This kind of individual madness was rare before 1960. Why is there now so much of it? Well, we know that Loughner, described by a high-school classmate as a pothead, had been rejected by the US Army in 2008 because of his admitted marijuana use. It was not minor. Pima police picked him up in 2007 because his car stank of marijuana. The offense was recorded, but, as so often in the alleged War on Drugs, nothing happened to him. Much of this information emerged some time after the shootings, when interest had cooled.

Even though it meshes with many similar stories from Canada, Japan and Europe, in which the perpetrators of crazy mass killings turn out to be long-term marijuana users, and with the arguments in Alex Berenson’s recent book Tell Your Children, in which a strong direct link is made between marijuana and many violent crimes, this issue never gets any further. The argument, by the way, isn’t that the killers were high at the time. Marijuana’s effects on the minds of many of its users are permanent. They may give it up entirely, but the damage has been done. Yet if nobody asks and nobody cares, nobody ever puts two and two together.

There’s a simple reason for this. Marijuana has been the beneficiary of one of the slickest, most sustained advertising campaigns in human history. Not only do millions believe it is some sort of medicine. Most people, even law enforcers, describe it as a ‘soft’ drug. This is an absurdity. Lifelong mental illness is not a ‘soft’ outcome. So, even when its use is clearly linked with mental illness and terrible crime, nobody even asks if it might be to blame. Its defenders chant ‘correlation is not causation’. Of course this is true. But correlation is also not necessarily not causation. Correlation is the foundation of epidemiology.