Imitation, Empathy,

  • View

  • Download

Embed Size (px)

Text of Imitation, Empathy,

  • 7/23/2019 Imitation, Empathy,


    Imitation, Empathy,and Mirror Neurons

    Marco Iacoboni

    Ahmanson-Lovelace Brain Mapping Center, Department of Psychiatry and BiobehavioSciences, Semel Institute for Neuroscience and Social Behavior, Brain Research InstituDavid Geffen School of Medicine at UCLA, Los Angeles, California 90095;email:

    Annu. Rev. Psychol. 2009. 60:65370

    First published online as a Review in Advance onSeptember 15, 2008

    TheAnnual Review of Psychologyis online

    This articles doi:10.1146/annurev.psych.60.110707.163604

    Copyright c2009 by Annual Reviews.All rights reserved


    Key Words

    social cognition, theory of mind, mirror neuron system, embodimen

    AbstractThere is a convergence between cognitive models of imitation, co

    structs derived from social psychology studies on mimicry and empthy, and recent empirical findings from the neurosciences. The ideom

    tor framework of human actions assumes a common representationformat for action and perception that facilitates imitation. Furthe

    more, the associative sequence learning model of imitation propos

    that experience-based Hebbian learning forms links between sensoprocessing of the actions of others and motor plans. Social psycho

    ogy studies have demonstrated that imitation and mimicry are pevasive, automatic, and facilitate empathy. Neuroscience investigatio

    have demonstrated physiological mechanismsof mirroring at single-cand neural-system levels that support the cognitive and social psych

    ogy constructs. Why were these neural mechanisms selected, and whis their adaptive advantage? Neural mirroring solves the problem

    other minds (how we can access and understand the minds of otheand makes intersubjectivity possible, thus facilitating social behavior


  • 7/23/2019 Imitation, Empathy,


    Hebbian learning:

    associative learning isimplemented bysimultaneousactivation of cells thatwould lead toincreased synapticstrength between thecells


    INTRODUCTION .. . . . . . . . . . . . . . . . . 654

    COGNITIVE MECHANISMSOF IMITATION . . . . . . . . . . . . . . . . . . 654

    The Ideomotor Frameworkof Imitation . . . . . . . . . . . . . . . . . . . . . 654

    Associative Sequence Learning . . . . . 656IMITATION AND EMPATHY

    IN SOCIAL BEHAVIOR. . . . . . . . . . 657

    Pervasiveness and Automaticityof Human Imitation. . . . . . . . . . . . . 657

    NEURAL MECHANISMSOF IMITATION . . . . . . . . . . . . . . . . . . 659

    Neural Precursorsin Nonhuman Primates . . . . . . . . . 659

    Macaque Mirror Neuronsand Imitation in Monkeys . . . . . . . 662

    Human Brain Mechanismsof M ir r or in g . . . . . . . . . . . . . . . . . . . . 6 6 3

    Neural Mirroring and Psychological

    Theories of Imitation . . . . . . . . . . . 665WHY NEURAL MIRRORING

    AND IMITATION? . . . . . . . . . . . . . . . 666


    Although mimicry is a pervasive phenomenonin the animal kingdom, imitation certainly

    achieves its highest form in humans. Pastauthorsfor instance, de Montaigne (1575),

    Adam Smith (1759), Poe (1982), Nietzsche(1881), and Wittgenstein (1980)have often

    associated imitation with the ability to em-pathize and understand other minds. The evo-

    lutionary, functional, and neural mechanismslinking imitation to empathy, however, have

    been unclearfor many years.Recently, there has

    been a convergence between cognitive mod-els of imitation, social psychology accounts of

    its pervasiveness and its functional links withempathy and liking, and the neuroscience dis-

    coveries of neural mechanisms of imitationand empathy. This convergence creates a solid

    framework in which theory and empirical datareinforce each other.

    Among cognitive models of imitationideomotor model and the associative sequ

    learning model seem to map well onto rophysiological mechanisms of imitation.

    ideomotor model assumes a common

    resentational format for action and pertion, whereas the associative sequence lear

    model puts at center stage Hebbian learnina fundamental mechanism linking sensory

    resentations of the actions of others to mplans. Furthermore, social psychology stu

    have documented the automaticity of imitaand mimicry in humans, a feature that also m

    well onto some recently disclosed neuropological bases of imitation.

    This review discusses cognitive modelscial psychology constructs, and neural m

    anisms of imitation under the hypothesis

    these mechanisms were selected because offer the adaptive advantage of enabling th

    derstanding of the feelings and mental statothers, a cornerstone of social behavior.


    The Ideomotor Frameworkof Imitation

    Theories of action can be divided intomain frameworks. The most dominant fra

    work may be called the sensory-motor frwork of action. It assumes that actions are

    tiated in response to external stimuli. Inframework, perception and action have i

    pendent representational formats. Stimulibe translated into motor responses by stim

    response mapping mechanisms. This frwork has generated a large literature an

    egant experimental paradigms, as for inst

    the work on stimulus-response compatib(Hommel & Prinz 1997, Proctor & R

    1990). Stimulus-response translational menisms, however, do not easily account fo

    similarity between the observed action andaction performed by the imitator that i

    quired by imitation. Indeed, one of the mproblems of imitation often discussed in

    654 Iacoboni

  • 7/23/2019 Imitation, Empathy,


    literature inspired by sensory-motor modelsis the so-called correspondence problem (Ne-

    haniv & Dautenhahn 2002). This problem

    can be summarized with the question: how isthe sensory input from somebody elses action

    transformed into a matching motor output bythe imitator?

    For the ideomotor framework of action, thecorrespondence problem of imitation is not a

    problem at all. Indeed, the ideomotor frame-work assumes a common representational for-

    mat for perception and action, an assump-tion that makes translational processes between

    stimuli and responses rather unnecessary. Theroots of the ideomotor framework were estab-

    lished by the work of Hermann Martin Lotze

    (Prinz 2005) and William James (1890). Thestarting point of actions, for Lotze and James,

    is not a response to a sensory stimulation, butrather the representation of the goal that the

    agent intends to achieve. When an intentionis unchallenged by a conflicting one, it acti-

    vates the representation of the intended goaland the motor plan necessary to achieve it. The

    coactivation of the intended goal and the mo-tor plan required to achieve itaccording to

    the ideomotor frameworkis the result of ourexperience. We have learned the effects of our

    own actions, and we expect certain effects when

    we perform certain acts. This previous learningmakes it possible that just thinking about the

    intended goal automatically activates the rep-resentation of the action necessary to obtain it.

    Thus, when I think about rebooting my com-puter, I automatically activate the representa-

    tion of the finger movement necessary to pressthe appropriate key.

    The ideomotor framework naturally ac-counts for imitation. According to this frame-

    work, when I see somebody elses actions andtheir consequences, I activate the representa-

    tions of my own actions that would produce

    those consequences. Here, consequences areconstrued in a very broad sense. For instance,

    a simple finger lifting has multiple perceptualconsequences, among them the sight of the fin-

    ger lifting. Thus, simply watching somebodyelse lifting a finger should activate my own mo-

    tor plan to lift the same finger. Brass and col-leagues tested this hypothesis in elegantly sim-

    ple experiments (Brass et al. 2000, 2001). Sub-jects were shown two movements of the index

    finger from the same starting position. In half

    of the trials the finger would move upward,and in the other half it would move downward.

    Subjects were instructed to respond as fast aspossible using their own index finger. Within

    each block of trials, subjects were instructed touse always the same motor response, either an

    upward or a downward movement. Thus, al-though perceptually subjects were seeing both

    upward and downward movements, motoricallythey were only executing one of the two move-

    ments. Given that response selection was notrequired, the identity of the stimulus was com-

    pletely irrelevant for the initiation of the motorresponse. Here, the sensory-motor framework

    would predict similar reaction times for re-

    sponses that were identical to the stimulus (e.g.,upward motor response for a stimulus showing

    an upward finger movement) and for responsesthat were different from the stimulus (e.g., up-

    ward motor response for a stimulus showing adownward finger movement). In contrast, the

    ideomotor framework would predict faster re-action times for motor responses identical to

    the stimulus compared to motor responses dif-ferent from the stimulus. The results demon-

    strated a large chronometric advantage for re-sponses identical to the stimuli, in line with the

    predictions of the ideomotor framework (Brass

    et al. 2000, 2001).The ideomotor framework also predicts that

    goals have higher priority than movements inimitation. Imitation experiments in children

    have confirmed thi