Marco Iacoboni, M.D., Ph.D.
David Geffen School of Medicine at University of California, Los Angeles
Introduction
The social sciences have documented the contagion of violence with carefully controlled studies, including longitudinal studies over long periods of time. Indeed, some have proposed for the contagion of violence a model that mimics the spreading of infectious diseases [for both these issues, see other contributors to this workshop summary]. This model captures well the phenomenon of contagion associated with violent behavior. The model, however, does not provide a biological mechanism that can plausibly account for the spreading of the behavior. Infectious diseases such as the flu have well-defined and well-studied causes, that is, the viruses that spread the flu from individual to individual. The missing link between the compelling social science studies on contagion of violence and the model of such contagion as an infectious disease is a biologically grounded mechanism. A recent neuroscience discovery, a type of brain cell called mirror neuron, may provide such a missing link. This paper summarizes what we know and do not know yet about mirror neurons and discusses the empirical findings from the neuroscience labs in light of potential implications for policy regarding the contagion of violence and its control.
Mirror Neurons: Original Findings
Mirror neurons were reported for the first time in the scientific literature exactly 20 years ago [Dipellegrino et al., 1992]. The scientists who discovered mirror neurons were investigating a region of the monkey brain that controls actions with the hand [e.g., grasping an object, holding it, manipulating it, and so on], and actions with the mouth [e.g., as ingestive actions like biting and drinking, but also facial gestures like lip smacking, a social communication gesture of positive valence in monkeys] [Gentilucci et al., 1988; Rizzolatti et al., 1988]. All of these actions are essential for normal everyday functioning. The scientists were studying the responses of the neurons, while the monkeys were performing those actions, to better understand how the brain controls motor behavior. Unexpectedly, the scientists found that some of the neurons were activated not only when the monkey was performing the action, but also when the monkey was simply observing somebody else making the same action. For instance, some grasping neurons activate when the monkey grasps a tiny object like a raisin [this type of grasp is called precision grip and is performed with the thumb and the index finger], but do not activate when the monkey grasps a large object like a banana [this type of grasp is called whole-hand prehension and requires the use of all fingers and the palm, too]. Among these grasping neurons for precision grip, there were some that activated when the monkey did not move at all, but simply watched somebody else grasping a tiny object [not necessarily a raisin, but any kind of tiny object] with a precision grip. The activity of these cells nearly suggested that while watching other people busy with their own activities, the monkey appeared to be seeing her own actions reflected by a mirror. Hence, the scientists decided to call these brain cells mirror neurons [Gallese et al., 1996].
The early studies on mirror neurons focused on the brain region in which these cells were originally discovered. These early studies demonstrated that there are two main classes of mirror neurons. While approximately one-third of mirror neurons activated for exactly the same action, whether performed or observed [these are called strictly congruent mirror neurons], about two-thirds of mirror neurons also fired for other kinds of observed actions [these are called broadly congruent mirror neurons] [Gallese et al., 1996]. These neurons would activate as long as the observed action achieved the same goal of the performed action. This property suggests that these cells implement a fairly sophisticated mapping of the perceived actions of other people onto the motor repertoire of the perceiver. But how sophisticated is this mapping? Studies have shown that mirror neurons can activate for observed actions that are not completely in sight [that is, vision is partially occluded] [Umilta et al., 2001] and for simply listening to the sound of the action [e.g., breaking a peanut] [Kohler et al., 2002]. The most compelling of these studies demonstrate that the majority of mirror neurons do not even code the action itself [e.g., grasping], but rather the intention associated with it [e.g., grasping to eat rather than grasping to place in a container] [Fogassi et al., 2005].
All these data suggest that when we watch other people's activities, mirror neurons automatically make our own motor system active as if we are performing those activities. This seems a wonderfully efficient mechanism for imitation, which is a fundamental behavior for learning and transmission of culture, and possibly for empathy. However, the mirror mechanism in the brain also suggests that we are automatically influenced by what we perceive, thus proposing a plausible neurobiological mechanism for contagion of violent behavior.
Recent Developments in Mirror Neuron Research
While the early studies on mirror neurons focused on hand and mouth actions, more recent studies have demonstrated the existence of mirror neurons for other kinds of actions [or specific aspects of the observed action] and most importantly in many brain regions. This new wave of studies suggests that the neural mirror mechanism is rather diffuse and pervasive.
In monkeys, three different labs have reported mirror neurons for reaching movements in two different brain areas [Cisek and Kalaska, 2004; Pfeifer et al., 2008; Dushanova and Donoghue, 2010]. Mirror neurons have also been reported for eye movements [Shepherd et al., 2009]. The neurons that code for eye movements tend to have a preferred direction. That is, some neurons activate for eye movements toward a specific sector of space, but not others. Mirror neurons for eye movements do the same. When the monkey is simply watching another monkey looking in the preferred direction of the neuron, the neuron activates as if the monkey was moving the eyes toward that direction. This mirroring mechanism may be important for gaze following and joint attention, two foundational behaviors for the development of social cognition.
Single-cell recordings require invasive brain surgical procedures and are obviously performed in experimental animals, but not in human subjects volunteering for research experiments. However, in some rather exceptional situations, it is possible to piggy-back on existing medical procedures to obtain recordings of individual cells in the human brain. A recent study indeed was able to do so [Mukamel et al., 2010]. The subjects of the study were patients with epilepsy who did not respond well to medications. In these situations, it is appropriate to treat epilepsy with brain surgery, in which the neurosurgeon removes the pathological brain tissue and spares the healthy tissue. To determine the epileptic focus or foci, the surgeon implants depth electrodes into the brain. While in the hospital, the patient stops taking medications and eventually seizes, thus allowing the electrodes implanted in the depth of the brain to show the surgeon exactly where the pathological brain tissue is.
Typically, this procedure only requires the registration of the electroencephalograph [EEG] signal that allows the surgeon to localize the affected tissue. However, with a slight modification of the electrodes used for this procedure, it is also possible to record the activity of individual neurons from the brain of patients. A recent study that recorded for the first time individual mirror neurons in humans reported mirror neurons in two areas that were previously not known to have these kinds of brain cells [Mukamel et al., 2010]. Note that the location of the electrodes in the study on neurosurgical patients is exclusively dictated by medical considerations, not by research questions. Thus, the study in human neurosurgical patients did not record at all from brain areas in which mirror neurons were found in the monkey brain. The two areas in which mirror neurons in humans were found are known to be important for initiating action and for memory. Mirror neurons in a brain region known to be important for memory suggest that we mirror the actions of others in a rather rich fashion. That is, when I watch somebody else grasping a cup of coffee, my brain not only mirrors the motor plans to perform the same action, but also retrieves memory traces of my previous grasping actions. This neural mechanism provides an “inner imitation” of the behavior of other people, which most likely allows us to learn by observation and imitation, and to empathize with others. However, it also makes us more prone to imitate what we see, thus facilitating social contagion.
The Study of the Mirror Neuron System with Noninvasive Techniques
So far we have discussed data from single-cell recordings in monkeys and, in one study, in humans. These data are probably the most compelling data one can obtain in neuroscience. However, single-cell recordings require invasive brain surgery, and their use in humans is obviously extremely limited. There is a large body of scientific literature on the study of mirror neurons in humans that uses noninvasive forms of brain investigation. The four main techniques used are functional magnetic resonance imaging [Iacoboni et al., 2005], EEG [Oberman et al., 2007], magneto-encephalography [Hari et al., 1998], and transcranial magnetic stimulation [Fadiga et al., 1995; Aziz-Zadeh et al., 2002]. Although the data obtained with these techniques are not as compelling as the data obtained with single-cell recordings, they are still extremely valuable. These techniques make it possible to study healthy subjects and neuropsychiatric patients, and to correlate the activity recorded in the brain with behavioral variables.
The data from this vast literature seem to confirm the initial intuitions about the role of mirror neurons in social behavior. Human brain areas with mirroring properties have been associated with imitation [Iacoboni et al., 1999] and empathy [Carr et al., 2003]. Indeed, some studies show correlations between individual differences in empathy and activity in mirror neuron areas [Pfeifer et al., 2008]. The more empathic the subject is, the higher the activity during imitation [Pfeifer et al., 2008], or simply during observation of actions [Kaplan and Iacoboni, 2006] or emotional facial expressions of other people [Pfeifer et al., 2008], including pain [Avenanti et al., 2005]. Furthermore, patients who find social interactions difficult, such as patients with autism spectrum disorders, seem to show reduced activity in mirror neuron areas [Dapretto et al., 2006]. These data support the idea that mirror neurons are important for the effortless, automatic understanding of the mental states of other people [Iacoboni, 2009], and may also be the basis of automatic imitation [Cross and Iacoboni, 2011].
Control of Mirroring to Prevent Contagion of Violence
If we have a mechanism in the brain that automatically activates our own motor system when we see others performing actions, we should also have a control mechanism to avoid continuous automatic imitation. Indeed, while humans tend to imitate others automatically and subconsciously, they tend to do that in a subtle way, imitating postures or the onset of movements [when I reach for the glass you may reach for the napkin], without overtly parroting the behavior of other people. Our behavior would be highly dysfunctional if we were imitating each other all the time. For instance, even during conversation humans tend to imitate each other, often using the same grammatical structures or noun selection [if we are talking about furniture in the living room and I say sofa, it is highly unlikely that the person talking to me will use a synonym like couch; that person most likely will also use the word sofa] [Garrod and Pickering, 2004; Pickering and Garrod, 2007]. However, we do not repeat word for word what the other person has just told us. What are the mechanisms and neural systems for control of mirroring?
The evidence, albeit not conclusive yet, points to a number of potential mechanisms for control of mirroring. Neurological patients with prefrontal lesions show imitative behavior, the rather dysfunctional tendency to imitate whatever other people do in front of them. The lesions that produce this rare behavior are very large, suggesting that multiple brain centers may be involved in mirroring control [Lhermitte et al., 1986; De Renzi et al., 1996]. Some imaging studies indeed suggest that multiple brain areas in the frontal lobe may implement some type of control of mirroring [Brass et al., 2005; Bien et al., 2009]. The differential role of these areas is unclear. Finally, other imaging data suggest that in some situations control is implemented by reconfiguring the connectivity among many different brain systems important for sensory-motor behavior [Cross and Iacoboni, 2011].
The study of the mechanisms of control of mirroring is potentially extremely important. If we can understand how the brain implements control of mirroring, we can in principle intervene and modulate its activity. In some cases, as in the case of autism, it may be beneficial to reduce control and increase mirroring. In some other cases, as in the case of individuals exposed to violent behavior who may be involved in spreading contagion of violence, it may be beneficial to increase control of mirroring, thus reducing imitative violence and possibly preventing the spreading of contagion of violence [Iacoboni, 2008].
Conclusions
Mirror neurons provide an important missing link between the social science data on contagion of violence and the model that draws similarities between contagious mechanisms in infectious diseases and contagion of violence. They provide a neurobiologically grounded mechanism that is fairly automatic and reflexive [albeit not entirely reflexive, of course]. It is important to pay attention to the neuroscience data because they suggest forms of human automatic behavior that require careful consideration when planning interventions and policy that attempt to reduce contagion of violence.