predictions in order to thrive in this competitive
environment. For example, if two individuals both want a food item, and
there isnt enough to share, the individual who can predict that an
intervention will lead the competitor away from the food will be the one
who gains the food. Given the fiercely competitive primate social
environment, making better predictions of behavior was instrumental
for gaining greater resources; better predictions were used to better
manipulate others behavior. As individuals gain a more sophisticated
theory of social action and greater predictive success, they up the stakes
for other members of their community, thus creating an evolutionary
arms race. Both active lies and withholding information such as food
alarm cries are examples of Machiavellian social intelligence.
The other version of the social intelligence hypothesis was introduced by
primatologist Allison Jolly (1966). Based on her expertise in lemur
behavior, Jolly suggests that cooperative social learning rather than
fierce social competition explains why social animals need greater
cognitive complexity. Social learning is a nonpedagogical method of
learning, which requires that a demonstrator tolerates the close
observation of the learner, and in many cases the learner gains some of
the benefits of the behavior being demonstrated. For example, in
orangutan food processing the mother will allow her infant to peer at
her complex manipulation of a ginger leaf or termite nest, and she will
allow her offspring to take pieces of processed food to eat. While this
sort of learning doesnt involve active teaching, it does require acting
differently toward individuals with differing abilities and responding
appropriately to different individuals depending on their current skill
levels.
We think that Jollys version of the hypothesis is more plausible for a
number of reasons. One of us has argued that the kind of predictions
emphasized by the Machiavellian Intelligence version of the social
intelligence hypothesis could be made without understanding the
content of other minds and without feeling what others feel (Andrews
2012). In addition, as researchers turn to examine cultural differences
between communities of a species, we are finding that social learning is
an essential part of the lives of social animals. Indeed, when we compare
wild apes with captive or rehabilitant apes, we see that the lack of social
learning opportunities among such individuals have led to harm for the
individuals and the new groups, leading to problems such as an inability
to properly care for offspring (rehabilitant orangutan mothers who
inadvertently drown their infants when crossing through streams, for
example) and inability to find nutritious food to eat. Further, the
traditions of ape societies such as orangutan habitual routes appear to
be learned by the infants as they are carried on their mothers backs;
juveniles have been observed to begin leading the way on habitual
routes and waiting for mother at the next stop on (p.198) the path
(Bebko 2013). Social learning leads to the development of cultural
behavior, defined as a behavior that is transmitted repeatedly until it
becomes widespread through a population (Whiten et al. 1999). A new
behavior may be introduced to the communitys behavioral repertoire by
an immigrant or by a community member who innovated the behavior.
Innovation is defined as the process that generates in an individual a
novel learned behavior that is not simply a consequence of social
learning or environmental induction (Ramsey, Bastian, & van Schaik
2007, 395; see also Reader & Laland 2003). Innovations are beneficial
behaviors, and as they spread through a community they make life
better for the individuals.
The way innovations or other learned behaviors spread through a
community is not unlike how some hunter-gatherer human adults pass
on their social knowledge. A recent ethnographic survey of learning in
hunter-gatherer societies concludes that [t]he sources discussed here
suggest that a range of learning processes are involved in acquiring
hunting skills, and that teaching and demonstration play a limited role
(MacDonald 2007, 398). In hunter-gatherer societies, facilitative
teaching is the norm, examples of which include allowing young
children to accompany adult experts on hunting trips or to play with the
adults tools or weapons at home.
Infant and juvenile nonhuman apes have much to learn from their
mothers as well (McGrew 1992). Much of this learning occurs via
facilitative teaching, as described by MacDonald, but there are also
reports of active teaching among chimpanzees. At the Fongoli research
site in Senegal, chimpanzees make a variety of sharp stick tools to hunt
small bush babies that can involve up to five steps to construct,
including trimming the tool tip to a point. The chimpanzees prepare the
tools, take them to a particular area, and then jab them forcefully into
tree hollows where the small primate prey nests. Pruetz has observed
what appeared to be a mother teaching the tool-making and hunting
techniques to infants not only by modeling the tool-making behavior but
also by physically correcting the youngsters tool (Pruetz & Bertolani
2007). In addition, observations of the chimpanzees of the Ta Forest in
Cte DIvoire suggest that they also engage in demonstration teaching
(Boesch 1991, 1993). An adult female named Ricci observed her daughter
Nina trying unsuccessfully to crack nuts with a stone hammer. Ricci
approached Nina, who immediately handed her mother the stone. With
Nina watching closely, Ricci turned the odd-shaped stone to its best
position for cracking the nut in a very slow and deliberate fashion. Then
Ricci cracked ten nuts, letting Nina eat almost all of them, dropped the
stone, and left. Nina picked up the stone and held it in the same position
Ricci had.
Teaching by inhibition, or by preventing another individual from acting,
is also apparent among chimpanzees. Wild chimpanzee mothers have
been (p.199) observed to pull their infants away from plants that are
not part of their regular diet (Hiraiwa-Hasegawa 1990). In captive
settings, researchers have observed mothers intervene when their infant
played with unusual and potentially dangerous objects, such as a heavy
metal chain (Hirata2009).
Kim Sterelny (2012) has argued that the complex culture we see in
human societies emerged from the kind of facilitative teaching we think
exists among the great apes, and which MacDonald describes in
contemporary hunter-gatherer societies. While Sterelny doesnt apply
his account to the great apes, we think that much of what he says about
the evolution of human culture through apprenticeship learning can also
be said of the other great apes. According to his apprenticeship learning
model, humans evolved in an environment organized by humans for
learning, and without explicit teaching or any specific cognitive adaption
for teaching humans were able to develop complex culture. Like
meerkats, whose young gradually learn how to kill and eat dangerous
scorpions from adults giving the young dead scorpions first and then
half-killed scorpions next, human experts often prepare gradual learning
steps for apprentices by task decomposition and ordering skill
acquisition (Sterelny, 2012, 35). In great apes societies, as it is with
human children, youngsters are given many opportunities for learning
by adults. MacDonald (2007) points out that in hunter-gatherer
societies, adults are tolerant of children closely looking at their activity
and playing with their tools. The same sort of tolerance has been
reported among chimpanzees and orangutans (see Van Schaik 2003 for a
review).
If were right and Jollys version of the social intelligence hypothesis is
correct, then there is a real relationship between understanding others
and the behaviors associated with different forms of teaching and
learning. We should expect, then, that empathy would have evolved in
order to facilitate teaching and learning and the transmission of social
traditions, which in addition to behaviors such as food processing can
also include behaviors that may be understood as examples of social
norms, such as the prohibition against infanticide in chimpanzee
societies (see Rudolf von Rohr, Burkart, & van Schaik 2011 for a review)
and the assistance male chimpanzees provide to females and children in
crossing roads (Hockings, Anderson, & Matsuzawa 2006) that we will
say more about in Section 5 below.
Etiological Considerations
The Risk-Need-Responsivity model (RNR) of offender rehabilitation
states that effective correctional interventions should follow the
principles of risk, need, and responsivity. While a number of conceptual
and practice problems have been identified in this model, most
researchers and practitioners working with offenders agree that ethical
and effective practice should be guided by the RNR principles (Ward &
Maruna, 2007; Ward & Stewart, 2003). One core requirement of RNR
practice is that clinicians concentrate their therapeutic efforts on
managing or eliminating dynamic risk factors. These psychological and
environmental variables are thought to causally contribute to the onset
of criminal events and their successful reduction typically results in
lowered reoffending rates (Andrews & Bonta, 2010). The theoretical
framework we derived from Kitchers multidimensional concept of
psychological altruism can easily accommodate the RNR principles in
the following way. Criminogenic needs such as offense-supportive
beliefs and attitudes, intimacy deficits, emotional regulation problems,
substance abuse, and impulsivity represent causal variables that are
likely to impair the ability of offenders to act in a psychologically
altruistic way. For example, offense-supportive beliefs, or what(p.224)
have been termed cognitive distortions, typically cast potential victims in
ways that permit the offender not to consider them as having the same
moral status as them or else as possessing desires and preferences that
make sexual abuse acceptable. This is a problem relating to the range
dimension. Two good examples of this type of cognitive distortions are
the belief that women are untrustworthy or dangerous, and that children
are sexual agents (Gannon & Polaschek, 2006). The former depicts
women as belonging to a class of beings whose desires and interests are
not that relevant when engaging in sex and the latter portrays children
as competent sexual beings who are capable of making decisions about
sex for themselves. We suggest that all of the dimensions of intensity,
range, scope, discernment, and empathetic skill can be linked to causal
factors resulting in a sexual offense, directly or indirectly (see below).
Assessment
The aim of the assessment phase of sex offender treatment is to
systematically collect clinically relevant information about individuals
offending, functional life domains, personal characteristics, and
developmental and social history. Once a sex offenders problems have
been identified a case formulation (or miniclinical theory) is
constructed in which the nature of the problems, their onset,
development, and interrelationships are described. Following the
development of a case formulation, clinicians construct an intervention
plan in which the various treatment goals, their sequencing, and
strategies for achieving them are noted. As outlined earlier, the
components of a comprehenisve sex offender treatment program should
include the following types of interventions: cognitive
restructuring/offense reflection, sexual reconditioning, sexual
education, social skill training, problem solving, (empathy) perspective
taking/constructing victim biographies/victim impact work, intimacy
work, acquiring emotional regulation skills, lifestyle/leisure planning
and experience, vocational training, and reentry or adjustment planning
including relapse prevention (Marshall, Marshall, Serran, & Fernandez,
2006; Laws & Ward, 2011).
When formulating a case the theoretical framework we derived from
Kitchers altruism dimensions can be used to direct and concentrate
clinical attention to certain kinds of problems. Drawing from the
assessment data (comprising interview information, psychological
measures, archive data, behavioral observations, etc.) practitioners can
ask the following questions, each covering one of the five dimensions of
altruism.
Range. Are there any individuals or classes of people explicitly excluded
from Xs list of altruism targets? Does he hold certain beliefs or
attitudes that (p.225) effectively disenfranchise persons from a
consideration of their interests, for example children or young adult
women? Does he lack the skills to communicate openly and honestly
with adults?
Scope. Are there any internal contexts in which Xs ability to act
altruistically are compromised in some way? For example, does he find
it hard to take account of someone elses interests when feeling angry,
sexually aroused, or lonely? What about external contexts? Does X
struggle to control his sexually deviant desires and preferences when
alone with a child or woman? What about if he is in the company of
certain groups of friends? Or when he is socially isolated?
Discernment. Does X lack an adequate understanding of the
psychological and developmental needs of children? Are his problem-
solving and inductive-reasoning skills of poor quality, making it difficult
for him to think through the consequences of acting in sexually abusive,
or offense reacted ways?
Empathetic skills. Does X struggle to accurately identify other peoples
mental states during an interaction? Is he able to adjust his actions in
light of his reading of others mental states?
Intensity. Does X possess the general practical reasoning and self-
management skills in order to frame other peoples situations in ways
that accurately describes what is going on for them? Having done this,
can he realign his own desires (and other relevant mental states) and
actions in order to respond in an appropriate manner? We view intensity
as a more global capacity that builds on the skills and so on aligned to
the other altruism dimensions.
It is anticipated that the answers to the above questions will enable
practitioners to pinpoint the reasons why a sex offender acted in ways
contrary to the desires and interests of their victim. This information
can then be recruited in the construction of the case formulation and
subsequent intervention plan. A clinical benefit of using the
psychological altruism framework in this way is that it functions a bit
like a fishing net: it guides clinical attention to a broad range of possibly
relevant offense variables, and then by drawing the threads of the net
together, focuses energies on the issue of altruism failures and building
competencies to promote psychological altruism. It has a certain
elegance about it.
Practice
In discussing the practice implications of the theoretical framework
derived from Kitchers concept of psychological altruism, we will
describe briefly a number of typical sex offender treatment modules and
trace their potential for strengthening altruistic actions. The description
of the modules content is based on our clinical experiences and the
work by Marshall and colleagues (p.226) (e.g., Marshall, Marshall,
Serran, & Fernandez,2006) and Ward and colleagues (e.g., Ward, Mann,
& Gannon, 2007).
Empathy Training
As stated earlier in this chapter, the major aim of the empathy module is
to encourage offenders to reflect on the impact of sexual abuse on
victims and their families. This is achieved through the use of victim
biographies, role-plays of the index offense, and the assimilation of
information about sexual abuse and its consequences for victims.
Offenders often describe this as an emotionally devastating experience
and report that it helped them to grasp the self-serving nature of their
behavior and the callous disregard for the wellbeing of vulnerable
children and unconsenting adults.
Victim perspective-taking and appropriate emotional responding are
therapeutic targets of this module, classical components of an empathy
response. In the language of psychological altruism, an expectation is
that empathetic accuracy is improved, discernment skills are sharpened,
and contextual features of high-risk situations that increase the
likelihood of sexual crime occurring, are discovered.
Emotional Regulation
Emotional regulation modules tend to look closely at offenders
competence on a number of emotional tasks. These include being able to
accurately identify and label an emotion, in oneself and in others; once
the emotions have been correctly identified, knowing how to act in
(adaptive) ways prompted by the emotion in question; and being able to
manage powerful emotional states so they do not overwhelm the person
concerned.
Powerful emotional states can disinhibit individuals and create immense
pressure on them to act non-altruistically. For example, if an offender is
experiencing strong feelings of anger, self-control could prove to be
particularly challenging. Norms directing him to attend to his potential
sexual partners desires or preferences may be overlooked and his own
desires are thought to trump all other motivations; he commits an
offense. Alternatively, another sex offender could use sex as a soothing
activity and when feeling vulnerable, anxious, or depressed seeks out a
sexual partner. These kinds of problems are unfortunately relatively
common and point to issues with psychological altruism. Perhaps the
most obvious issue relates to one of internal context, where failure to
effectively modulate certain moods makes it hard for an offender to
interact in a psychologically altruistic manner; his own desires and
needs take precedence in a context when the reverse should be true.
Problem Solving
The final module we will consider is that of problem solving. Basically,
in this module offenders learn how to frame problems and work toward
effective solutions. The aim is to increase their ability to step back from
social and personal crises in order to reflect on the nature of the
difficulty, and by thinking in a flexible and pragmatic way arrive at a
workable solution. Offenders learn the various phases of problem
solving and how to seek relevant information when deciding between a
number of options to resolve their difficulties.
The acquisition of good problem-solving skills is most likely to improve
the way offenders think about the consequences of their actions
(discernment dimension) although it does have implications for the
other dimensions as well. (p.229) For example, when faced with an
interpersonal problem or experiencing a negative emotion such as
intense fear, the offender would ideally sit back and ask himself what is
going on. Creating cognitive space between feeling and acting should
open up further opportunities to explore his difficulties and to consider
alternative ways of dealing with them. This could result in a shift of
focus from his own needs to what the potential victim is experiencing,
and ultimately, to a decision to realign his own desires to those of the
other person and not to engage in a sexual assault. It is also to be
expected that improved problem-solving skills could impact in a positive
manner on offenders cognitive distortions and thus contribute to
dealing with any possible altruism failures associated with the
dimension of range.
Conclusions
In this chapter we have explored the relevance of the concept of
empathy for sex offender research and practice. In doing so, it became
apparent that empathy may play an important role in motivating
individuals to act in morally acceptable ways, and importantly, to cease
offending. After examining empathy and its conceptualization in the
sexual offending field more closely, we concluded that the concept of
psychological altruism and its associated five dimensions could
incorporate valued aspects of empathy, while avoiding some of the
conceptual and practice-related problems that attend it. After describing
Kitchers concept of psychological altruism, and using it as the basis for
an altruism theoretical framework, we investigated its implications for
practice. In our view, the capacity of the psychological altruism concept
to provide an ethical and theoretical framework for viewing correctional
practice is encouraging. It reminds practitioners that work with sex
offenders has a strong normative as well as a scientific or empirical
dimension and that the concept of psychological altruism is much better
positioned to provide this broader perspective than that of empathy.
Empathy and Morality in Ethnographic
Perspective
Douglas Hollan
What is the relationship between empathy and morality? Is empathy
inherently prosocial, leading to altruistic acts and behaviors as some
evolutionary-minded psychologists (Hoffman 2011, Haidt 2012, Batson
2012), neurobiologists (Harris 2007), ethologists (de Waal 2009), and
care ethicists (Slote 2007) have suggested? Or is it instead certain moral
climates that promote the development of human empathy and that
encourage or use empathy to promote peoples care and well-being? In
this chapter, I address these and related questions by reviewing some of
the recent ethnographic work on empathy and morality, work that
attempts to describe and analyze these phenomena in social and cultural
contexts, as a part of ongoing, naturally occurring behavior. I begin by
discussing recent anthropological definitions and conceptions of
empathy and morality before turning to the ethnographic evidence as
to how these two phenomena are related to one another. I argue that
while there is ample evidence from neurobiology and ethology to
suggest that basic forms of empathy (Stueber2006)rooted in
automatic, biologically based, embodied forms of imitation and
attunementare critical to human sociality and communication
anywhere, the ways in which these basic forms become developed and
elaborated into more complex, marked forms of empathy (p.231)
(Hollan & Throop 2008, Hollan & Throop2011b) that can be mobilized to
help or to harm others may vary considerably across communities,
across individuals within communities (Hollan 2008, 2011, Groark
2008), and through time.
(p.246) Conclusion
I have used some contemporary ethnographic studies of empathy and
morality to examine some of the claims from moral psychology and
philosophy about the significance of empathy in human life.
Ethnographic studies are helpful in this regard because rather than
presume the centrality of empathy on theoretical or conceptual grounds,
they examine how empathic processes manifest themselves and unfold
in the course of everyday, naturally occurring behavior, and how these
processes are related to, if not contingent upon, other forms of social
knowing, awareness, and communication. One of the first things
ethnographic studies make evident is just how difficult it is to define
empathy in a cross-cultural and historical context. While many groups
appear to have concepts that resemble or overlap with the formal
academic definition of empathy as a quasi-first-person perspective on
anothers situation that involves both affective resonance and
imaginative perspective taking, few have terms or concepts that are
identical to it. Many groups seem to have concepts that shade more
closely, both semantically and behaviorally, into what English speakers
would call sympathy, compassion, love, pity, or some
combination and blending of these terms. This raises the issue of
whether empathy per se is ever found as a relatively isolated
experience, or whether it is an awareness emanating from very basic
processes of intersubjectivity that must be culturally developed and then
carved out of other closely related social sentiments, with boundaries
that remain semantically and behaviorally fuzzy and open to cultural
and symbolic mediations of various types, including cultural and
symbolic suppression and inhibition.
Beyond these basic definitional issues, though, I have used ethnographic
data to evaluate claims about the extent to which empathic processes are
inherently prosocial and altruistic, whether they are inherently biased
toward the near and dear, and whether they can be used as a reliable and
consistent framework for moral evaluation and action. While it is clear
that many cultural and linguistic groups identify and label forms of
prosocial, positively valenced behavior and sentiments that English
speakers sometimes refer to as empathy, it is equally evident that
many of the same groups hold ambivalent attitudes toward such
sentiments. Among many of these groups, there is an acute awareness
that though empathic-like knowledge can be used to help people, it may
also be used to hurt or harm them. This wariness regarding first-person
perspective taking takes particularly stark form in the Pacific region,
where many groups assert the fundamental opaqueness of other
peoples hearts and minds. I have argued, following Keane (2008), that
this opacity doctrine is not so much an epistemological claim that one
cannot know the mind of another as it is a political and moral one about
what is proper to say or publicly(p.247) acknowledge about another
persons unexpressed feelings or intentions. However, the doctrine does
seem to express not only an empathic-like respect for the autonomy of
other peoples minds, but also an acknowledgement of peoples
vulnerability to the intimate knowledge others might have of them.
From an ethnographic perspective, then, it is clear that empathic-like
like knowledge of others is rarely, if ever, considered an unambiguously
good thing, despite the positive connotations empathy has in many
European-American contexts.
The question of whether empathic-like attitudes are necessarily limited
to the more near and dear at the expense of those who are less similar,
less familiar, and less proximate remains an open question from an
ethnographic point of view. Of course there are many groups around the
world that do indeed act preferentially toward immediate kin. Yet this
preference may not be as limited and automatic as De Waal (2009),
Preston (2007), and others contend. Mageo (2011), for example, has
shown that the circle of people toward whom empathy is directed and
encouraged is much larger in community-oriented Samao than among
most middle-class Americans in the United States. And among Banabans
on Fiji, the circle is extended even further, to include almost any
genuinely forlorn and needy person, no matter what ethnicity or religion
(Hermann 2011). Further, there is much ethnographic evidence to
suggest that morally salient and motivating forms of empathy can be
and are extended to nonhuman animals and numinous beings as well.
While some of these extensions are based on relatively pure forms of
faith and imagination, many of them involve embodied cultural practices
such as dream interpretation that establish a face to whom empathy
may be directed, though not necessarily the flesh and blood one that de
Waal posits. The question of whether these culturally mediated faces are
as powerfully eliciting of empathic processes as are the flesh and blood
ones can only be answered by more direct and explicit comparisons
between the two.
The ethnographic evidence regarding the relative fallibility or infallibility
of empathy as a framework for moral evaluation and action is more
definitive. I have already mentioned that even where empathic attitudes
and behaviors are valued and encouraged, people often fear their
inaccuracy or misuse. There seems to be a relatively widespread
awareness that empathic-like feelings and processes can be easily
overridden by the empathizers own needs, fears, and concerns, leading
him or her to misinterpret the targets behavior or to confuse the targets
situation with his or her own. People such as the Toraja also recognize
that even in the best of circumstances, individuals differ in their
willingness or ability to empathize in culturally appropriate ways, even
though they might not attribute these differences to differences in
disposition or developmental experiences, the way an outside observer
might. Apart from these relatively innocent ways in which empathy may
be lost, suppressed, or (p.248) overridden, we have far too many
ethnographic examples, unfortunately, of people deliberately and self-
consciously denying empathy to other classes or types of persons for
political or economic reasons. Whether for more innocent or sinister
reasons, then, ethnographic evidence suggests that empathy may be a
highly fallible form of moral orientation, subject to disruptions and
countervailing influences of many different kinds.
None of this is to deny the importance of basic embodied processes of
mirroring, attunement, affective resonance, and perspective taking in
providing a scaffold for many forms of human sociality and
communication, including morality. And indeed, one of the primary
lessons from contemporary ethology and evolutionary studies is that
Adam Smith and David Hume were probably right in arguing that
socially oriented emotions and intuitions play a critical role in enabling
and promoting human sociality everywhere. But these foundational
intersubjective processes, as Battaly (2011) suggests, are probably best
thought of as involuntary capacities, or, when they are self-consciously
practiced and honed, as voluntary skills, not as morally virtuous
attitudes or sentiments in themselves. They can perhaps become
powerfully motivating social sentiments, of course, but only if people
become self-conscious of them, value them in themselves and others,
and make it politically and economically feasible for them to be
culturally elaborated into more marked, complex forms of social
knowing and awareness (Hollan 2012), rather than politically or
culturally suppressed, inhibited, or punished.
The anthropologist Robert Levy (1973) once argued that cognitive and
emotional states could be differentially heightened and made
experientially vivid by hypercognizing them, that is, by naming them,
labeling them, and bringing cultural and moral attention to them.
Conversely, such states could be experientially downplayed or elided by
hypocognizing them, that is, by not naming them, not labeling them,
and by not bringing cultural and moral attention to them. In village
Tahiti in the 1960s, anger, according to Levy, was hypercognized while
grief was hypocognized. Levy argued that this was because anger was an
extremely dangerous emotion in the small face-to-face communities of
Tahiti. By hypercognizing and bringing conceptual and symbolic
mediation to emotional states of anger, Tahitians brought heightened
social and self awareness to them, which made their open display more
threatening and therefore less likely to occur. By knowing anger well, so
to speak, Tahitians could manage it better. Prolonged grief, on the other
hand, was thought to interfere with and disrupt the casual attitude that
one ideally was expected to bring to interpersonal attachments. By
hypocognizing grief, Tahitians culturally elided it, depriving people of
the conceptual and linguistic resources necessary to know it and
experience it in a focused, explicit way. As a result, people might report
symptoms of physical illness such as head- or (p.249) stomach-ache in
the aftermath of a broken or lost attachment, but they rarely complained
of experiencing a focused or prolonged sense of grief per se. We can
certainly imagine that empathic processes become differentially
enculturated in a similar way, making them more or less central to a
communitys behavioral repertoire and more or less central to its moral
orientations.
Levy also noted that the cultural centrality or marginality of certain
emotions and experiential states could change over time. For example,
in 1960s village Tahiti, shame, relative to guilt, was still a highly salient,
hypercognized emotion that played an important role in local morality
and social control. When Levy asked people why they did they not
commit socially disapproved of behaviors, they invariably told him that
they did not do such things for fear of being caught and shamed, not
because the behaviors were inherently wrong or because they felt guilty
about committing such acts. Levy makes the point that in face-to-face
communities like village Tahiti, people were likely to be seen if they
committed disapproved of acts, so that is was unnecessary to
hypercognize other types of inhibitory emotions, such as guilt, when
shame remained such a powerful means of inhibition and social control.
As Tahiti changed, however, and became more a society of relative
strangers, Levy hypothesized that Tahitians too, like people in many
other modern, developed societies, would begin to hypercognize guilt
more, since it is an inhibitory emotion that relies on an internalized
audience for its motivational force and saliency, not an actual audience
of neighbors and friends, as with shame.
With these examples, Levy is pointing out that human behavioral and
emotional repertoires are not set in stone, not simple expressions of an
underlying human nature, but rather they are in dynamical relationship
with the social environment, both responding to changes in the social
world, but also initiating changes within it.3Although Levy made his
observations at a time when Tahitians were hypercognizing shame and
hypocognizing guilt, he could imagine a time when this pattern would be
reversedthough both shame and guilt remaining in dynamical
relationship with each other and with other elements of the social and
moral ecology. Empathic processes are also embedded in larger
behavioral and social ecologies, of course. Though they can be given
more or less moral valence, and perhaps even be made morally virtuous,
nowhere do they operate in a social or behavioral vacuum. They operate
in the context of, with and beside, other behaviors and social restraints.
This is important to remember because while I have underscored how
morally fallible empathy appears to be from an ethnographic
perspective, no other moral (p.250) orientation is any less subject to
disruptions and countervailing influences of various kinds. The ecology
of morality in any place or time is a complex, dynamic one. Empathic
processes can and do play an important role in these ecologies, but the
exact ways in which basic embodied capacities for attuning to others are
elaborated into more complex forms of empathy and related to other
social sentiments and behavioral repertoires still requires much more
ethnographic analysis and specification.