Was it really me?

by 2300 2,300 words
  • Read later or Kindle
    • KindleKindle

Was it really me?

Illustration by Matt Murphy

Neuroscience is changing the meaning of criminal guilt. That might make us more, not less, responsible for our actions

Steve Fleming is a cognitive neuroscientist. He is a postdoctoral fellow at New York University and a blogger at The Elusive Self.

2300 2,300 words
  • Read later
    • KindleKindle

In the summer of 2008, police arrived at a caravan in the seaside town of Aberporth, west Wales, to arrest Brian Thomas for the murder of his wife. The night before, in a vivid nightmare, Thomas believed he was fighting off an intruder in the caravan – perhaps one of the kids who had been disturbing his sleep by revving motorbikes outside. Instead, he was gradually strangling his wife to death. When he awoke, he made a 999 call, telling the operator he was stunned and horrified by what had happened, and unaware of having committed murder.

Crimes committed by sleeping individuals are mercifully rare. Yet they provide striking examples of the unnerving potential of the human unconscious. In turn, they illuminate how an emerging science of consciousness is poised to have a deep impact upon concepts of responsibility that are central to today’s legal system.

After a short trial, the prosecution withdrew the case against Thomas. Expert witnesses agreed that he suffered from a sleep disorder known as pavor nocturnus, or night terrors, which affects around one per cent of adults and six per cent of children. His nightmares led him to do the unthinkable. We feel a natural sympathy towards Thomas, and jurors at his trial wept at his tragic situation. There is a clear sense in which this action was not the fault of an awake, thinking, sentient individual. But why do we feel this? What is it exactly that makes us think of Thomas not as a murderer but as an innocent man who has lost his wife in terrible circumstances?

Automatism implies that the accused person had no control over his actions, that he acted like a runaway machine

Our sympathy can be understood with reference to laws that demarcate a separation between mind and body. A central tenet of the Western legal system is the concept of mens rea, or guilty mind. A necessary element to criminal responsibility is the guilty act — the actus reus. However, it is not enough simply to act: one must also be mentally responsible for acting in a particular way. The common law allows for those who are unable to conform to its requirements due to mental illness: the defence of insanity. It also allows for ‘diminished capacity’ in situations where the individual is deemed unable to form the required intent, or mens rea. Those people are understood to have control of their actions, without intending the criminal outcome. In these cases, the defendant may be found guilty of a lesser crime than murder, such as manslaughter.

In the case of Brian Thomas, the court was persuaded that his sleep disorder amounted to ‘automatism’, a comprehensive defence that denies there was even a guilty act. Automatism is the ultimate negation of both mens rea and actus reus. A successful defence of automatism implies that the accused person had neither awareness of what he was doing, nor any control over his actions. That he was so far removed from conscious awareness that he acted like a runaway machine.

The problem is how to establish if someone lacks a crucial aspect of consciousness when he commits a crime. In Thomas’s case, sleep experts provided evidence that his nightmares were responsible for his wife’s death. But in other cases, establishing lack of awareness has proved more elusive.

It is commonplace to drive a car for long periods without paying much attention to steering or changing gear. According to Jonathan Schooler, professor of psychology at the University of California, Santa Barbara, ‘we are often startled by the discovery that our minds have wandered away from the situation at hand’. But if I am unconscious of my actions when I zone out, to what degree is it really ‘me’ doing the driving?

This question takes on a more urgent note when the lives of others are at stake. In April 1990, a heavy-goods driver was steering his lorry towards Liverpool in the early evening. Having driven all day without mishap, he began to veer on to the hard shoulder of the motorway. He continued along the verge for around half a mile before he crashed into a roadside assistance van and killed two men. The driver appeared in Worcester Crown Court on charges of causing death by reckless driving. For the defence, a psychologist described to the court that ‘driving without awareness’ might occur following long, monotonous periods at the wheel. The jury was sufficiently convinced of his lack of conscious control to acquit on the basis of automatism.

The argument for a lack of consciousness here is much less straightforward than for someone who is asleep. In fact, the Court of Appeal said that the defence of automatism should not have been on the table in the first place, because a driver without ‘awareness’ still retains some control of the car. None the less, the grey area between being in control and aware on the one hand, and in control and unaware on the other, is clearly crucial for a legal notion of voluntary action.

If we accept automatism then we reduce the conscious individual to an unconscious machine. However, we should remember that all acts, whether consciously thought-out or reflexive and automatic, are the product of neural mechanisms. For centuries, scientists and inventors have been captivated by this notion of the mind as a machine. In the 18th century, Henri Maillardet, a Swiss clockmaker, built a remarkable apparatus that he christened the Automaton. An intricate array of brass cams connected to a clockwork motor made a doll produce beautiful drawings of ships and pastoral scenes on sheets of paper, as if by magic. This spookily humanoid machine, now on display at the Franklin Institute in Philadelphia, reflects the Enlightenment’s fascination with taming and understanding the mechanisms of life.

Modern neuroscience takes up where Maillardet left off. From the pattern of chemical and electrical signaling between around 85 billion brain cells, each of us experiences the world, makes decisions, daydreams, and forms friendships. The mental and the physical are two sides of the same coin. The unsettling implication is that, by revealing a physical correlate of a conscious state, we begin to treat the individual not as a person but as a machine. Perhaps we are all ‘automata’, and our notions of free choice and taking responsibility for our actions are simply illusions. There is no ghost in the machine.

Under the influence of alcohol, people become more likely to daydream and less likely to catch themselves doing so

In his book Incognito (2011), David Eagleman argues that society is poised to slide down the following slippery slope. Measurable brain defects already buy leniency for the defendant. As the science improves, more and more criminals will be let off the hook thanks to a fine-grained analysis of their neurobiology. ‘Currently,’ Eagleman writes, ‘we can detect only large brain tumours, but in 100 years we will be able to detect patterns at unimaginably small levels of the microcircuitry that correlate with behavioral problems.’ On this view, responsibility has no place in the courtroom. It is no longer meaningful to lock people up on the basis of their actions, because their actions can always be tied to brain function.

While it is inevitable that defence teams will look towards neuroscientific evidence to shift the balance in favour of a mechanistic, rather than a personal, interpretation of criminal acts. But we should be wary of attempts to do so. If every behaviour and mental state has a neural correlate (as surely it must), then everything we do is an artifact of our brains. A link between brain and behaviour is not enough to push responsibility out of the courtroom. Instead we need new ways of thinking about responsibility, and new ways to conceptualise a decision-making self.

Responsibility does not entail a rational, choosing self that floats free from physical processes. That is a fiction. Even so, demonstrating a link between criminal behaviour and conscious (or unconscious) states of the brain changes the legal landscape. Consciousness is, after all, central to the legal definition of intent.

In the early ’70s, the psychologist Lawrence Weiskrantz and the neuropsychologist Elizabeth Warrington discovered a remarkable patient at the National Hospital for Neurology and Neurosurgery in London. This patient, known as DB, had sustained damage to the occipital lobes (towards the rear of the brain), resulting in blindness in half of his visual field. Remarkably, DB was able to guess the position and orientation of lines in his ‘blind’ hemifield. Subsequent studies on similar patients with ‘blindsight’ confirmed that these responses relied on a neural pathway quite separate from the one that usually passes through the occipital lobes. So it appears that visual consciousness is selectively deleted in blindsight. At some level, the person can ‘see’ but is not aware of doing so.

Awareness and control, then, are curious things, and we cannot understand them without grappling with consciousness itself. What do we know about how normal, waking consciousness works? Hints are emerging. Studies by Stanislas Dehaene, professor of experimental cognitive psychology at the Collège de France in Paris, have revealed that a key difference between conscious and unconscious vision is activity in the prefrontal cortex (the front of the brain, particularly well-developed in humans). Other research implies that consciousness emerges when there is the right balance of connectivity between brain regions, known as the ‘information integration’ theory. It has been suggested that anesthesia can induce unconsciousness by disrupting the communication between brain regions.

Some fear that an increased understanding of consciousness will dissolve our sense of personal responsibility

Just as there are different levels of intent in law, there are different levels of awareness that can be identified in the lab. Despite being awake and functioning, one’s mind might be elsewhere, such as when a driver zones out or when a reader becomes engrossed. A series of innovative experiments have begun to systematically investigate mind-wandering. When participants zone out during a repetitive task, activity increases in the ‘default network’, a set of brain regions previously linked to a focus on internal thoughts rather than the external environment. Under the influence of alcohol, people become more likely to daydream and less likely to catch themselves doing so. These studies are beginning to catalogue the influences and mechanisms involved in zoning out from the external world. With their help we can refine the current legal taxonomy of mens rea and put legal ideas such as recklessness, negligence, knowledge and intent on a more scientific footing.

An increased scientific understanding of consciousness might one day help us to determine the level of intent behind particular crimes and to navigate the blurred boundary between conscious decisions and unconscious actions. At present, however, we face serious obstacles. Most studies in cognitive neuroscience rely on averaging together many individuals. A group of individuals allows us to understand the average, or typical, brain. But it does not follow that each individual in the group is typical. And even if this problem were to be overcome, it would not help us to adjudicate cases in which normal waking consciousness was intact, but happened to be impaired at the time of the crime.

Nonetheless, the brain mechanisms underpinning different levels of consciousness are central to a judgment of automatism. Without consciousness, we are justified in concluding that automatism is in play, not because consciousness itself is not also dependent on the brain, but because consciousness is associated with actions worth holding to a higher moral standard. This perspective helps to arrest the slide down Eagleman’s slippery slope. Instead of negating responsibility, neuroscience has the potential to place conscious awareness on an empirical footing, allowing greater certainty about whether a particular individual had the capacity for rational, conscious action at the time of the crime.

Some worry that an increased understanding of consciousness and voluntary action will dissolve our sense of personal responsibility and free will. In fact, neurological self-knowledge could have the opposite effect. Suppose we discover that the brain mechanisms underpinning consciousness are primed to malfunction at a particular time of day, say 7am. Up until this discovery, occasional slips and errors made around this time might have been put down to chance. But now, armed with our greater understanding of the fragility of consciousness, we would be able to put in place counter-measures to make major errors less likely. For Brian Thomas, a greater understanding of his sleep disorder might have allowed him to control it. He had stopped taking his anti-depressant medication when he was on holiday, because he believed it made him impotent. This might have contributed to the night terrors that caused him to strangle his wife.

Crucially, increased self-knowledge often percolates through to laws governing responsible behaviour. A diabetic who slips into a coma while driving is held responsible if the coma was the result of poor management of a known diabetic condition. Someone committing crimes while drunk is held to account, so long as they are responsible for becoming drunk in the first place. A science of consciousness illuminates the factors that lead to unconsciousness. In reconsidering the boundary between consciousness and automatism we will need to take into account the many levels of conscious and unconscious functioning of the brain.

Our legal system is built on a dualist view of the mind-body relationship that has served it well for centuries. Science has done little to disrupt that until now. But neuroscience is different. By directly addressing the mechanisms of the human mind, it has the potential to adjudicate on issues of capacity and intent. With a greater understanding of impairments to consciousness, we might be able to take greater control over our actions, bootstrapping ourselves up from the irrational, haphazard behaviour traditionally associated with automata. Far from eroding a sense of free will, neuroscience may allow us to inject more responsibility than ever before into our waking lives.

For references to the scientific research discussed in this essay, see Steve Fleming's blog The Elusive Self.

Read more essays on free will, law & justice and neuroscience


  • http://twitter.com/ravenseldon6 Naveen Ramalingam

    An Excellent Article..but my opinion is that should not the advances in neuroscience be used to rehabilitate the offenders rather than punishing them as they only get to suffer and no one gains as lives lost are lost and not gonna come back..of course it will be a great resource hog but i think thats the right way IMHO

    • Nun Yerbizness

      I agree punishment is not morally sound and will not effectively change anything but short of rehabilitation they should be confined (humanely) to protect the rest of us

      • Hominid

        Punishment for crime is a moral obligation of the state to the aggrieved.

    • Hominid

      Nope - it should be to protect others from the destructive actors.

      When a bear discovers tasty kids on the edge of town are ready prey, he is only being a bear. But, we remove him one way or another. The notion that behavior that is somehow 'hard-wired' or 'innate' should exonerate the perpetrator or that he can be "rehabbed" is humanist nonsense. The primary concern should be public safety.

  • Barry

    The article is excellent but the point of view taken by the crown is absurd and if possible, over-civilized. Thomas killed someone. As it turns out someone in the bed with him. Everything else is pseudo-sophisticated blather. He is responsible for his act and for all of our sakes should be dealt with accordingly.

    • Samuel B.

      Consideration for how consciousness might affect intent = pseudo-sophisticated blather? Really??

      • Hominid

        Yes - "really" - because neither YOU nor anyone else KNOWS how "consciousness might affect intent."

    • Oliver Milne

      So if you sleepwalked out a third-story window, fell on a passing granny and as a result she died, you'd be a murderer?

      • Hominid

        Straw man - childishly so.

        • Oliver Milne

          He killed her in his sleep! If that doesn't remove criminal responsibility, what does? If you want a better analogy, suppose he was sleepwalking and got into his car and drove, and ran over somebody because, y'know, he was asleep the whole time. Would he be criminally responsible? Hell no. It'd be a bizarre tragedy, nothing more. The same applies here.

          • Hominid

            Your perspective is wrong.

          • Oliver Milne

            Fancy explaining why?

          • Hominid

            Sure. You're looking out for the benefit of the fukk-up who hurt someone instead of the someone (& people) who are hurt by the fukk-up. You're a wacked-out Lib-thinker - totally irrational.

          • Oliver Milne

            But that's just the point, he wasn't a 'fukk-up', he was a totally normal guy who woke up and found he'd killed his wife in his sleep. You can't deter that with prison. There was no mens rea. He's as much a victim as any other husband whose wife died suddenly. Are you trolling? What's wrong with you?

          • Hominid

            Only an irrational Lib-thinker believes he knows what's in another's mind. To accept such a defense - "I was asleep and not responsible for the murder I committed" - is lunacy only a whacked-out Lib pervert would buy.

          • Oliver Milne

            If you had said from the beginning "like hell he was asleep" you could've saved us both a lot of confusion. If there was conclusive evidence he really was asleep, would you agree he shouldn't be criminally liable?

          • Hominid

            You probably don't realize how nonsensical your question is.

          • Oliver Milne

            And you're avoiding answering it, because you agree with me but don't want to admit it now you've called me a 'whacked-out Lib pervert'. For my part, you've got a good point, now you've finally spitten it out - I don't know what evidence would convince me a guy who strangled his wife was asleep at the time. Yet you seem kind of invested in belittling a stranger on the internet, at the expense of actually having an interesting discussion about this stuff. Why are you so dedicated to fighting your bizarre American culture war? What good does it do?

  • Al Rodbell

    Interesting exposition.

    With the advances of neuroscience we can identify those specific syndromes that do more than contribute, but actually cause homicidal behavior. Our discussion group has been considering this recently based on the recent interview with Lee Malvo, one of the snipers who killed a dozen people ten years ago.

    There is no medical syndrome to explain his actions, but there is a psychological one so profound that the jury gave him a lessor sentence of life rather than execution. These are the more common cases, where early experiences of being surrounded a violent childhood preordains one's criminal actions as definitively as any neuronal aberration.

    The concept of free will, and of personal guilt that flows from such decisions is a necessary fiction found in some form in all societies. It is a way that a culture imposes its rules that are inherently unfair to anyone who thinks about it all. Those who resist the absolution of guilt to those with rare neurological conditions fear a world where those with the best advocates will be able to explain away any abominable crime. This fear is not unfounded, and until we can establish a more refined framework of analysis, the insights of neuroscience on aberrant behavior must rightly be limited.

    For more see alrodbell.blogspot.com

    • Hominid

      Psychological syndromes are BS. Psychology co-opted the sexier term 'neuroscience' to acquire scientific legitimacy in the eyes of the scientifically illiterate.

  • FormFactor

    What physical processes are responsible for the writing of this piece and believing it has any correlation to reality?

  • saksin

    For once, an article that avoids every one of the embarrassing non sequiturs that abound in popular expositions of the legal and moral implications of progress in neuroscience, errors of reasoning to which even some neuroscientists have succumbed. As Steve Fleming so judiciously shows, there are no good grounds for believing that neuroscience findings will overturn the validity of traditional legal distinctions regarding culpability and legal responsibility, but may in fact refine the practice of their application. The day is not yet here, but is likely to come, when neuroscience can pinpoint the circuitry of the neural function we experience as a consciously decision-making self as well as its malfunctions. Such knowledge in no way threatens classical legal conceptions regarding moral and legal responsibility for our actions, distinctions which contrary to the comment made by "Barry" are essential ones to a legal and moral order capable of serving justice. That such distinctions can be abused to exonerate the guilty is a good reason to curtail the abuses, but not to abolish the disctinstions themselves. In Brian Thomas' case an innocent man was aquitted of a crime HE did not commit - because HE was not present at the scene of the crime at the time it took place (though his sleep-acting body was), assuming the facts are as stated in the article. I would myself have had trouble believing that such an act were possible while dreaming a very different incident, except that I myself once woke up from a dream in which I was defending myself from a sudden attack by my younger brother. I was awakened by the shocked scream of my girlfriend who had been sleeping next to me, only to be awakened by being pummeled by my hard blows!!! We had the best of relationships at the time, and no conflict which might have supplied me with unconscious motivation to punish her in this way. I was as shocked as she was, and luckily there was no reason to involve the police.

    • Al Rodbell

      Well said. Your personal experience makes what occurred to Thomas real to you, but most Jurors and citizens have not had the "benefit" of this. See my comment below for the real limits of neurological defects to criminal justice, which is the need for the fiction of "free will" even for those vastly larger numbers of victims of birth who we must condemn, morally and legally.

  • mahood

    For the classic, and incisive philosophical account of just these issues you should read P.F. Strawson's essay "Freedom and Resentment."

  • Wayne Wilner

    Recent brain research suggests that our unconscious mind is the motivating force behind all of our actions. If the legal system seeks a responsible party to punish for committed crimes, it should seize upon our unconscious. Fleming's anticipation of fine-grained distinctions of consciousness assumes that the legal system must retain its antiquated notion that our consciousness is most responsible for our actions. I expect that further research will completely undermine that assumption.

    • John Rumbold

      This idea has been in circulation since Freud's time and revived after Libet's research. You just demonstrate the need for people to understand the difference between descriptive and normative. Neuroscience will not destroy the criminal justice system.

    • Hominid

      Your unconscious is expressing your preferred preconceptions and biases and wishing they were reality for emotional reasons.

  • http://www.facebook.com/perry.greenbaum1 Perry Greenbaum

    Steve: This is an excellent and insightful article that touches upon both legal and moral issues while bringing in the evolving knowledge of neuroscience.

  • Bernecky

    The fact that Steve Fleming makes his living as a scientist doesn't excuse the lack of an hypothesis or thesis statement in this which is an exploratory exam (of the reader, presumably so that Fleming may better know himself).

    "Free will," like "instinct," is a term that the political scientist employs when he wishes to demonstrate proficiency in a field in which all others are ignorant: ignorance.
    Infants, born to be the interlocutors of scientists, possess both qualities (if, indeed, they're different qualities), but even both qualities together are useless, worthless, to the infant who is left alone in the wilderness, to survive on his own devices.

    We might subscribe to the scientist's belief that Anders Breivik is an anomaly: the person who, for all we know, survived on his own devices. The question is what we're going to do with the person whose aim is (not to commit murder, but) to commit murder as a means of causing the rest of us to build, and to maintain, a facility for him alone, within a prison.

    A caveman, a cavewoman, a cavechild and, now, a second caveman (who attempted to take the woman). Can the family survive if, while the husband is hunting, the wife sets aside her foraging and mothering, and tends to the second man? There's not reason, in our scenario, for the first man to build a cage to house the second man, or even for him to fashion a rope with which to tether him to a tree. There is infinite reason for him to do anything but.

    The idea is to supplant ignorance.

    The death penalty, far from being the last resort, is the first. The promise of the death penalty is (not death, but) to keep the criminal from breeding.

  • Doug Hensley

    The standard for guilt needs to take into account issues of deterrence and public self defense. If the offender did it, and is likely to do something similar again, then he has to be locked up so he doesn't do it again. If the offender did it, and would have profited from doing it but for getting caught, then he has to be punished enough that others, watching how it played out, will conclude that doing likewise won't pay.

    In this way of looking at things, people who zone out while driving have no defense. Zoning out is a habit as much as anything. Drivers make conscious decisions about how long to drive. If economics impels them to dangerous decisions, then the prospect of punishment can be put in the balance to tilt their thinking in a direction consistent with public safety. (Or, in the future, machinery that monitors driving performance and intervenes if needed can be attached to the vehicle.) If the mind is a machine, after all, then the prospect of punishment will inevitably be factored in to the decisions the "mind" makes. For this much is beyond doubt: humans take in information, process it, and their actions are then (mostly, and with allowances for difficulty processing complexities*) rationally related to that information.

    *So laws should always be written with a determination to keep them simple.

  • http://twitter.com/alokpi Alok Prasanna Kumar

    There seems to be a slight conceptual confusion here.

    While automatism is sufficient to show that there was no mens rea (as say in a situation where the crime committed is purely accidental), medical conditions such as schizophrenia, autism (in extreme cases) are cases where the "crime" has been committed with both mens rea and actus reus (i.e., the act) but the person committing the crime is not responsible.

    Whereas in the first case no crime has been committed at all, the latter forms a justification for crime, even if it has been established. The plea of "self-defence" is also alike. The law chooses not to punish people in such situations even if they've committed a crime.

    At the same time, it is still open to the law to prescribe that persons who cannot be held "responsible" for crimes committed may still be committed to a juvenile centre or mental health institution.

    • John Rumbold

      You have this slightly wrong. Someone who is insane cannot commit a crime, and insanity is more like an excuse than a justification. Self-defence is certainly a justification, although some would call it a 'denial-of-proof'.

  • Nun Yerbizness

    "It is no longer meaningful to lock people up on the basis of their
    actions, because their actions can always be tied to brain function."

    if locking people up is an act of punishment I would agree but if locking people up is an act of protection from people (a threat to themselves or others) with diseased brain function it is very meaningful.

  • Arash Farzaneh

    First off, excellent article, Steve! This is an issue I have been struggling with and continue to do so. However, to get to the root of the problem, we need to look at the
    notion of free will here. And this is exactly what's troubling me.

    If we don't have free will (and I believe we generally don't) and we do not freely choose our actions then the question of morality and moral responsibility becomes mere tautology and even redundant.

    What will the neuroscientist say? We have consciousness, but if my (admittedly very limited and rudimentary) understanding of neuroscience is correct, then the
    neurons fire first, before we even become aware of them. Put differently, my brain has already decided that I will squash the fly on my table before I actually do so. I am not the one running the show; consciousness may become reduced to sheer awareness ... I am aware that something is being done and we become mere spectators
    of the unfolding actions of our lives.

    It sounds scary, as if we were all Frankensteins after all, which is why I am so troubled by it. It is the biblical pronouncement that “they do not know what they are
    doing” because they are victims of their brains and bodies.

    Now when it comes to the law, I disagree with the idea of morality. Laws should not exist for the purposes of passing judgement or issuing punishment. When people are
    declared guilty and locked away, it is, in my view, not a question of morals, but rather a social necessity.

    People are then imprisoned not because they know (or ever did for that matter) what they are doing, but simply so that they do not continue to pose a threat to our society.
    Laws should be practical in the sense of avoiding (and even eliminating) harm to society.

    To me, it is nonsensical to release sex offenders, particularly pedophiles into society when the authorities know and explicitly state that these people are prone and at risk to re-offend. It is akin to releasing a lion into our society while crossing our fingers that it does not attack or devour anybody.

    Again, I am not passing judgement here. Certain people are sick. They pose a threat. The man with the sleeping condition does too; if he does not take his medication, he
    too can re-offend. I do not blame him for his action the same way I do not blame an alcoholic or drug addict for their condition. They all need help.

    As for the rest of us, as long as we keep to ourselves, do not cause harm to others and do not break the law, we are free to exercise our lack of free will, not knowing what it is we are doing but doing it nonetheless.

  • Raleigh

    Is it so bad that our society could detect these small little ticks that change our behavior? Maybe its what we need, to be more understanding of those who have slight differences. In many liberal countries drug addicts, thieves, and even murderers aren't sent to a prison where they can be boxed up and put our of sight and mind. They are given therapy, and live normal lives, and are treated and reintroduced into society as functional members.

    • Hominid