Every morning when I drive my son to school, we pass a small, nondescript building that houses a daycare center. It’s been a daycare center ever since we moved to this neighborhood, 17 years ago, before the kids were born.

The name has changed though. When we moved here, it was called The Kid’s Playhouse, and—because I didn’t have any kids at the time—that was all I knew about it. Until I was coming home one day—September 22, 1997, to be exact—and noticed yellow police tape wrapped around the building. One of the center’s big picture windows had been shattered, and wicked-looking shards of glass still hung in the frame. Oh my God, I thought; a car in the parking lot must have backed into the window with its bumper. I hope all the kids were OK.

In fact, as I discovered when I got home and saw the local news, the window had been broken by bullets. Three bullets, all fired by police officers; one of the bullets also struck the man who had walked into the daycare center that morning, grabbed an eighteen-month-old boy, and put a knife to his throat.

There were many shocking elements to this crime, not least of which was this: Both the child and his attacker survived. The child required 20 stitches to close the knife wound that came within a millimeter of his carotid artery and jugular vein. There is less information about the recovery of the man who attacked the boy—the bullet hit him in the eye, and he was conscious and talking when he was taken to the hospital. Beyond that, all I know is that he was sentenced to 55 years in prison.

There is information publicly available about the aftermath of this crime, about the police response, about the child’s recovery, about the trial and sentencing of the attacker. But there is very little information now—and as I recall it, there was just as little at the time—about why it happened.

Not that there’s ever a compelling explanation for stabbing a child. But when someone, especially a child, is the victim of such nightmarish, random violence, we want to know why. We need the event to make some kind of sense, so we can figure out what we ought to do to prevent it from happening again.

In this case, most people decided that what happened in the daycare center that day made sense because the attacker had, for several weeks beforehand, been living in a halfway house for the mentally ill.

“He was crazy,” was the unspoken judgment of the police and the media, and everyone seemed content to leave it at that. The news reports said nothing about the man’s past, if he’d been violent before, whether he’d been on or off medication. The whole incident was explained by virtue of the fact that he was mentally ill. And that seemed to be enough for people.

In fact, it turns out that the man had three prior felony convictions (I learned this from the court records when he subsequently—and unsuccessfully—appealed his sentence). He also had a history of alcohol abuse. In fact, he had a six-pack of beer in the bag he carried into the daycare center, the same bag from which he pulled an eight-inch knife.

Now if you look at research into violence and mental illness, you’ll find some surprising things. Surprising because they contradict our instincts, and also because they give the lie to the dominant social representations of mental illness. People with mental illnesses—even quite serious ones—are only slightly more likely to commit violence than members of the general population. Of course, the general population are no angels; my point is that the mentally ill are, by and large, no worse. With this exception: When there is a prior history of violence, or alcohol or other substance abuse, then, yes, the mentally ill are more likely to be violent. Just as, again, everyone else is.

This data aligns with some of my personal experience with violence. Violence is innate and not especially complicated, but, I have discovered, it requires considerable effort. Committing violence actually takes planning, logic, and concentration—even spur of the moment crimes like abuse and assault are really quite carefully calculated, if you look at the perpetrators’ patterns of behavior. So it seems to me that someone who’s mentally ill would have a lot more hurdles to overcome than your average would-be criminal.

At any rate, the man who stabbed a child at The Kid’s Playhouse had two very definite indicators for violence—alcohol and a violent history—and we might ask why, in that case, the social workers who had presumably been in contact with him hadn’t realized he posed a threat.

We might ask that, but I don’t know that anyone ever did. The staff at the halfway house the attacker had reportedly been living at were never, to my knowledge, asked for comment. If it came up during the trial, none of the papers reporting the proceedings thought it was worth mentioning. One story suggested that the assailant had in fact used up his allotted time at the halfway house, and been turned out into the street shortly before the crime occurred. Given that Texas ranks 49th in mental health funding, this shouldn’t surprise anyone.

So for all we know, the staff at the halfway house may have felt the man was indeed a threat, but they had treated him for as long as their employers—the state of Texas and its citizenry—had authorized. And then, somehow, he wasn’t our problem any more. We don’t want to spend our money on the mentally ill.

We do, though. We spend a fortune on the mentally ill, and we do it by putting them in prison. As a whole, our country has over three times more mentally ill people in prisons right now than we have in mental institutions. In Texas, 16% of all prison inmates are seriously mentally ill. And consider, for a moment, whether this approach makes us safer in the long run. If someone—mentally ill or not—doesn’t have a history of violence when they go into prison, the odds are very good that they will when they come out. They’re also more likely to abuse drugs when they get out of prison than they were when they went in.

Mental illness isn’t in itself dangerous, but it becomes dangerous in a society like ours where good medical care is unaffordable, alcohol is cheap, and weapons are abundant.

Compassion is pretty thin on the ground around here as well. Most of us simply don’t know how to deal with people who are mentally ill. They may seem different, and that scares us. And if we know someone is mentally ill and they don’t seem too different from us, that really scares us.

I happened to sit next to a woman on the bus once who was pretty clearly delusional. I figured this out as soon as I sat down next to her and she started talking, and I realized why, on a crowded bus, that seat happened to be empty. But I stayed where I was, and had a long, strange conversation with her, in which she told me about the “facility” she was currently living in, and how she had moved to Texas from Colorado to escape certain unnamed persons there who were persecuting her in some vague way. And I really didn’t know how to respond to her, because my self-defense training calls for me to validate people’s instincts and not to dismiss their fears. Yet it seemed unlikely that she really was being stalked by Coloradoans, and it also seemed deeply unkind to agree with her that, Yes, her life was undoubtedly in danger.

So at one point, when she had told me for the second or third time how she had moved to Austin from Colorado, just up and moved because it wasn’t safe for her there anymore, she knew it wasn’t safe, even though her friends told her she was imagining things—at some point when she paused for breath, I told her something we say all the time in self-defense workshops. “Well, if you really felt unsafe,” I said, “I’m glad you did what you felt you needed to do to put yourself in a safer place. I think it’s important for people to make good decisions about their own safety.”

I wasn’t sure if this was the right thing to say, but it sounded like she already had more than enough people trying unsuccessfully to talk her out of delusions. And it struck me that, while her sense of danger and persecution was probably illogical, her response to it was completely logical, and showed that she still had a sense of agency and responsibility for her own welfare. That seemed like something worth celebrating, because God knows she didn’t have much else to cheer about.

My fellow passenger’s response to this casual remark was revealing, and in a way, heartbreaking. She stopped talking—a striking change, because she had been talking rapidly and consistently for fifteen minutes. Then she looked at me for a moment, puzzled, and finally said “Thank you for saying that.” For a brief moment, she was able to depart from the compulsive way she talked and thought about herself, and speak spontaneously, directly to me.

You could tell she didn’t hear supportive comments very often, and I’m sure if you’re delusional you hear a lot of very reasonable contradiction from people around you. But the encounter made me realize how hard it must be for mentally ill people to deal with the constant discouragement that comes not from their illness, but from people’s reaction to it. That kind of round-the-clock negative interaction would be difficult to process even if you didn’t have any mental health problems. To be constantly corrected, avoided, feared; to always have an empty seat next to you on the bus. It seems unbearably cruel.

The fear that mental illness inspires in people is the motivation behind a lot of discriminatory laws, policies, and attitudes that separate “them” from “us.” If we identify the mentally ill as dangerous, and shun them, we feel safer. If we call violent people “crazy,” we can explain to ourselves why they are violent. There’s just something wrong with them; they are abnormal.

But in fact violence is depressingly normal. If we face this rather nasty fact, it becomes clear that the problem we should be focusing on isn’t the rare deranged homeless man with a knife, but the larger system we’ve created. In this system the mentally ill not only get no support and no care, but are funneled into prisons that would make any human being more inclined to violence. Forty percent of seriously mentally ill people in America have been in prison at some point in their lives. Almost half. Is that any way to treat an illness?

I’ve said many times that if you make decisions about your safety based on what you wish were true, you’re exposing yourself to much greater dangers down the road. That’s exactly what our attitude toward mental illness has done. The stigma of mental illness is one of the most common and convenient ways that we, as a country, excuse ourselves from solving social problems like poverty, health care, homelessness, and what our wars do to the people who fight them for us. Stigmatizing the mentally ill is a defense mechanism. It allows us to feel safer, but that sense of safety means we leave the real problems unsolved. And that’s far more dangerous.