The Gift of Fear: Survival Signals that Protect us from Violence (16 page)

There is a similar dynamic with more personal predictive questions, such as “Will smoking kill me?” Smokers can easily predict that it will likely kill them, but the outcome is so remote in time that it loses much of its significance.

Is the context of the situation clear to the person making the prediction? Is it possible to evaluate the attendant conditions and circumstances, the relationship of parties and events to each other?

Are there detectable pre-incident indicators that will reliably occur before the outcome being predicted? This is the most valuable of the elements. If one were predicting whether a governor might be the object of an assassination attempt at a speech, pre-incident indicators could include the assassin’s jumping on stage with a gun—but that is too recent a PIN to be very useful (as it provides little time for intervention). The birth of the assassin is also a PIN, but it is too dated to be valuable. Even though both of these events are critical intersections on the map of this particular prediction, one hopes to be somewhere between the two, between the earliest possible detectable factor and those that occur an instant before the act. Useful PINs for assassination might include the assassin’s trying to learn the governor’s schedule, developing a plan, purchasing a weapon, keeping a diary, or telling people “something big is coming.”

Ideally, an outcome would be preceded by several reliable PINs, but they must also be detectable. Someone’s getting the idea to kill and making the decision to kill are both extraordinarily valuable PINs, but since these occur in the mind, they might not be detectable on their own. Later I’ll discuss the PINs for workplace violence, spousal killings, homicides by children, and public-figure attack. They are always there, though not always known to the people making the predictions.

Does the person making the prediction have experience with the specific topic involved? A lion tamer can predict whether or not a lion will attack more accurately than I can because he has experience. He can do an even better job if he has experience with both possible outcomes (lions that do not attack and lions that do).

Can you study or consider outcomes that are comparable—though not necessarily identical—to the one being predicted? Ideally, one relies on events that are substantively comparable. Predicting whether a senator will be shot by a mentally ill member of the general public, one might study cases in which mayors were shot by deranged pursuers, as this is substantively the same situation and the relationship between the players is similar. One can learn about the PINs in the mayor cases and consider whether they apply to the present prediction. On the other hand, studying cases of senators shot by their spouses or senators who shot themselves would not likely improve the success of a prediction about a stranger’s shooting a senator.

Is the person making the prediction objective enough to believe that either outcome is possible? People who believe only one outcome is possible have already completed their prediction. With the simple decision to make a decision before the full range of predictive tests has been completed, they have hit the wall of their intuitive ability. Asked to predict whether a given employee will act violently, the person who believes that kind of thing never happens is not the right choice for the job. People only apply all their predictive resources when they believe either outcome is possible.

To what degree is the person making the prediction invested in the outcome? Simply put, how much does he or she care about avoiding or exploiting the outcome? Does he or she have reason to want the prediction to be correct? If I ask you right now to predict whether I will oversleep tomorrow, you won’t bring your best predictive resources to the question because you don’t care. If, however, you are relying on me to pick you up at the airport early tomorrow morning, your prediction will be far better.

Is it practical to test the exact issue being predicted by trying it first elsewhere? Asked to predict if water in a pot will boil when heated, you need not heat
this
water to improve the prediction. You can test the issue, replicate it exactly, by heating other water first. It is a low-cost experiment for a low-stakes prediction. While replicability is the cornerstone of most scientific predictions, it is nearly useless in high-stakes predictions of human behavior. I cannot test whether an angry employee will shoot a supervisor by giving him a gun and watching him at work.

Does the person making the prediction have
accurate
knowledge about the topic? Unless it is relevant and accurate, knowledge can be the sinking ship the fool insists is sea-worthy, because knowledge often masquerades as wisdom. If a corporate executive has knowledge that most perpetrators of workplace violence are white males between thirty-five and fifty years-old, he might ignore someone’s bizarre behavior because the employee does not “fit the profile.”

(In my firm, we use a predictive instrument that assigns point values to each of these eleven elements. The scale and its ranges appears in
appendix 6
, along with some examples of popular predictions.)

▪ ▪ ▪

The most advanced concept of prediction has to do with deciding just when it is that a thing starts to happen. The prediction of earthquakes gives us an extreme example: There are, contrary to popular belief, reliable pre-incident indicators for earthquakes. The problem is that the PINs might be ten thousand years long, and for this reason earthquakes remain, in human terms, unpredictable. In geological terms, however, it is fair to say that the next earthquake in Los Angeles has already started. In geology, calling something a catastrophe means that the event occurs in a time period short enough to be meaningful to man. The earth’s moving is not the issue, because the ground you are on right now is moving. The suddenness is the issue.

In predicting violence, a pre-incident indicator that takes a long time begs the question of whether we need to wait until something becomes a catastrophe versus trying to detect it at a midway mark. Does an assassination attempt begin when the gun is fired at the victim, or when it is drawn, or when it is carried into the arena, or when it is loaded, or when it is purchased, or when assassination is first thought of? Prediction moves from a science to an art when you realize that
pre-incident indicators are actually part of the incident
.

When you apply this concept to human beings, you can see that behavior is like a chain. Too often, we look at just the individual links. When we ask why a man committed suicide, someone might say, “He was despondent over major financial losses,” as if this could possibly explain it. Many people are despondent over financial losses and don’t kill themselves. Though we want to believe that violence is a matter of cause and effect, it is actually a process, a chain in which the violent outcome is only one link.
The process of suicide starts way before the act of suicide
.

The same is true for homicide. Though we might try to explain a murder using simple cause-and-effect logic (e.g., “He learned his wife was having an affair so he killed her”), it doesn’t aid prediction to think this way. Like the earthquake, violence is one outcome of a process that started way before this man got married. If you were making a prediction of what a friend of yours might do if he lost his job, you wouldn’t say, “Oh, he’ll commit suicide” unless there were many other PINs of suicide present. You’d see the loss of his job as a single link, not the whole chain.

▪ ▪ ▪

By this point, you have read a lot about successful predictions, more perhaps than can easily be recalled. Still, there’s no need for a memory test because the information is already in your mind. I know that because it came from your mind in the first place. These elements of prediction are the same ones our ancestors relied on to survive. If they seem new to you, it’s because they have been largely ignored by modern Westerners. We perceive less need for them because we are at a point in our evolution where life is less about predicting risks and more about controlling them.

Endowed with great intellect with which to protect ourselves, we have developed extraordinary technologies for survival. Chief among them is modern medicine; though we are no less vulnerable to injury, we are far less likely to die from it. Technology has also provided the ability to call for help, so we rarely feel isolated in an emergency. We also have rapid transportation that can rush us to medical care, or rush it to us. Even with all this, we have more fear today than ever before, and most of it is fear of each other.

To be as free from it as possible, we need to recapture our inherent predictive skills. In the following chapters, the elements of prediction and intuition that I’ve discussed will come together in practice. You’ll see that just as hearing intuition is no more than reading the signals we give ourselves, predicting human behavior is no more than reading the signals others give us.

“Man is a coward, plain and simple. He loves life
too much. He fears others too much.”
—
Jack Henry Abbott

“I am going to kill you.” These six words may have triggered more high-stakes predictions than any other sentence ever spoken. They have certainly caused a great deal of fear and anxiety. But why?

Perhaps we believe only a deranged and dangerous person would even think of harming us, but that just isn’t so. Plenty of people have thought of harming you: the driver of the car behind you who felt you were going too slowly, the person waiting to use the pay-phone you were chatting on, the person you fired, the person you walked out on—they have all hosted a fleeting violent idea. Though thoughts of harming you may be terrible, they are also inevitable. The thought is not the problem; the expression of the thought is what causes us anxiety, and most of the time that’s the whole idea. Understanding this will help reduce unwarranted fear.

That someone would intrude on our peace of mind, that they would speak words so difficult to take back, that they would exploit our fear, that they would care so little about us, that they would raise the stakes so high, that they would stoop so low—all of this alarms us, and by design.

Threatening words are dispatched like soldiers under strict orders: Cause anxiety that cannot be ignored. Surprisingly, their deployment isn’t entirely bad news. It’s bad, of course, that someone threatens violence, but the threat means that at least for now, he has considered violence and decided against doing it. The threat means that at least for now (and usually forever), he favors words that alarm over actions that harm.

For an instrument of communication used so frequently, the threat is little understood, until you think about it. The parent who threatens punishment, the lawyer who threatens unspecified “further action,” the head of state who threatens war, the ex-husband who threatens murder, the child who threatens to make a scene—all are using words with the exact same intent: to cause uncertainty.

Our social world relies on our investing some threats with credibility while discounting others. Our belief that they really will tow the car if we leave it here encourages us to look for a parking space unencumbered by that particular threat. The disbelief that our joking spouse will really kill us if we are late to dinner allows us to stay in the marriage. Threats, you see, are not the issue—context is the issue.

For example, as you watch two people argue, an escalation of hostility that would otherwise cause alarm causes none if it is happening between actors on stage at the theater. Conversely, behavior that is not normally threatening, such as a man’s walking up some stairs, becomes alarming when it is an uninvited audience member marching up onto that same stage. It is context that gives meaning to the few steps he takes.

A single word between intimates, perhaps meaningless to others, might carry a strong message of love or threat, depending on context. Context is the necessary link that gives meaning to everything we observe.

Imagine a man arriving for work one morning. He does not go in the unlocked front door where most people enter the building but instead goes around to a back entrance. When he sees someone ahead of him use a key to get in, he runs up and catches the door before it re-locks. Once he is inside the building, he barely responds as a co-worker calls out, “The boss wants to see you.” “Yeah, I want to see him too,” the man says quietly. He is carrying a gym bag, but it appears too heavy to contain just clothes. Before going to his boss’s office, he stops in the locker room, reaches into the bag, and pulls out a pistol. He takes a second handgun from the bag and conceals both of them beneath his coat. Now he looks for his boss.

If we stopped right here, and you had to predict this man’s likely behavior on the basis of what you know, context would tell the tale, because to know just one thing changes every other thing: This man is a police detective. If he were a postal worker, your prediction would be different.

▪ ▪ ▪

Though knowing context is key to predicting which threats will be acted upon, people are often reluctant to put it ahead of content. Even some experts believe that threat assessments are aided by identifying and considering so-called key words. The assumption is that these words are significant by their presence alone, but the practice is rarely enlightening. As a person creates a communication, his selection of words is part of that creation, but they are instruments, not the final product.