Neuroscience, ethics, and the law

Advances in neuroimaging techniques have the potential for combating crime. Edmund Naylor and colleagues explain the philosophical implications

Advances in neuroscience have important implications in a legal setting, adding fuel to a centuries old debate over the nature of free will and offering the possibility of “mind reading” after and even before a crime has been committed. These issues have important ethical consequences.

Libertarian free will—the concept that an individual human can be the “uncaused cause” of their actions—initially appears central to legal responsibility, but its existence has long been controversial. The concept requires acceptance of the dualist belief that mind and brain are separate, implying that the conscious, the “ghost in the machine,” is able to decide on actions independently of external influences.

If one takes the determinist view that the mind is a construct of the brain then libertarian free will is impossible. The brain is a physical object in a physical universe, subject to the laws of causality: thus all neural impulses and eventual decisions are predetermined. Every action has a reaction even if separated by complex neural processing. If the latter argument is accepted then every defendant could legitimately claim the automaton defence: as Gazzaniga aptly put it: “I had no choice but to murder—my brain made me do it” ^w1

When science meets philosophy

Neuroscience aims to elucidate the mechanisms of brain function—in other words, determine the physical causality of brain processes. As a result, neuroscience is often used by those arguing for a determinist explanation of the human condition, adding hard evidence to what was previously a purely philosophical debate. Libet showed that neuronal impulses initiating conscious action are detectable hundreds of milliseconds before the perception of a conscious decision, suggesting to some that consciousness is an illusion.^w2 This, and other similar evidence, casts doubt on the extent of libertarian free will.

Others question the impact of this evidence on the dualism-determinism debate. Neuroscience has yet to come close to explaining integrated command and response mechanisms within the brain: why, for example, a pupil raises his arm when told to by his teacher.^w3 Until neuroscience can explain the mind as neural mechanisms it does not justify giving up the dualistic explanation: we do, after all, perceive our own mind and brain as separate.

Other philosophers argue that determinism cannot threaten human free will, no matter the extent of neuroscientific evidence for the neural mechanisms of the mind. Compatibilists consider that free will can exist despite determinism, with free will lost only if an individual is forced to act by another person. As a result, our actions are free despite being predetermined: it is a fallacy to consider our actions imposed on us by the laws of nature.

The consequentialists provide another powerful argument, asserting that libertarian free will is not necessary for responsibility. From a consequentialist viewpoint, laws, and specifically punishments for breaking them, play a part solely in improving an individual’s future actions, enabling society to function for overall benefit. In this system, responsibility is determined only by an ability to understand the law and know the consequences of breaking it; as a result the equation of personal responsibility with libertarian free will falls. Thus, the extension of the determinist argument to absolve legal responsibility is hardly without challenge.

Potential of neuroimaging

Advances in our ability to probe the living brain may have a more practical impact on legal practice. Functional neuroimaging has led to psychiatry exploding into a “hard” scientific discipline. Imaging studies have shown that psychiatric illnesses are associated with characteristic changes within the brain, correctable by pharmacological or psychological treatment. This promises to revolutionise our understanding of these conditions and improve treatments.

The extension of psychiatric neuroimaging into the legal field has revealed abnormalities in the brains of antisocial individuals, showing an association with various crimes.^w4 This raises the possibility that individuals with a certain combination of structural and metabolic abnormalities could be pre-emptively imprisoned because of a possibility of future criminality. If we take the same argument to its opposite extreme, those accused of crimes could cite similar abnormalities, detected by neuroimaging, in their defence—if an individual’s brain predisposes them to violence by its structure, then are they less culpable for their crime?

These potential uses of advanced neuroimaging in a legal setting raise several important objections. It must be remembered that neurophysical states are in themselves neither bad nor good^w5: they may be only correlates of bad or good actions. Currently there is little evidence for the predictive validity of such tests, making any use in pre-emptive sectioning unsupportable. For predictive tests to be introduced, the predictive value of any test would have to be established, with society deciding how specific such a test should be: would 10% false positives be acceptable? Or 1%?

Pre-empting crime?

The use of neuroinvestigation in pre-emptive imprisonment also raises objections on the grounds that it would violate certain fundamental rights. It could lead to a new level of invasion of patients’ mental privacy, dangerously challenging free thought and speech and going so far as to introduce Orwellian “thought crime.” As there is presently little consensus on which brain structures and activity could be construed as being either “criminal” or abnormal, the use of brain scans to either pre-empt crime or acquit a suspect seems unlikely in the near future.

In spite of these reservations, the careful use of neuroimaging in this area could have an important positive effect. If treatment rather than imprisonment were given to individuals with neurophysical states suggestive of likely future criminality or indeed to those already convicted of crimes, then a positive outcome could be achieved for the individual and society. Ethical issues would remain over imposition of treatment in this situation, however, and the classification of individuals as physiologically hard wired to commit future crime would cause controversy.^w6

Functional neuroinvestigation may also have a role in assessing quality of testimony (see box). Researchers have reported the use of functional magnetic resonance imaging (fMRI) to detect lying, albeit in an experimental setting, to an accuracy of more than 90%.^w7 In comparison, traditional polygraphs have no established reliability or accuracy and are highly operator dependent. Neuroinvestigation could also be used to determine whether an individual had been present at the crime scene: Electroencephalograms have been shown to indicate if a person is familiar with a particular situation.^w1 These legal uses of neuroscience could have a positive effect by increasing the accuracy of conviction.

The accuracy of neuroinvestigations in determining the veracity of witness testimony remains an issue, however, and would have to be proved in situations where subjects are highly motivated to lie. A further objection to the use of neuroscience in court arose when a US judge concluded that the use of neuroscientific evidence should be limited on the grounds that it threatens the jury’s role in determination of credibility in a trial.^w8 It is important that neuroimaging based analysis is viewed only as a further piece of evidence to be assessed by the jury, rather than as a direct insight into the defendant’s mind.

What lies ahead

Recent neuroscientific advances have important ethical implications in relation to the moral basis of the law and in the more practical realm of legal practice. While neuroscience helps to bring the determinist challenge to free will into focus, it does not currently threaten legal responsibility: there are powerful arguments that responsibility remains in spite of determinism. More robust testing of the predictive and diagnostic power of functional neuroimaging will be required before its legal use can be increased, and even then ethical objections will remain. Neuroscience can have a positive effect in the legal field, but careful navigation will be required to achieve it.

fMRI: coming soon to a court near you?

fMRI based lie detection has yet to make the jump from laboratory to courtroom, but researchers predict that its accuracy will soon reach a level at which its use will be irresistible.^w9 Two US companies, No Lie MRI and Cephos Corporation, are developing fMRI lie detection with the aim of using it in the legal setting
For the companies to succeed in moving fMRI lie detection into the US courtroom, the technique must fulfil the Daubertcriteria, which state that a scientific technique should have general acceptance from the scientific community and a known error rate before it can be used as evidence
Scientists in the field believe that once a 95% rate for lie detection is reached, general acceptance will follow; in comparison the polygraph machine, which is largely excluded from US courts, is right between 60% and 85% of the time.^w9 With recent evidence suggesting that fMRI is around 90% accurate in detection of lying in an experimental setting,^w7 this 95% threshold does not seem far off. In the United Kingdom there is no equivalent of the Daubertcriteria, potentially making courts more accessible to fMRI on this side of the Atlantic. None the less, cultural objections to lie detection will remain a barrier in both countries^w8
A second direction for No Lie MRI and Cephos’s energies is towards the large market for commercial lie detection: 40,000 polygraphs are carried out in the US each year by insurance companies and government agencies,^w10 and the use of lie detection to manage sex offenders has been trialled in Britain. In 2006 a South Carolina delicatessen owner, accused of arson by his insurance company, used No Lie MRI to support his claim of innocence in an attempt to make the insurers pay up.^w11 The out of court use of fMRI is likely to spread quickly, especially if a US loophole allowing fMRI but not polygraph testing of potential employees can be exploited^w12

Competing interests: None declared.

Provenance and peer review: Not commissioned; externally peer reviewed.

References w1-12 are on studentbmj.com.

Edmund Naylor fifth year medical student Oxford University Medical School, New College, Oxford
Email: edmund.naylor@medschool.ox.ac.uk
Daniel Wood sixth year medical student Oxford University Medical School, Queen’s College, Oxford
Sheheryar Kabraji fifth year medical student Oxford University Medical School, Merton College, Oxford
Julian Savulescu Uehiro professor of practical ethics Uehiro Centre for Practical Ethics, OxfordStudent BMJ 2008;1:56-57 | 17

Gazzaniga MS. The Ethical Brain. New York: Dana Press; 2005.
Libet B, Gleason CA, Wright EW, Pearl DK. Time of conscious intention to act in relation to onset of cerebral activity (readiness-potential). The unconscious intention of a free voluntary act. Brain 1983;106 (Pt3):623-42.
Morse SJ. Moral and legal responsibility and the new neuroscience. In: Iles J, editor. Neuroethics. Oxford: Oxford University Press; 2006. p. 33-50.
Eastman N, Campbell C. Neuroscience and legal determination of criminal responsibility. Nat Rev Neurosci 2006;7(4):311-8.
Buller T. What can neuroscience contribute to ethics? J Med Ethics 2006:32:63-64.
Wolfe T. Sorry, but your soul just died. In: Hooking Up. London: Picador; 2000. p. 89-109.
Kozel FA, Johnson KA, Mu Q, Grenesko EL, Laken SJ, George MS. Detecting deception using functional magnetic resonance imaging. Biol Psychiatry 2005:58(8):605-13.
Greely HT. The social effects of advances in neuroscience: legal problems, legal perspectives. In: Iles J, editor. Neuroethics. Oxford: Oxford University Press; 2006. p. 33-50.
Rosen J. The Brain on the Stand. New York Times 2007 Mar 11.
Ireland C. Symposium: ‘Will Brain Imaging be Lie Detector Test of the Future.’ Harvard University Gazette 2007 Feb 8.
McKenna P. Can a brain scan prove you’re telling the truth? New Scientist 2007 Feb 10.
No Lie MRI. Customers – Corporations. [Online]. 2006 [cited 2007 Dec 2]. Available from: URL:http://www.noliemri.com/customers/GroupOrCorporate.htm