University of Pennsylvania neuroscientist Adrian Raine, the author of “The Anatomy of Violence,” believes that advances in brain imagery are helping to explain the biological roots of crime. American Enterprise Institute scholar and psychiatrist Sally Satel, a coauthor of “Brainwashed,” is wary of the seduction of brain scans. The Washington Post brought them together for a conversation about the promises and pitfalls of brain imagery. An abridged version of that conversation follows.
Outlook: Adrian, could you start us off with one of the images from your work? Tell us what we’re seeing and what some of your research suggests.
Adrian Raine: One that strikes me is Donta Page, who robbed, raped and killed a young woman in Denver in 1999. I was an expert witness in that case. Compared to normal controls, brain scans revealed he had a distinct lack of activation in the ventral prefrontal cortex: the region that helps regulate our emotions and control our impulses.
He was also brought up in a horrible environment, neglected and physically and sexually abused. He was thrown out a car window when he was just 9 months old and suffered multiple head injuries as a child. He also had a family history of mental illness. He was referred 19 times for psychological treatment, but he never once got a treatment session.
So the key conceptual point is biosocial: Combine the brain with the social environment, and you have a predisposition for violence and crime that should be taken into account. [A three-judge panel gave Page a life sentence rather than the death penalty.]
Outlook: Sally, talk about some of your concerns about brain imaging and how it’s used. You’ve written, “Naive media, slick neuroentrepreneurs, and even an occasional overzealous neuroscientist exaggerate the capacity of scans to reveal the contents of our minds, exalt brain physiology as inherently the most valuable level of explanation for understanding behavior, and rush to apply underdeveloped, if dazzling, science for commercial and forensic use.”
Sally Satel: Neurocentrism, as [my coauthor Scott Lilienfeld and I] define it, is the notion that explanations that reside at the level of the brain are inevitably the most informative, authentic, truthful explanations of complex behavior. Brain imaging can inadvertently, if it’s not interpreted correctly, or if it’s positioned in a tendentious way, feed into that [bias], because you have this stunning biological portraiture. Things appear to be lit up.
That’s really not how the brain works. It’s always on, it’s always firing — some circuits are going to be more active at any one time than another, but [neural activity] is highly distributed and not static.
Many people think that if [there is activity] in the brain, whatever behavior flows from it is involuntary. Sometimes that is the case, but you cannot draw that inference just from looking at a brain scan.
There’s also a lot of misapplication, a lot of premature application of underdeveloped science. Using [brain imaging] in the service of lie detection is one. It doesn’t mean we won’t get much better to the point where it may well be informative. But it’s not there yet, and there’s a lot of pseudo-neuroscience out there.
Outlook: The Supreme Court ruled this past week that police can take DNA samples from people arrested in connection with serious crimes. How about brain imagery? What role do you see brain scans playing in law enforcement and criminal justice, now or in the future?
Raine: Not now, but there’s potential for the future in the area of probation and parole decisions: Do we keep someone in prison, do we let them out early, are they a risk to society? Those decisions are made every day by judges, by [probation officers]. In California, it’s based on 20 indicators — social and behavioral indicators. Question is: Can these assessments be more accurate?
Two new studies have come out — one that I’m linked to and one by another group — that show brain-scan data can give added value to social and behavioral predictors of future offending. One study, this was [University of New Mexico psychologist] Kent Kiehl’s group, showed that individuals with lower levels of anterior cingulate functioning are twice as likely to reconvict after release from prison. The study I was involved in showed that individuals with reduced volumes of the amygdala were three times more likely to commit an offense in the three-year period post-release. Now, caution here, these are just two studies. But if there’s replication and extension, then there’s some potential for better decisions to be made.
Satel: I’m looking at the study you’re referring to with Kent Kiehl, and while it was a well-done study, they found what’s to me an unacceptable number of false positives and false negatives: 40 percent of people who they thought would reoffend based on reduced activity in the anterior cingulate cortex did not reoffend, and almost 50 percent of the people who they didn’t expect to offend based on the activity in that one area did offend. So that’s a not a lot better than chance.
Raine: I would say even if you statistically raise the prediction a little bit, that’s got to be better than nothing.
I think brain imaging and other neurobiology can also be used in mitigation in capital cases. If you’re against the death penalty, well, you’re in search of anything that can be thrown into that situation. Neuroscience has something to add in the penalty phase of capital cases. Outlook: If there is some biological basis for violence, what does that do to concepts of moral responsibility?
Raine: I would say it gives us more humanity. We need to understand behavior more. It will never be a full explanation. But just as with witches — we used to burn them. Once we find out more about the etiology and causes of behavior, it gives us a more benign and humane perspective.
I talk about the case of Michael, who was a 40-year-old man, married. And then he began to develop a sexual interest in his prepubescent stepdaughter. Michael was found guilty of child molestation and diagnosed with pedophilia. The night before his prison sentence, he went to hospital complaining of a headache. An astute neurologist brain-scanned him and found a tumor growing from the base of the orbital frontal cortex — a part of the brain very critical for emotional regulation. They resected the tumor, and Michael was completely normal. He successfully completed a [therapy] program and then went back home to live with his wife and stepdaughter. Sort of a happily ever after. But then the headaches came back. He began to develop interest in child pornography again. The tumor had grown back. They resected this tumor for the second time, and for six years after that, to our knowledge, Michael’s been quite normal. It’s as close as you can get to causality.
Satel: I’m all for humanity. But the case of the pedophile: That is an amazing case. However, one wonders how many of those there are.
Raine: Or take head injury: It changes behavior.
Satel: Yes, but anything dramatic like that I don’t think people question so much. My point about the pedophilia case is that the night he [went to the hospital], at one point he was afraid that if he were released he might rape his landlady.
Raine: Yes, knowledge and forethought.
Satel: This man’s moral sense was intact enough for him to stop it. And that’s always interested me because of so many of these cases where people claim that they couldn’t control themselves. All you know is that they didn’t control themselves. Our science is not good enough yet to tell us when they truly couldn’t control themselves.
Raine: I think there are degrees of selfcontrol. You know what Michael said? I’ll try and use his exact words. He said: “There was a little voice in the back of my head saying, ‘You shouldn’t do this.’ But there was a much bigger voice in the front of my head saying, ‘Go on, why not?’ ” So he did have a sense. He knew what he was doing, and he knew that what he was doing was wrong. That’s why, when I put that case to judges, 90 percent of judges will say Michael is responsible for his behavior, because he fits the legal definition as it stands.
In the future, what about partial responsibility? Not just based on brainimaging data, but on all data combined. You know, you’re either responsible or not responsible: I don’t buy into that. I think there are shades of gray. And I think that there’s a future potential for a change in the justice system in terms of bringing the concept of partial responsibility into the guilt phase of a trial.