In many ways, violence is like cholera, passing from person to person and treating it in similar ways is working to reduce it.
Usually, facial trauma doesn't kill you, but it can cause significant disfigurement. Working as a surgeon in Glasgow in the early 2000s, Christine Goodall treated hundreds, if not thousands, of patients with injuries to the neck, face, head and jaw.
One young man came into the hospital in the night with a knife wound across his face. Goodall dreaded the next day when she would have to tell him that it would be impossible to reduce the scar's appearance.
But his reaction surprised her.
"He was very offhand about it," she says. "His friends came to see him later that afternoon, and I realised why it wasn't going to be a problem for him - because they all had one. He'd just joined the club."
In 2005, the United Nations declared Scotland the most violent country in the developed world. The same year, a World Health Organisation (WHO) study of crime figures in 21 European countries showed that Glasgow was the "murder capital" of Europe .
More than 1 000 people a year required treatment for facial trauma alone, many of them as the result of violence.
Goodall, who has spent most of her life in Glasgow, would stitch up the wounds and work to repair the damaged tissue. But for most patients, the problems continued after they were discharged. Chronic pain, posttraumatic stress disorder, self-medication with alcohol and drugs.
Often, the same people would come back through the casualty department, repeated victims and perpetrators of violent attacks. "We were really good at patching injuries up," says Goodall.
"But I started to think: What can we do to prevent them coming here in the first place?"
Humans engage in a wide array of risky behaviours that can lead to serious health problems such as overeating or unsafe sex. It's the accepted wisdom that doctors should encourage patients to change their behaviour - let's say, go on a diet or use a condom - rather than wait to treat the obesity-related heart attacks or HIV infections that could be the result
Yet, when it comes to violence, there's an assumption that it is innate in the people who perpetrate it and that they are beyond redemption.
More often than not, solutions have been sought in the criminal justice system, with stricter sentencing or increasing stop-and-search tactics.
In 2005, Karyn McCluskey, principal analyst for the Strathclyde police department in western Scotland, wrote a report pointing out that traditional policing was not reducing violence.
It included a list of recommendations. "One was tongue-in-cheek," recalls Will Linden, who worked for McCluskey as an analyst.
"'Do something different.' I don't think it was meant to stay in there. But the chief constable said, 'Okay, go do something different.'"
McCluskey's team started pulling together evidence on the drivers of violence. "In Scotland, it was poverty, inequality, things like toxic masculinity, alcohol use, all these factors - most of which were outside the bounds of policing," Linden explains.
One of the primary indicators that someone will carry out an act of violence is being the victim of one, the American Psychological Association says on its website.
Next, they looked around the world to find pioneering programmes to prevent violence. This was the foundation of the department's Violence Reduction Unit (VRU) that eventually became a national centre of excellence with Linden as its acting director.
The VRU takes a "public health approach" to violence: a school of thought that suggests that violent behaviour itself is an epidemic that spreads from person to person.
The idea that violence spreads between us, reproducing itself and shifting group norms, explains why one locality might see more stabbings or shootings than another area with many of the same social problems.
It's also become popular with the WHO, as the body explains in its guidance on violence prevention: "Violence can be prevented and its impact reduced, in the same way that public health efforts have prevented and reduced pregnancyrelated complications ... " experts write.
"Despite the fact that violence has always been present, the world does not have to accept it as an inevitable part of the human condition."
Watch: 'I became a bully because my dad abused my mom'
Since the VRU was launched in 2005, the murder rate in Glasgow has dropped by 60%, according to 2017 reports by Scotland's The Herald news service.
How did Glasgow do it? The VRU looked to Chicago.
In the 1980s and early 1990s, epidemiologist Gary Slutkin was in Somalia as one of six doctors working across 40 refugee camps housing a million people. His focus was containing the spread of TB and the waterborne illness cholera, which causes diarrhoea and dehydration.
Stopping infectious diseases relies heavily on data. First, public health officials map out where the most transmissions of the disease are occurring. Then they focus on containing it by getting people in these areas to change their behaviour even when larger structural factors can't be tackled.
For instance, diarrhoeal disease is often in large part caused by poor sanitation. It takes a long time to improve plumbing systems - but in the meantime, thousands of lives can be saved by giving people oral rehydration solutions to prevent deadly dehydration.
But changing behavioural norms is about far more than merely giving people information. To change behaviour - whether it's by using rehydration solutions or avoiding dirty water - relatable messengers are essential.
"In all of these outbreaks we used outreach workers from the same group [as the target population]," Slutkin says. "Refugees in Somalia to reach refugees about TB or cholera, sex workers to reach other sex workers about HIV ... "
Slutkin returned to his native Chicago in the late 1990s to a different kind of problem: a skyrocketing homicide rate.
His ideas about tackling this problem began as a nerdy project: he gathered maps and data on gun violence in Chicago.
As he did, the parallels with the disease outbreaks he was accustomed to were unavoidable. "The epidemic curves are the same, the clustering. One event leads to another, which is diagnostic of a contagious process. Flu causes more flu, colds cause more colds and violence causes more violence."
This was a radical departure from mainstream thinking about violence at the time, which primarily focused on enforcement.
"The idea that's wrong is that these people are 'bad' and we know what to do with them, which is punish them," he says.
"Behaviour is formed by modelling and copying."
He spent years gathering funding for a pilot project that would use the same steps against violence as the WHO takes to control cholera or TB: interrupt transmission, prevent future spread and change group norms.
In 2000, Slutkin launched his first project in Chicago's West Garfield Park neighbourhood. Within the first year, there was a 67% drop in shootings, statistics on the project's website argue.
More funding came, more neighbourhoods were piloted. Everywhere Slutkin launched his Cure Violence programme, violence dropped by at least 40%.
Although many deeper structural factors were contributing to Chicago's violence - poverty, lack of jobs, exclusion, racism and segregation - Slutkin argued that lives could be saved by changing the behaviour of individuals and slowly shifting how communities think about and frame violence in their daily lives.
As in many places, the discussion of violence in Chicago often takes on a highly racialised tone. The city is deeply racially segregated. Many South Side neighbourhoods are over 95% African American; others are more than 95% Mexican American. Most of these areas are severely socioeconomically deprived and have suffered years of state neglect. Homicide rates can be up to 10 times higher than in more affluent, predominantly white areas of Chicago.
But Slutkin emphasises that this clustering is less to do with race and more to do with patterns of behaviour - usually among a small section of the population, often young and male - that are transmitted between people.
Today, Slutkin's organisation Cure Violence works in 13 Chicago neighbourhoods, and versions of it run in New York, Baltimore and Los Angeles, as well as in other countries. In each place, it trains local organisations that then find community members in the area to do the work.
Although there is a level of debate about Cure Violence's use of statistics, the method's overall effectiveness has been shown by numerous academic studies. A 2009 study by Northwestern University found that crime went down in all neighbourhoods examined where the programme was active.
In San Pedro Sula, Honduras, the first five zones in which Cure Violence worked saw a drop from 98 shootings during January to May 2014 to just 12 in the same period in 2015, according to a Cure Violence evaluation.
How do we stop violence? It starts with our thinking.
Demetrius Cole is 43, a gentle, soft-spoken man. He grew up in an area of Chicago afflicted by violence and, at the age of 15, saw his best friend die in a shooting. Nonetheless, he'd had a stable home life and stayed out of gangs. He planned to join the Marines.
But when he was 19, a friend bought a new car. Boys from the neighbourhood tried to steal it and shot Cole's friend.
Cole didn't stop to think.
He retaliated. In those few minutes, his life changed entirely.
"I reacted off emotion," he says.
His friend was left paralysed and Cole was sent to prison for 12 years for his response.
But since October 2017, Cole has been working for Cure Violence in Chicago's West Englewood area.
He finds people in the same situation he was in and tries to persuade them to pause. That can mean calming people down before disputes turn violent or preventing retaliations after shootings.
Cole is a "violence interrupter". Violence interrupters use numerous techniques, some borrowed from cognitive behavioural therapy, an approach to psychology that focuses on managing problems by changing the way you think and act, the United Kingdom National Health Service explains.
Cole reels them off.
"Constructive shadowing", which means echoing people's words back to them; "babysitting", or just staying with someone until they have cooled down; and emphasising consequences.
Interrupters' effectiveness depends on their credibility. Many, such as Cole, have served long prison sentences and can speak from experience. Most also have close relationships with the community. They can respond when violence happens but also know when it might be brewing and defuse tensions.
Cole explains: "We may not be able to reach everybody, but for the few people we do reach, it's a beautiful thing."
The surprising new science emerging around violence - and the solution to helping stop it.
In the nearby neighbourhood of Grand Crossing, one of Cure Violence's newest centres opened in December 2017.
When I visit one spring afternoon, a boy in his teens with a visible facial scar sits in the reception. Young men from the area pass in and out to speak to their outreach workers and use the computers.
Supervisor Demeatreas Whatley is catching up on paperwork. Whatley has been working for Cure Violence on and off since 2008 when he finished a 17-year prison sentence.
His first assignment was in his old neighbourhood, Woodlawn.
Ignoring his family's protests, he moved on to the block worst affected by gang tensions and worked to build "peace treaties". Initially, it was about convincing the two groups to stay out of each other's way. Gradually, he then worked on getting individuals to shift their views about violence as well as find work or go back to school.
"I knew I was making a difference when I saw the old folks back out on the porch again drinking their coffees," he says.
Although it's adapted for each new location, Cure Violence follows roughly the same steps everywhere. First, map the violence to see where it happens. The streets and blocks where incidents cluster are the focus. Next, hire would-be interrupters with a local connection.
These interrupters patrol the streets, get to know the neighbours and build links with young men and women deemed to be at high risk. Grand Crossing's 11 interrupters devote at least six hours of their eight-hour shifts to being out in the neighbourhood.
"You have to be there," Whatley says.
Outreach workers also spend as many as two years trying to change people's attitudes about violence and connect them with job opportunities, counselling or education.
"You have to have a few tricks up your sleeve," says Jermaine Peace, an outreach worker at Grand Crossing.
Sometimes he might hook someone's interest by telling them he can help them get photo identification, such as a driving licence.
"Once you start showing you care and you call them, they may call one day and say, 'Man, I ain't ate in two days.' You go over there and buy them something to eat and you get more chances to talk to them."
Slutkin emphasises how quickly this model can reduce homicides, and how it costs less than mass incarceration, but it takes a lot of workers to get results. Some of Chicago's gang territories are small, just a few blocks. A violence interrupter respected in one area may be unknown, or even mistrusted, in another. To work, there must be at least one interrupter with strong connections in each district, so that if a conflict erupts, someone with the trust of the group can mediate.
Epidemiologist Gary Slutkin says treating violence like a disease can slash homicide rates and is cheaper than funding mass incarceration.
In Glasgow, Christine Goodall didn't only see patients immediately after they'd been attacked. Sometimes people would come to her months, even years after their injury, desperately seeking a solution. Goodall had to do something, and in 2008, with two other surgeons, she founded a charity called Medics Against Violence , which partners with the Violence Reduction Unit (VRU).
When the VRU was established in 2005, Karyn McCluskey and John Carnochan of the Strathclyde police department had searched the world for solutions to Glasgow's violence problem. Eventually, they came up with an approach that blended Gary Slutkin's thinking with that of David Kennedy, a Boston-based criminologist, who gave gang members one choice: Get an education or a job - or face harsh penalties.
In Glasgow, this meant ramping up traditional penal measures - increased stop-and-search and stricter sentencing for knife possession - alongside preventive measures in line with the public health approach.
Acting VRU director Will Linden says they had no choice but to get tough.
"If we hadn't had that evolution, we'd spend most of our time firefighting the press about [allegations of] being soft on crime."
Today Linden estimates that around 90% of the emphasis - and the funding - is on prevention.
The VRU is run by the police force, with support from the Scottish government. Scotland has the world's only police force to have formally adopted a public health model.
But alongside police, a range of public officials - from doctors to social workers - are involved in VRU's work.
The WHO breaks the public health approach to violence down into four steps. The first is uncovering as much knowledge as possible about all aspects of violence. The second is investigating why violence occurs - looking at causes, correlations and risk factors. The third is exploring ways to prevent violence using this information. The fourth is implementing these strategies.
To see a serious effect, the body cautions, work needs massive levels of collaboration - and longer than a four- or five-year election cycle. Linden notes that Scotland has an unusual degree of political consensus, with successive governments funding this work.
On a sunny evening in downtown Chicago, I watch Slutkin give a talk to an audience of young professionals. Homicides reached a 20-year high in 2016 in the city.
Slutkin presents graphs showing that every time Cure Violence's funding is cut in a particular area, shootings spike, and when it returns, they drop (critics argue that it is impossible to conclude causality because of other factors at play).
Whatley, the site supervisor at the Grand Crossing branch, has experienced these funding cuts first-hand, repeatedly losing his job and being rehired in the 10 years since he started working for Cure Violence. The project he initially worked on, in his home area of Woodlawn, was cancelled.
The old folks no longer drink their coffees out on the porch.
"Despite massive amounts of data, it's hard to get funding for this," Slutkin tells the audience. "Mass imprisonment has no good data - but it's funded. This is the only epidemic health problem not being tackled by the health department."
This is an edited version of an article which first appeared by Wellcome on Mosaic and is republished here under a Creative Commons licence.