Your coverage of the psychobiological roots of violence in “The Violent Brain,” by Daniel Strueber, Monika Lueck and Gerhard Roth, was interesting and compelling. Although their report seems to be accurate, I find problematic the article's near exclusion of a discussion of social factors involved in violence.

Certainly psychobiology can help explain the behaviors of some chronic violent offenders. Yet these extreme cases are rare; it is far more common to find offenders who commit violence as a result of weak bonds to society, goal frustration or other social problems. In fact, the Dunedin study profiled in the article mentions social factors as a likely reason for continued violence among life-course persistent offenders—not psychobiology.

In addition to being misrepresentative, the article's almost total focus on psychobiological roots of violence is only a short step away from a eugenics argument. Considering that 65 percent of U.S. state and federal prisoners in 2005 were nonwhite (according to the Bureau of Justice Statistics), it would be easy for readers to misinterpret this article as promoting the idea that racial and ethnic minorities are “hardwired” to be violent. This potential harm necessitates that research on the psychobiology of violence be discussed within the context of its limited reach and alongside the social causes of violence that criminologists have been promoting for the past century.

Aaron Kupchik
Department of Sociology and Criminal Justice University of Delaware


“Brain Scans Go Legal,” by Scott T. Grafton, Walter P. Sinnott-Armstrong, Suzanne I. Gazzaniga and Michael S. Gazzaniga, points out the problems involved in using brain scans in criminal cases. But these concerns should not discourage their use for civil cases, in which the standards of proof are significantly lower (that is, “more likely than not” rather than “beyond a reasonable doubt”).

An especially important civil application could be determining whether a patient is suffering from pain. HMOs and disability insurers frequently refuse to honor their policies for patients with chronic pain because they say the pain cannot be “objectively demonstrated” and as such is subject to abuse by malingerers and drug seekers. Legitimate patients are thus often denied proper treatment for their symptoms.

Many reports in the literature over the past several years have demonstrated that on a PET or fMRI scan, activation of the somatosensory cortex indicates the sensation of pain and activation of part of the anterior cingulate cortex indicates the emotional aspects of pain. Other reports document that telling a lie activates other regions of the anterior cingulate cortex and, not too surprisingly, parts of the frontal lobe that are connected with creativity.

Such scans might provide the “objective” evidence of pain that insurers desire, while giving them the confidence that they are not paying to treat fakers.

Harvey S. Frey
Santa Monica, Calif.


In “Taking a Closer Look” [Facts and Fictions in Mental Health], Scott O. Lilienfeld and Hal Arkowitz give an informative and enthusiastic overview of eye movement desensitization and reprocessing (EMDR). The questions they pose are good ones, but the answers they give deserve some clarification.

The article says, for instance, that EMDR patients must repeatedly visualize the traumatic material, suggesting that EMDR is therefore just another exposure therapy. It is true that at the beginning of treatment, the EMDR subject will target the original traumatic scene, but it is misleading to leave it at that. Unlike exposure therapy, EMDR does not repeatedly return the client's attention to the original event. Rather the flow of experience can lead in many directions: past, present or future; emotional, physical, cognitive or perceptual. The job of the EMDR therapist is to make sure this spontaneous experience occurs freely, without editing, manipulation or interpretation. EMDR reprocessing includes associated feelings and perceptions. Whether to call this adaptation, exposure, desensitization or habituation is not clear.

Several comparison studies and meta-analyses show that EMDR works as well as cognitive and behavioral exposure therapies and better than purely supportive therapies. EMDR has received the imprimatur of the American Psychiatric Association and the Department of Defense and Veterans Affairs for treatment of posttraumatic stress disorder (PTSD).

Although we do not yet know how EMDR works, it is more important that it does work. EMDR does not demand that the client describe the event in detail, so it is beneficial to people who cannot—or prefer not to—do so. EMDR does not require learning new skills or habits. In less severe cases, EMDR may require only one or two sessions. Compare this with 30 to 100 hours of homework in cognitively oriented therapies. When cognitive skill and structured activity are important, other trauma therapies may be preferable. But for people who can quickly move into their memories and associations, EMDR may be a better match.

Robert S. Marin
Department of Psychiatry University of Pittsburgh School of Medicine

The authors state that EMDR is not more effective than standard behavioral and cognitive-behavioral therapies. I have heard that EMDR is less stressful than standard therapies for PTSD and thus has a lower dropout rate and that this difference is not taken into account in studies of EMDR (because only patients who complete therapy are included in the studies). Is there indeed a difference in dropout rates?

Phil Thompson
Los Altos, Calif.

ARKOWITZ AND LILIENFELD REPLY: We received many spirited letters raising a variety of intriguing questions concerning the efficacy of EMDR. Yet none present data challenging the central conclusions of our column—namely that EMDR is no more efficacious than behavioral and cognitive-behavioral therapies that rely on exposing clients to anxiety-provoking stimuli and that the eye movements of EMDR do not contribute to its efficacy.

As Marin notes, EMDR may work better for certain individuals than standard exposure-based treatments, and we encourage research to investigate this possibility. Such data could help us choose whether to use EMDR or traditional exposure for specific clients.

Thompson's letter raises the useful question of whether dropout rates are lower in EMDR than in comparable therapies. A 2004 meta-analysis of 25 studies by Elizabeth A. Hembree of the University of Pennsylvania and her colleagues found no significant differences in dropout between EMDR and other behavioral and cognitive-behavioral therapies.

In our columns, we rely on sound scientific data to shed light on the controversies in mental health. Although personal testimonials and organizational endorsements such as those discussed in several of the letters can be thought-provoking, they do not constitute scientific evidence.


With regard to Mariette DiChristina's From the Editor, I have found a counterexample to our remembering bad things rather than good things. We tend to remember well the very good teachers that we have had, whereas we have to try hard to recover even a sketch of the bad ones. Isn't that a happy and telling exception?

Mark Economos
Scarsdale, N.Y.