Adam Lanza was a loner: highly intelligent with a ghostly pallor; awkward but pleasant seeming; described by his own brother as a "nerd".
On December 14 last year, Lanza, 20, walked into Sandy Hook Elementary School in the affluent town of Newtown, Connecticut, and over the next two hours shot 20 children and six adults dead, before turning the gun on himself. Already that day, he had killed his mother with her shotgun.
Why had this young man, as opposed to millions of other "geeky" outsiders, murdered 27 innocents? The media talked about Lanza's mother's gun collection; the fact he had no Facebook page and no photo of him appeared in his high-school yearbook, only the words "camera-shy". But at the same time, scientists at the University of Connecticut were embarking on a different line of inquiry. The genetics department was analysing Lanza's DNA.
The university refused to give any details about these investigations - possibly of cells from Lanza's brain, but equally likely from cells taken from his hair or the gun he used - or what they hoped they could reveal from the analysis. But the news shone a light on an area of behavioural genetics that provokes deeply divergent opinions both within the scientific community and in wider society. Is it possible that there is a gene that makes some people "evil"? Could future murderers be spotted before they have committed a crime? And should they be punished if they are simply prisoners of their own biology?
Some scientists rejected the announcement, saying it was "almost inconceivable" there was a common genetic factor among mass murderers. But others applaud the initiative.
"Only by studying individuals [like Lanza] as thoroughly as possible will we some day be able to reduce the frequency of these sad episodes," says Dr Art Beaudet, chairman of the Department of Molecular and Human Genetics at Baylor College of Medicine in Houston, Texas. "The genetic research should include genomic copy number analysis, whole genome sequencing and epigenetic analysis of post-mortem brain tissue."
Kathleen Taylor, a neuroscientist at Oxford University, agrees that no line of investigation into criminal behaviour should be dismissed out of hand. "More knowledge is always useful," she says. "We don't know nearly enough about extreme violence and its effects on the brain. There's only one person involved in this analysis, so we can't really draw any conclusion, but if they find Lanza has a very unusual type of gene or a particular type of neurotransmitter, they could at least be making a start."
This is thought to be the first time that scientists have analysed the genetic blueprint of a "spree killer", but it's far from the first attempt to examine a murderer's biology. In 1931, the brain of the "Vampire of Dusseldorf", Peter Kurten, a serial killer, was removed from his corpse after his execution for examination, although no useful conclusions were published. Today, it is displayed in Ripley's Believe It Or Not! Museum in Wisconsin.
Over the past decade, Dr Kent Kiehl, a neuroscientist at the University of New Mexico, has visited eight high-security prisons in two US states with a mobile MRI unit, scanning the brains of criminals to see if those defined as psychopaths have different brain structures from "someone who commits a robbery out of poverty", as Kiehl puts it.
Dr Kiehl's and others' research has found that psychopaths' brains tend to have very low levels of density in the paralimbic system, the area of the brain associated with the processing of emotion, something that may be genetically determined. The result is that psychopaths tend to have impulsive personalities and show little evidence of feeling guilt, remorse or empathy.
In contrast, "spree killers" tend to be extremely depressed, to the point of suffering from a delusional psychosis accompanied by voices or hallucinations, or - as in Lanza's case - to be young people with physiologically immature brains, who in their state of ultra-sensitivity decide to exact "revenge" on the world for perceived injustices.
Recent years have seen huge advancements in DNA research, with researchers now able to identify specific genes that are linked to anti-social or aggressive behaviour, in particular the MAO-A gene (nicknamed "the warrior gene"), which appears to be hereditary.
A study of Danish twins concluded that a Danish man who has an identical twin with a criminal record is about 50 per cent more likely to have been in prison himself than the average Danish male. Non-identical twins are between 15 and 30 per cent more likely to both have criminal records. Similarly, adoption studies around the world have shown that a child of criminal parents is more likely to become a criminal, even if the adoptive parents are law-abiding.
Irving Gottesman, a psychologist at the University of Virginia who has worked on the Danish twin study, believes the results show that "criminals are not born, but the odds at the moment of birth of becoming one are not even".
But so controversial are the links between biology and violence that only the bravest scientists have dared tackle it. "There are many political objections and that means there's not been enough research into the area," says Kiehl.
In the United States, projects set up to investigate the issue have been shelved after public uproar. Scientists are haunted by the ghosts of the pseudoscientific eugenics movement, which held that mankind could be improved by breeding out the bad. Peaking in the early 20th century, its influence led to the sterilisation of mental patients and prisoners and, by extension, to Hitler's "final solution" of eliminating the "Untermenschen".
Today, opponents of such research worry that because minority groups tend to commit most crimes, we will blame their colour, which is genetically determined, rather than their low socio-economic status. Others fear that if criminal urges are shown to have biological or genetic origins, then crime will no longer be associated with parents, society, education - or anything else that has the potential to be improved.
"Looking at a criminal's environment is hard," says Taylor. "If you say Lanza was a loner, he watched too much television and Hollywood culture is far too violent, then it spreads the responsibility, it makes us all feel uneasy. It's far easier to say it's all about the genes." Lanza spent hours on the internet, a fact likely to make any parent of a teenager uncomfortable. His parents had also divorced bitterly - tough news for similar couples to swallow.
The truth is, history is full of examples of "normal" people committing terrible atrocities. Thousands of "ordinary" Germans, Rwandans and Cambodians committed unspeakable atrocities in genocides, of whom only a tiny percentage could have been mentally unstable or psychopathic.
In the infamous Stanford Prison experiment of 1971, where student recruits were instructed to act as prisoners and guards, the latter group behaved with such astonishing cruelty that the experiment was aborted. The results, said the experiment's instigator, Prof Philip Zimbardo, showed that "the majority of us can be seduced into behaving in ways totally atypical of what we believe we are".
Yet, as Taylor points out in her book Cruelty: Human Evil and the Human Brain, even in the worst atrocities brain chemistry is at work. Perpetrators' neural pathways harden so that they end up believing, in the unwavering way that we believe the sun will rise the following morning, that certain members of society are contaminants that must be harmed or exterminated. But if all criminal behaviour can be attributed to neurological tics, then what happens to the age-old question of free will? If a war criminal was brainwashed, or a murderer can be proved to be insane, will they no longer be responsible for their actions? Will the concept of evil become extinct, replaced by a black-or-white notion of a functioning or malfunctioning brain?
By the same token, would we continue to reward virtue? Would we continue to praise heroes, or would we simply acknowledge that they had a well-developed amygdala?
Such questions have already made the new fields of neurolaw and neuroethics hot topics. Universities, law schools and, increasingly, the judicial system are all reviewing cases where it seems "the brain" rather than "the person" might be culpable.
In the US, several killers have been sentenced for manslaughter rather than murder after DNA evidence was produced to show the perpetrator had unusually high levels of MAO-A. In Italy, in 2009, a judge reduced the sentence of an Algerian called Abdelmalek Bayout for the same reason. Bayout had stabbed a man to death in the northern city of Udine after a comment about his appearance.
Naturally, such cases are greeted with horror by victims' families. The flip side to such developments is an Orwellian world where any potential criminal is locked up, and the key thrown away. Many fear that carriers of "criminal" genes could be identified even in the womb, raising a question of what to do with the embryos.
Essi Vidling, professor of developmental psychopathology at University College London, who has carried out extensive research into psychopathy in children, describes the research into Lanza's genes as "a complete bloody waste of time". She says: "Colleagues and I are perplexed as to what would be the point. The authorities want to reassure people, 'We are doing our best to explain why this happened,' but the aim of the exercise is not scientifically informative because it only involves one person. It's a desire not to leave any stone unturned."
Yet she agrees juries love nothing more than brain scans. "They are seriously fascinated by these images and studies have shown they are much more likely to believe an argument when it's backed up by pictures of the brain. But everyone I know in the field who does responsible research, doesn't think these data are ready to go into a courtroom. In any case, even if you have a vulnerability, you still have responsibility and, since you do, society has no right to punish you pre-emptively."
Even scientists who have proved the unusual brain structures of psychopaths believe it is still impossible to dismiss environmental factors. Nearly all psychopaths suffered physical and emotional abuse as children. "Genes like MAO-A are not crime-promoting genes in themselves but they can create a vulnerability in someone who has already had a deleterious childhood," says Dr Nigel Blackwood of the Institute of Psychiatry at King's College London, who has done research in the field. "Even then, you can't say this person will definitely commit a crime. Some children are very sensitive to maltreatment, others are not."
Scientists also point out that brain structures can be altered by environmental factors, meaning no one is destined for a life of crime, just as someone with large hips is not doomed to heart disease if they eat healthily and exercise. "Twin studies show DNA can be influenced by body chemistry, what you eat, what you hear, what you feel, by the presence of various hormones," says Taylor.
With so many other factors to consider, what is the point in investigating criminals' brain chemistry? According to Dr Blackwood, it's to tailor appropriate therapy for those who commit crimes in "hot blood", as opposed to the differently wired "cold-blooded" criminals. "For most conduct-disorder children parent-training programmes work well, but they're less successful with children with callous and unemotional traits," he says. "We need to adapt treatments for this group." What treatments would be best is still far from clear, although researchers largely agree that with this group, rewarding good behaviour works far better than punishments, to which they are usually indifferent.
Could things have been different for the Lanza family if their youngest son's brain had been scanned at an early age? It's all conjecture. As Dr Kent Kiehl says: "The only thing we can be sure of is that if Lanza's mother had locked away her guns, this tragedy might have been avoided."