In a scene from the well-known movie Minority Reporta husband enters the house and finds his wife with another man. Upon discovering the deception, and armed with a pair of scissors, he sets out to kill her. Then a police "pre-crime" unit, capable of predicting violent crimes before they happen, breaks into the house and prevents the murder.

Although in this case it is science fiction, the use of neuroscientific techniques to try to "read minds" and predict future violent behavior, especially recidivism, has been promoted in several countries for years. Thus, for example, programs of study has proliferated, using functional magnetic resonance imaging (fMRI) and other brain imaging techniques, correlating the Degree activation of certain encephalic regions, such as the anterior cingulate cortex (ACC), with the risk of violence.

Supporters of the so-called "neuroprediction" argue that these subject findings allow the justice system to take measures to prevent recidivism and thus improve public safety. Such measures range from denying parole to even extending sentences beyond those established in a court sentence. However, this subject of interpretations is based on neuroessentialist views of behavior, which ignore the decisive role of the environment and the specific circumstances surrounding a crime. As the well-known neuroscientist David Eagleman and his team remind us, "lives are complex, and crime is contextual".

The problem of neuroprediction is closely related to that of free will. Although it is a very slippery philosophical concept, there are two main positions on the subject when it comes to applying it to the criminal field. One of them is the absolute or retributive view of punishment, which seeks to assign punishments proportional to the seriousness of the crimes committed and undervalues any underlying biological conditioning factors.

Is it all about neurons?

The opposite position is perfectly represented in an allegorical way by the masterly Duelo a garrotazos in which two men, trapped in the mud, seem condemned to beat each other to death.

This is deterministic fatalism, according to which we lack any hint of freedom and responsibility for our actions, which are inevitable. This mentality is certainly compatible with a reductionist approach to the nervous system: everything would be limited to neurons and their physicochemical interactions.

Behind the defense of neuroprediction as a means of avoiding recidivism lies, precisely, a fatalistic view of human behavior; but this defense faces three major objections.

• First: as Alfred Mele explains in his excellent work Freethere are good arguments to defend that neuroscience, contrary to what some neuroscientists claim, has not demonstrated the non-existence of free will.

• Second, neuroprediction is based on a probabilistic estimate of the risk of violence and, as such, does not offer absolute certainty about the future. For example, a leave CCA activity seems to be associated with a higher probability of recidivism, but recidivism does not always occur.

• Third: by assuming that the subject is not free to prevent it, the system would be taking the future crime for granted, and yet it would be granting itself the power to prevent it. This way of understanding a crime as determined and not determined at the same time, besides being paradoxical, could be used to justify tyrannical forms of exercising power.

However, neuroscience can contribute much to the study of recidivism without endorsing neuroprediction or, conversely, ignoring the biological determinants of crime. In a work conducted with a colleague from INACIPE in Mexico, and edited by David Eagleman, we have argued that the best possible approach to this question is "neuroprevention": using neuroscience to reduce the risk of future crime, never to predict it.

Citizen security and prisoner reintegration

Naturally, this prevention should focus on citizen security. But it must be accompanied by intervention strategies designed to improve cognitive skills usually related to criminality (impulsivity, empathy, planning, etc.) that are dynamic, i.e., modifiable.

In this sense, techniques such as fMRI can provide valuable complementary information to the traditional psychological assessment tests applied for decades.

This approach would allow inmates to influence their own future by proactively working towards (if possible) their eventual social reintegration. The role of the system would thus be to accompany their evolution; worlds like Minority Report must remain in the realm of fiction.

Neuroscience can contribute a great deal to the justice system, but we will do it a disservice if we misinterpret its findings, overestimate its scope and adopt neuroessentialist views that ignore the context in which the brain is framed: body, environment and personal relationships. As in so many other major problems, conceptual reflection and dialogue of this exciting discipline with the social sciences and Humanities are essential if we aspire to achieve a more complete (and fairer) vision of human action.

This article was originally published in The Conversation. Read the original.