Imagine you are serving on a jury: the defendant is charged with murder, but he also suffers from a brain tumor that causes erratic behavior. Is he to be held responsible for the crime? Now imagine you are the judge: What should the defendant’s sentence be? Does the tumor count as a mitigating circumstance?

The assignment of responsibility and the choice of an appropriate punishment lie at the heart of our justice system. At the same time, these are cognitive processes like many others—reasoning, remembering, decision-making—and as such must originate in the brain. These two facts lead to the intriguing question: How does the brain enable judges, juries, and you and me to perform these tasks? What are the neural mechanisms that let you decide whether someone is guilty or innocent?

A recent study published in the December 2008 issue of the journal Neuron, by Joshua Buckholtz and his colleagues at Vanderbilt University tackles exactly this question. Until recently, such topics would have been out of the reach of cognitive neuroscience for lack of methods; today, functional magnetic resonance imaging (fMRI) allows researchers to watch the brain “in action” as normal human participants make decisions about responsibility and punishment. In the new study, Buckholtz and colleagues asked participants to read vignettes describing hypothetical crimes that a fictitious agent, “John,” commits against another person. The stories were divided into three conditions: in the first, the “responsibility” (R) condition, the perpetrator was fully responsible for the negative consequences of his action against the victim; for instance, John might have intentionally pushed his fiancée’s lover off a cliff. In the “diminished responsibility” (DR) condition, mitigating circumstances were present that reduced John’s responsibility; imagine that John committed the same crime, but suffered from a brain tumor.

And finally, the “no crime” (NC) condition consisted of stories that did not describe crimes. The participants had to make judgments regarding the degree of punishment that John should receive, on a scale from one to nine.

The authors then analyzed the brain activation linked to these judgments. To identify neural correlates of responsibility, they contrasted activation in the R and DR conditions. Note that the stories in two conditions are identical, except for the degree to which John is responsible for his crime. This contrast thus aims to identify which regions of the brain are involved in assigning responsibility for a crime, holding constant the crime itself. Buckholtz and colleagues found a peak of activation in the right doroslateral prefrontal cortex (rDLPFC), a brain region on the top surface of the right frontal lobe that is known to be involved in high-level cognitive processes such as reasoning and decision-making. In addition, this same region was more active when subjects thought a diminished-responsibility crime deserved punishment compared with when it did not.

Thus, these findings suggest that rDLPFC might be involved in assigning responsibility for crimes, or making judgments about appropriate punishments. Based on this finding, one might have expected that activation in rDLPFC should be higher when participants decide that very severe punishments are appropriate. Buckholtz and colleagues found no correlation between neural activation and punishment magnitude in rDLPFC, however, suggesting that this brain region does not directly underlie the decision on the amount of punishment. In contrast, there was some evidence that activation in emotion-related areas, such as the amygdala, correlates with the degree of punishment subjects assign to John: higher punishment scores were associated with higher activation in these regions during the decision period.

Reconciling the Findings

Have we found, then, the brain center for jurisprudence? Probably not: the brain regions identified in this new study, in particular right DLPFC, have previously been highlighted in a number of other studies addressing related but slightly different questions. Unifying patterns do exist, however. We therefore first describe some related studies, and then outline a possible reconciliation between the different findings.

What does rDLPFC do when it isn’t busy assigning responsibility for crimes? One answer comes from a study by Alan Sanfey and colleagues in 2003: these authors found activation in rDLPFC when subjects decided whether to accept or reject a low offer in a two-person economic game called Ultimatum Game. In addition, Daria Knoch and her colleagues in 2006 found that when rDLPFC was deactivated with a technique called repetitive transcranial magnetic stimulation (TMS), participants became less able to reject low offers in this game, although they still judged these offers as very unfair. A different line of work by Joshua Greene and colleagues in 2004 suggests that rDLPFC may be involved in moral reasoning. They presented participants with moral dilemmas such as the decision whether or not to kill one’s own crying child to keep it raising the attention of enemy soldiers and thereby endangering the whole group. The rDLPFC region was activated when subjects acted in the interest of greater overall welfare, against their emotional impulses. Finally, rDLPFC was also highlighted by another study involving social decision-making by Manfred Spitzer and colleagues in 2007: these authors asked participants how much of their wealth they wanted to share with another player. This amount wasn’t very much, usually—unless participants were threatened with punishment. Under the punishment threat, participants transferred more money, and rDLPFC was more active. Moreover, the more subjects changed their behavior under the punishment threat relative to the situation without a threat, the more rDLPFC was activated, suggesting that rDLPFC played a key role in adapting behavior when facing the sanctioning threat.

The Big Picture

How does the new study fit together with the previous ones, and to what extent is a unifying interpretation of the role of rDLPFC across all these studies possible? The findings described above are all consistent with a role for rDLPFC in inhibiting what psychologists call “prepotent responses,” such as knee-jerk reactions. Rejecting a low Ultimatum Game offer means losing money and thus requires overriding the impulse to accept the money. Making utilitarian rather than emotion-driven moral decisions, and resisting the impulse to make low transfers to one’s partner require the suppression of the impulses to save one’s baby, in the one case, and to keep the money, in the other. The activation of rDLPFC in these studies is consistent with the view that rDLPFC is involved in overriding such responses. In addition, the TMS study mentioned above even suggests a key causal role of rDLPFC for overriding prepotent impulses because when rTMS inhibits the recruitment of rDLPFC, subjects are less able to resist the temptation to accept positive, yet unfair, money offers.

Is the new study by Buckholtz and colleagues consistent with this view of the role of rDLPFC? The crucial new element of this study is the fact that participants were instructed to determine “appropriate” judgments of punishment from a third-person perspective. Thus, participants acted like criminal judges; the fact that their punishment decisions correlated strongly with the prison sentences they deemed appropriate for the crimes in question attests to the fact that they saw themselves in this role. Inasmuch as judges are expected to act impartially and objectively, this task, too, requires the suppression of prepotent responses to the crimes described, to produce “just” and impartial punishments. Reading stories about severe crimes may well cause arousing emotional responses that may be associated with a strong desire to punish. Indeed, Buckholtz and colleagues report that activation in the amygdala correlates with punishment judgments, consistent with the role of this brain region in the representation of arousing emotional events. However, the demands of impartiality often require overriding these impulses to produce a reasonable judgment. The higher activation of rDLPFC in the condition where John is responsible for his crime, and when participants decide if and how much to punish him, is therefore consistent with a role for rDLPFC in the suppression of prepotent emotional reactions.

Although it is unlikely that rDLPFC will enter the textbooks as the seat of the law in the brain, this intriguing new study by Buckholtz underscores the role of this region in high-level cognitive processes in general, and judgment and decision-making in particular. More specifically, it illustrates that third-person judgment situations, such as those used in their study, may rely on similar neural mechanisms as two-person economic and social exchanges, and highlights rDLPFC as a candidate for the neural suppression of impulsive reactions in these situations. So, should you punish John for the murder, despite his brain tumor? Of course! says your gut. Not so fast, says your rDLPFC.

Are you a scientist? Have you recently read a peer-reviewed paper that you want to write about? Then contact Mind Matters editor Jonah Lehrer, the science writer behind the blog The Frontal Cortex and the book Proust Was a Neuroscientist. His next book, How We Decide, will be available in February 2009.