Illustration by Jim Cooke

When people think of someone with an “addictive personality,” the image typically isn’t a pretty one. “When is an addict lying?” goes a joke told by addiction counselors: the snide answer is “when his lips are moving.” Media portrayals of addiction tend to depict people with addictions as “fiends” or “demons” whose debauchery is driven by a ravenous hedonism, not a human or understandable search for safety and comfort. Consequently, the “addictive personality” is seen as a bad one: weak, unreliable, selfish, and out of control.

The temperament from which addiction springs is also seen as defective, unable to resist temptation. Even when we laugh about having an addictive personality it’s usually to justify an indulgence or to signal our guilt about pleasure, even if only ironically.

However, the idea that people with addiction are all alike in these extremely negative ways is inaccurate— based more on the racist history of American drug policy and the criminalization of certain drugs than it is on science. Although addiction was originally framed by both Alcoholics Anonymous and psychiatry as a form of antisocial personality or “character” disorder, research did not confirm this idea. Despite decades of attempts, no single addictive personality common to everyone with addictions has ever been found.

If you have come to believe that you yourself or an addicted loved one, by nature of having addiction, has a defective or selfish personality, you have been misled. As George Koob, the director of the National Institute on Alcohol Abuse and Alcoholism, told me when I interviewed him for Nature, “What we’re finding is that the addictive personality, if you will, is multifaceted,” says Koob. “It doesn’t really exist as an entity of its own.”

Basically, the idea of a general addictive personality is a myth. Research finds no universal character traits that are common to all addicted people. Only half have more than one addiction (not including cigarettes)—and many can control their engagement with some addictive substances or activities, but not others. If there were an addictive personality, this shouldn’t be possible.

Some are shy; some are bold. Some are fundamentally kind and caring; some are cruel. Some tend toward honesty; others not so much. The whole range of human character can be found among people with addictions, despite the cruel stereotypes that are typically presented.

Only 18% of addicts, for example, have a personality disorder characterized by lying, stealing, lack of conscience, and manipulative antisocial behavior. This is more than four times the rate seen in typical people, but it still means that 82% of us don’t fit that particular caricature of addiction.

Although people with addictions or potential addicts cannot be identified by a specific collection of personality traits, however, it is often possible to tell quite early on which children are at high risk. Children who ultimately develop addictions tend to be outliers in a number of measurable ways. Yes, some stand out because they are antisocial and callous—but others stand out because they are overly moralistic and sensitive.

While those who are the most impulsive and eager to try new things are at highest risk, the odds of addiction are also elevated in those who are compulsive and fear novelty. It is extremes of personality and temperament—some of which are associated with talents, not deficits—that elevates risk. Giftedness and high IQ, for instance, are linked with higher rates of illegal drug use than having average intelligence.

Whether these extreme traits lead to addictions, other compulsive behaviors, developmental differences, mental illnesses, or some mixture depends not just on genetics, but also on the environment, people’s own reactions to it, and those of others to them. Addictions and other neurodevelopmental disorders rely not just on our actual experience but on how we interpret it and how our parents and friends respond to and label the way we behave. They develop in brains designed to change with experience—and that leaves us vulnerable to learning things that create damaging patterns, not just useful habits.

The impact of all these factors together can be seen most clearly in studies that follow participants from infancy into adulthood (which are rare because they take so long to conduct and are thus very expensive). In these types of data, some strong patterns emerge. One of the earliest and best known longitudinal studies related to drug use followed 101 children—mainly middle class, two-thirds white—raised in Berkeley in the 1970s.

Conducted by psychologists Jonathan Shedler and Jack Block, then at the University of California, the research was published in 1990 and its main finding generated much controversy. The authors discovered that the most mentally and psychologically healthy teens were not those who abstained entirely from alcohol and other drugs, but rather the kids who experimented with weed and drinking, but didn’t overdo it. In this study, occasional teen drinking and marijuana use was normal adolescent behavior. However, while it was common, it was typically not problematic.

Unsurprisingly the teens who became frequent users and drinkers had the problems you might expect like depression, anxiety, and delinquent behavior. Then again, many of the same psychiatric problems were also seen in the adolescents who rejected the idea of drinking and drugs entirely.

That’s probably because, in order to avoid any experimentation as a kid growing up around the Berkeley campus in the ’70s (when nearly two thirds of high school seniors nationally reported at least trying marijuana), you’d have to be either a loner with few friends or a person who was unusually fearful and/or resistant to peer pressure. Not using drugs may well have been a wise choice for these youth—but good decisions aren’t always made for healthy reasons.

And indeed, that’s exactly what the study found. The youth who abstained did not tend to do so because they rationally recognized the risks. Instead, they were overly anxious, uptight, and lacking in social skills; some may not have had to say no because they didn’t even get the chance to say yes. Similar data have been published on teen drinking as well. Moderate drinkers—not nondrinkers—are the most well adjusted, at least in countries where drinking is a social norm. The healthiest patterns are found in the middle of the curve, not at the extremes.

To understand how having these outlying traits increases risk for addiction, we have to look at how they affect development. Critically, in Shedler and Block’s data, the traits that marked both abstainers and heavy users could be seen long before drug use began. After all, the authors had started following these children in preschool. Once they knew how the participants behaved in adolescence, they could look back and see what early traits were linked to particular problems.

Longitudinal studies looking at addiction risk like Shedler and Block’s have found three major pathways to it that involve temperamental traits, all of which can be seen in nascent form in young children. The first, which is more common in males, involves impulsivity, boldness, and a desire for new experience; it can lead to addiction because it makes it hard for people to control their own behavior. Preschoolers who were more impulsive than others the same age tended to stay that way, putting them at risk.

The second path, which tends to be seen more in women, involves being sad, inhibited, and/or anxious. While these negative emotions can also deter experimentation, when they do not do so, people may find themselves on a “self-medicating” path to addiction, where drugs are used to cope with painful feelings. People in this group who were shy, withdrawn preschoolers tended to drink or take other drugs “to fit in” and then stuck with them.

Being bold and adventurous and being sad and cautious seem like opposite personality types. However, these two paths to addiction are actually not mutually exclusive. The third way involves having both kinds of traits, where people alternatively fear and desire novelty and behavior swings from being impulsive and rash to being compulsive, fear driven, and stuck in rigid patterns.

This is where some of the contradictions that have long confounded the study of addiction come into play—namely, some aspects seem precisely planned out, while others are obviously related to lack of restraint. My own story spirals around this paradoxical situation: I was driven enough to excel academically and fundamentally scared of change and of other people—yet I was also reckless enough to sell cocaine and shoot heroin.

If we look more closely, however, the paradoxes disappear. All three pathways really involve the same fundamental problem: a difficulty with self-regulation. This may appear predominantly as an inability to inhibit strong impulses, it may be largely an impairment in modulating negative emotions like anxiety, or it may have elements of both. In any case, difficulties with self-regulation lay the groundwork for learning addiction and for creating a condition that is hard to understand. The brain regions that allow self-regulation need experience and practice in order to develop. If that experience is aberrant or if those brain regions are wired unusually, they may not learn to work properly.

A problem with self-regulation, however, is not a personality disorder. If we want better ways to manage addiction, we have to recognize that the “addictive personality” is a harmful stereotype— not a useful concept to guide treatment or prevention.

This is an excerpt from the new book UNBROKEN BRAIN: A Revolutionary New Way of Understanding Addiction by Maia Szalavitz. Copyright © 2016 by the author and reprinted by permission of St. Martin’s Press, LLC.