A man walks into a pub. "Hey, I've just not been mugged," he says.
"That's amazing, " a bloke at the bar says as he puts his pint down. "I didn't strangle my wife today."
"You two are weird," said a third bloke. He was a journalist.
Reading the papers this week, you'd be forgiven for thinking there is carnage on our streets - partly because one newspaper said exactly that.
The truth is, there isn't carnage on our streets and very few of us are victims of or witnesses to crime more than once or twice in our lives - even fewer are victims of serious crime.
In any rational description of the world, our risk of dying in, say, a knife attack or a serious assault ranks way behind our risk of dying in a road accident or from the effects of cigarettes or alcohol. And if you happen not to be a city dweller and are over 25, your chances of being shot or stabbed are vanishingly small: your chances of being attacked or killed by a stranger, approaching nil.
Except, we don't learn about the world from a rational description. We learn about it, for the most part, from "news" - and crime is news precisely because it is both shocking and uncommon. Except, of course, when it seems to confirm our society is sick and broken. Then, the more common and apparently true-to-type the gruesome violence can be made to seem, the better.
One of journalism's great father figures, American commentator, media critic and diplomatist Walter Lippmann, struggled nearly 90 years ago with this paradox. On the one hand, news is the way we learn about the world; on the other hand, you would be mad to rely on it to learn about the world.
"All the reporters in the world working all the hours of the day could not witness all the happenings in the world," he wrote. So, a thing becomes becomes "news" only when it is a "manifestation" at one of the places journalism has "watchers stationed" - the police station, the courts, the crime scene.
So how to put the deaths of Amar Aslam or of Jimmy Mizen or of any of the other 20 teenage victims of violence so far this year into context? They are terrible, sad events and we all have great sympathy for the boys' families. But beyond the personal tragedies that they represent, they tell us nothing about teenagers, gangs, knives or crime. Most of all, they tell us nothing about how concerned or fearful we should be for ourselves and our own families.
Three years ago, a Home Office survey found that 4% of 10-to-17-year-olds had carried a knife at some time in the previous 12 months: or to put it another way, 96% had not. This month, in Operation Blunt2, the Metropolitan Police seized 193 weapons in more than 4,000 searches: or to put it another way, 95% of those stopped were not carrying knives.
Or try this question. In England and Wales, were you more or less likely to be murdered last year than five years ago? Were you more or less likely to be stabbed last year than five years ago? Bludgeoned to death? Strangled? You'll have guessed the answer to each - according to official statistics (pdf file) - is less likely. There were 734 murder victims in England and Wales last year - almost 10% down on the 805 murdered in 2001-2002 and about 5% below the average for 2001-2006.
Is there another, better way of reporting crime that doesn't risk distorting what we think we know about our world? Take the trial at the Old Bailey of the two young men and two youths charged with the murder of 14-year-old Martin Dinnegan.
What are the alternatives to covering the trial as the news story that it is? Not covering it at all? Holding back details of the evidence? Pointing out repeatedly that most 14-year-old boys don't get stabbed? Using a chart within the story to show how deaths by stabbing and beating are falling not rising?
Maybe part of the answer is for us all - journalists and audiences - to understand that "news" is what it is: a semi-ritualised set of snapshots of a small sector of our common lives. No more, no less.
Maybe journalists should resist the temptation to make links where none exist: are we really in a "battle to fix broken Britain" as the Sun's banner for each report of teen violence claims? And maybe audiences have a job to do, too: to understand the limitations of "news" that Walter Lippmann wrote about all those years ago. And to realise it's the unusual that's weird, not the everyday.