Chapter 3 Discover Mode and the Need for Risky Play

Chapter 3

Discover Mode and the Need for Risky Play

In recent decades, America and many other Western nations made two contradictory choices about children’s safety, and both were wrong. We decided that the real world was so full of dangers that children should not be allowed to explore it without adult supervision, even though the risks to children from crime, violence, drunk drivers, and most other sources have dropped steeply since the 1990s. [1] At the same time, it seemed like too much of a bother to design and require age-appropriate guardrails for kids online, so we left children free to wander through the Wild West of the virtual world, where threats to children abounded.

To take one example of our shortsightedness, a powerful fear for many parents is that their child will fall into the hands of a sexual predator. But sex criminals nowadays spend most of their time in the virtual world because the internet makes it so much easier to communicate with children and to find and circulate sexual and violent videos involving children. To quote a 2019 New York Times article, “Tech companies are reporting a boom in online photos and videos of children being sexually abused—a record 45 million illegal images were flagged last year alone—exposing a system at a breaking point and unable to keep up with the perpetrators.” [2] More recently, in 2023, The Wall Street Journal ran an exposé that showed how “Instagram connects pedophiles and guides them to content sellers via recommendation systems that excel at linking those who share niche interests.” [3]

To offer another example: Isabel Hogben, a 14-year-old girl in Rhode Island, wrote an essay in The Free Press that demonstrated how American parents are focusing on the wrong threats:

I was ten years old when I watched porn for the first time. I found myself on Pornhub, which I stumbled across by accident and returned to out of curiosity. The website has no age verification, no ID requirement, not even a prompt asking me if I was over 18. The site is easy to find, impossible to avoid, and has become a frequent rite of passage for kids my age. Where was my mother? In the next room, making sure I was eating nine differently colored fruits and vegetables on the daily. She was attentive, nearly a helicopter parent, but I found online porn anyway. So did my friends.

Hogben’s essay is a succinct illustration of the principle that we are overprotecting our children in the real world while underprotecting them online . If we really want to keep our children safe, we should delay their entry into the virtual world and send them out to play in the real world instead.

Unsupervised outdoor play teaches children how to handle risks and challenges of many kinds. By building physical, psychological, and social competence, it gives kids confidence that they can face new situations, which is an inoculation against anxiety. In this chapter, I’ll show that a healthy human childhood with a lot of autonomy and unsupervised play in the real world sets children’s brains to operate mostly in “discover mode,” with a well-developed attachment system and an ability to handle the risks of daily life. Conversely, when there is society-wide pressure on parents to adopt modern overprotective parenting, it sets children’s brains to operate mostly in “defend mode,” with less secure attachment and reduced ability to evaluate or handle risk. Let me explain what these terms mean, and why discover mode is one of the keys to helping the anxious generation.

Discover Mode Versus Defend Mode

The environments that shaped hominid evolution over the last few million years were extraordinarily variable, with periods of safety and abundance alternating with periods of scarcity, danger, drought, and starvation. [4] Our ancestors needed psychological adaptations to help them thrive in both settings. The variability of our environments shaped and refined older brain networks into two systems that are specialized for those two kinds of situations. The behavioral activation system (or BAS) turns on when you detect opportunities, such as suddenly coming across a tree full of ripe cherries when you and your group are hungry. [5] You’re flooded with positive emotions and shared excitement, your mouths may begin to water, and everyone is ready to go! I’ll give BAS a more intuitive name: discover mode. [6]

The behavioral inhibition system (BIS), in contrast, turns on when threats are detected, such as hearing a leopard roar nearby as you’re picking those cherries. You all stop what you are doing. Appetite is suppressed as your bodies flood with stress hormones and your thinking turns entirely to identifying the threat and finding ways to escape it. I’ll refer to BIS as defend mode . For people with chronic anxiety, defend mode is chronically activated.

The two systems together form a mechanism for quickly adapting to changing conditions, like a thermostat that can activate either a heating system or a cooling system as the temperature fluctuates. Across species, the default setting of the overall system depends on the animal’s evolutionary history and expected environment. Animals that evolved with little daily risk of sudden death (such as top predators in a food chain, or herbivores on an island with no predators) often seem serene and confident. They are willing to get close to humans. Their default setting is discover mode, although they will shift into defend mode if attacked. In contrast, animals such as rabbits and deer, which evolved in the presence of constant predation, are skittish; they are quick to bolt and run. Their default setting is defend mode, and they shift into discover mode only slowly and tentatively when they perceive that the environment is unusually safe.

In humans (and other highly sociable mammals, such as dogs), the default setting is a major contributor to their individual personality. People (and dogs) who go through life in discover mode (except when directly threatened) are happier, more sociable, and more eager for new experiences. Conversely, people (and dogs) who are chronically in defend mode are more defensive and anxious, and they have only rare moments of perceived safety. They tend to see new situations, people, and ideas as potential threats, rather than as opportunities. Such chronic wariness was adaptive in some ancient environments, and may still be today for children raised in unstable and violent settings. But being stuck in defend mode is an obstacle to learning and growth in the physically safe environments that surround most children in developed nations today.

Students on the Defensive

Discover mode fosters learning and growth. If we want to help young people thrive—at home, in school, and in the workplace—shifting them into discover mode may be the most effective change we can make. Let me lay out the differences between the modes as we might see them in a college student. Figure 3.1 shows what a student arriving at a university would look like if her childhood (and her genes) gave her a brain whose default setting was discover mode versus defend mode. It’s obvious that students in discover mode will profit and grow rapidly from the bountiful intellectual and social opportunities of a university. Students who spend most of their time in defend mode will learn less and grow less.

This contrast explains the sudden change that happened on many college campuses around 2014. Figure 3.2 shows how the distribution of mental challenges changed as the first members of Gen Z arrived and the last members of the millennial generation began to graduate. The only disorders that rose rapidly were psychological disorders. Those disorders were overwhelmingly anxiety and depression.

Two Basic Mindsets

Discover mode (BAS)

Scan for opportunities

Kid in a candy shop

Think for yourself

Let me grow!

Defend mode (BIS)

Scan for dangers

Scarcity mindset

Cling to your team

Keep me safe!

Figure 3.1. Discover mode versus defend mode, for a student arriving at a university.

As soon as Gen Z arrived on campus, college counseling centers were overwhelmed. [7] The previously exuberant culture of millennial students in discover mode gave way to a more anxious culture of Gen Z students in defend mode. Books, words, speakers, and ideas that caused little or no controversy in 2010 were, by 2015, said to be harmful, dangerous, or traumatizing. America’s residential universities are not perfect, but they are among the safest, most welcoming and inclusive environments ever created for young adults. Yet campus culture changed around 2015, not just in the United States but also at British [9] and Canadian [10] universities. How could such a big change happen so quickly and internationally?

Self-Reported Disabilities, College Freshmen

Figure 3.2 . Percentage of U.S. college freshmen reporting various kinds of disabilities and disorders. (Source: Annual Freshman Survey, by UCLA’s Higher Education Research Institute.) [8]

In the rest of this chapter, I’ll show how the play-based childhood is nature’s way of wiring up brains that tend toward discover mode, and how the phone-based childhood shifted a generation of children toward defend mode.

Kids Are Antifragile

In the late 1980s, a grand experiment was launched in the Arizona desert. Biosphere 2 was (and still is) the largest attempt to build a closed artificial ecosystem, as a prelude to (someday) building self-sustaining ecosystems in outer space. Biosphere 2 was designed to support eight people, who would attempt to live within it for several years. All of the oxygen they breathed, the water they drank, and the food they ate was to be generated within the facility.

That goal was never reached. The complexity of biological interactions among species and social interactions among humans proved to be too much, but a great deal was learned from the multiple failures. For instance, many of the trees they planted to create a rain-forest ecosystem grew rapidly but then fell over before reaching maturity. The designers had not realized that young trees need wind to grow properly. When the wind blows, it bends the tree, which tugs at the roots on the windward side and compresses the wood on the other side. In response, the root system expands to provide a firmer anchor where it is needed, and the compressed wood cells change their structure to become stronger and firmer.

This altered cell structure is called reaction wood, or sometimes stress wood. Trees that are exposed to strong winds early in life become trees that can withstand even stronger winds when full grown. Conversely, trees that are raised in a protected greenhouse sometimes fall over from their own weight before they reach maturity.

Stress wood is a perfect metaphor for children, who also need to experience frequent stressors in order to become strong adults. The Biosphere trees illustrate the concept of “antifragility,” a term coined by my NYU colleague Nassim Taleb in his 2012 book, Antifragile: Things That Gain from Disorder . Taleb noted that some things, like wineglasses, are fragile. We protect fragile things from shocks and threats because we know they cannot withstand even a gentle challenge, such as being knocked over on a dinner table. Other things are resilient, such as a plastic cup, which can withstand being knocked off the table. But resilient objects don’t get better from getting dropped; they merely don’t get worse.

Taleb coined the word “antifragile” to describe things that actually need to get knocked over now and then in order to become strong. I used the word “things,” but there are very few inanimate objects that are antifragile. Rather, antifragility is a common property of complex systems that were designed (by evolution, and sometimes by people) to function in a world that is unpredictable. [11] The ultimate antifragile system is the immune system, which requires early exposure to dirt, parasites, and bacteria in order to set itself up in childhood. Parents who try to raise their children in a bubble of perfect hygiene are harming their children by blocking the development of their antifragile immune systems.

It’s the same dynamic for what has been called the psychological immune system [12] —the ability of a child to handle, process, and get past frustrations, minor accidents, teasing, exclusion, perceived injustices, and normal conflicts without falling prey to hours or days of inner turmoil. There is no way to live with other humans without conflicts and deprivations. As the Stoics and Buddhists taught long ago, happiness cannot be reached by eliminating all “triggers” from life; rather, happiness comes from learning to deprive external events of the power to trigger negative emotions in you. In fact, the best parenting book [13] that my wife and I read when our children were toddlers urged us to look for opportunities to frustrate our children every day by laying out and enforcing the contingencies of life: If you want to watch Teletubbies , you must first put away your toys. If you persist in doing that, you’ll get a time-out. Yes, your sister got something you didn’t, and that happens sometimes.

Well-intentioned parents who try to raise their children in a bubble of satisfaction, protected from frustration, consequences, and negative emotions, may be harming their children. They may be blocking the development of competence, self-control, frustration tolerance, and emotional self-management. Several studies find that such “coddling” or “helicopter parenting” is correlated with later anxiety disorders, low self-efficacy (which is the inner confidence that one can do what is needed to reach one’s goals), and difficulty adjusting to college. [14]

Children are intrinsically antifragile, which is why overprotected children are more likely to become adolescents who are stuck in defend mode. In defend mode, they’re likely to learn less, have fewer close friends, be more anxious, and experience more pain from ordinary conversations and conflicts.

Antifragile Kids Need Risky Play to Stay in Discover Mode

Antifragility is the key to solving many puzzles about human development, such as this one: Why do children add risk to their play? Why is it that once a skill is mastered, such as skateboarding down a gentle slope, a child will move on to a steeper slope, then a staircase, then perhaps the staircase railing? Why would children choose activities that pretty much guarantee that they’ll get hurt, multiple times? Play researchers have long known the answer. As the Norwegian researchers Ellen Sandseter and Leif Kennair wrote in 2010, thrilling experiences have anti-phobic effects. [15]

Sandseter and Kennair begin with a puzzling fact long known in clinical psychology: Phobias are concentrated around a few animals and situations that kill almost nobody, such as snakes (even tiny ones), tightly enclosed places, the dark, public speaking, and heights. Conversely, very few people develop phobias to things that kill many modern people, including cars, opioids, knives, guns, and junk food. Furthermore, phobias in adults can rarely be traced to a bad experience in childhood. [16] In fact, kids who fall out of trees often turn into the adults who are least afraid of climbing trees.

We can resolve the puzzle by taking an evolutionary view. Common phobias evolved over millions of years of hunter-gatherer life, with some (such as snakes) being shared by other primates. We have an “evolved preparedness” to pay attention to some things, such as snakes, and to acquire a fear very easily from a single bad experience or from seeing others in our group show fear toward snakes. Conversely, as a child gains exposure, experience, and mastery, fear usually recedes.

As children become more competent, they become increasingly more intrigued by some of the things that had frightened them. They may approach them, look to adults and older kids for guidance, learn to distinguish the dangerous situations from the less dangerous ones, and eventually master their fears. As they do so, their fear turns into thrill and triumph. You can see the transition on a young child’s face as he reaches out to touch a worm under a rock you just lifted up for him on a nature walk. You can see the mix of fear and fascination turning into a shriek of delight and disgust as he pulls his finger away, laughing. He did it! Now he’ll be less afraid the next time he encounters a worm.

While I was writing this chapter, in the fall of 2022, my family got a puppy. Wilma is a small dog, and she weighed only seven pounds when we first started taking her for walks on the crowded sidewalks of New York City. At first she was visibly afraid of everything, including the parade of larger dogs, and she had trouble relaxing enough to “do her business.”

Over time, she habituated somewhat, and I began to let her run off-leash, early mornings, in parks with other dogs. There too she was afraid at first, but the way she handled it made it seem as though she had read Sandseter and Kennair. She would approach much larger dogs, slowly, and then bolt away like lightning when they’d take a step toward her. Sometimes she’d run toward me for safety, but then her anti-phobic programming would kick in. Without slowing down, she’d execute a high-speed turn around my legs and sprint back toward the larger dog for another round of thrills. She was experimenting to find the balance of joy and fear that she was ready for at that moment. By repeatedly cycling through discover and defend mode, she learned how to size up the intentions of other dogs and she developed her own abilities to engage in rough and joyful play, even as she occasionally got knocked over in a scramble of paws and tails.

Figure 3.3. Wilma, age 7 months, executing a hairpin turn as her sprint toward a German shepherd sharply angled into a sprint away, which was followed by play position and more sprinting toward the larger dog. You can see the video of this interaction in the online supplement.

Kids and puppies are thrill seekers. They are hungry for thrills, and they must get them if they are to overcome their childhood fears and wire up their brains so that discover mode becomes the default. Children need to swing and then jump off the swing. They need to explore forests and junkyards in search of novelty and adventure. They need to shriek with their friends while watching a horror movie or riding a roller coaster. In the process they develop a broad set of competences, including the ability to judge risk for themselves, take appropriate action when faced with risks, and learn that when things go wrong, even if they get hurt, they can usually handle it without calling in an adult.

Sandseter and Kennair define risky play as “thrilling and exciting forms of play that involve a risk of physical injury.” (In a 2023 paper, expanding on their original work, they add that risky play also requires elements of uncertainty. [17] ) They note that such play usually takes place outdoors, during free-play time rather than during activities organized by adults. Children choose to do activities that often lead to relatively harmless injuries, particularly bruises and cuts.

Figure 3.4. An overly dangerous playground in Dallas, Texas, year unknown. [18]

Sandseter and Kennair analyzed the kinds of risks that children seek out when adults give them some freedom, and they found six: heights (such as climbing trees or playground structures), high speed (such as swinging, or going down fast slides), dangerous tools (such as hammers and drills), dangerous elements (such as experimenting with fire), rough-and-tumble play (such as wrestling), and disappearing (hiding, wandering away, potentially getting lost or separated). These are the major types of thrills that children need. They’ll get them for themselves unless adults stop them—which we did in the 1990s. Note that video games offer none of these risks, even though games such as Fortnite show avatars doing all of them. [19] We are embodied creatures; children should learn how to manage their bodies in the physical world before they start spending large amounts of time in the virtual world.

You can see children seeking out risks and thrills, together, in many playground photos taken before the 1980s. [20] Some of them, such as figure 3.4, show playgrounds that are clearly too dangerous. If children fell from such a great height, they could suffer severe injury, perhaps even a broken neck.

In contrast, figure 3.5 shows a playground spinner (or merry-go-round), which is, in my opinion, the greatest piece of playground equipment ever invented. It requires cooperation to get going: the more kids who join in, the faster it goes and the more screaming there is, both of which amplify the thrills. You get physical sensations from the centrifugal force that you don’t get anywhere else, which makes it educational as well as experientially unique. You get consciousness alteration if you lie in the center (dizziness). To top it all off, it offers endless opportunities for additional risk-taking such as standing up, hanging off the sides, or throwing a ball with the other kids while it’s spinning.

Figure 3.5. A playground spinner (or merry-go-round), a staple of 1970s playgrounds. [21]

On the playground spinner you can get hurt if you’re not careful, but not badly hurt, which means you get direct feedback from your own skillful and unskillful moves. You learn how to handle your body and how to keep yourself and others safe. Researchers who study children at play have concluded that the risk of minor injuries should be a feature, not a bug, in playground design. In the U.K., they are acting on this insight, adding construction materials, hammers, and other tools (which are used with adult supervision). [22] As one enlightened summer camp administrator told me, “We want to see bruises, not scars.”

Unfortunately, playground spinners are rare nowadays, because they carry some risk, and therefore in a litigious country like the United States they carry some risk of a lawsuit against whoever is responsible for the playground. You can see the decimation of risky play since the 1990s in most American playgrounds. Figure 3.6 shows the most common kind of structure in the playgrounds my children used in New York City in the early 2010s. It’s hard to hurt yourself on these things, which means children don’t learn much about how to not get hurt.

Figure 3.6. An overly safe playground, offering little opportunity for antifragile kids to learn how to not get hurt. [23]

These ultrasafe structures were entertaining when my kids were three or four, but by age 6 they wanted bigger thrills, which they found at Coney Island. Amusement parks around the world are designed to give children two of Sandseter and Kennair’s six kinds of thrills: heights and high speeds. The rides offer differing doses of fear and thrill (with close to zero risk of injury), and a major topic of conversation in the car whenever I took my kids and their friends to Coney Island was, who is going to try which scary ride today?

Figure 3.7. Coney Island, New York City, offers a wide range of dosages of thrills. [24]

Perhaps your first reaction to those old playground photos is “good riddance!” What parent wants to take any risks with their child? But the harms of eliminating all risky outdoor play are substantial. While writing this chapter, I met with Mariana Brussoni, a play researcher at the University of British Columbia. Brussoni guided me to research showing that the risk of injury per hour of physical play is lower than the risk per hour of playing adult-guided sports, while conferring many more developmental benefits (because the children must make all choices, set and enforce rules, and resolve all disputes). [25] Brussoni is on a campaign to encourage risky outdoor play because in the long run it produces the healthiest children. [26] Our goal in designing the places children play, she says, should be to “keep them as safe as necessary, not as safe as possible.” [27]

The play researchers Brussoni, Sandseter and Kennair, and Peter Gray all help us see that antifragile children need play that involves some risk to develop competence and overcome their childhood anxieties. Like my dog, Wilma, only the kids themselves can calibrate the level of risk they are ready for at each moment as they tune up their experience-expectant brains. Like young trees exposed to wind, children who are routinely exposed to small risks grow up to become adults who can handle much larger risks without panicking. Conversely, children who are raised in a protected greenhouse sometimes become incapacitated by anxiety before they reach maturity.

I am often asked why I urge parents to be more vigilant and restrictive about their children’s online activities when I’ve been talking for years about how parents need to stop over-supervising their children and start giving them independence. Can’t children just as well become antifragile online? Don’t they experience setbacks, stressors, and challenges there?

I see few indications that a phone-based childhood develops antifragility. Human childhood evolved in the real world, and children’s minds are “expecting” the challenges of the real world, which is embodied, synchronous, and one-to-one or one-to-several, within communities that endure. For physical development they need physical play and physical risk-taking. Virtual battles in a video game confer little or no physical benefit. For social development they need to learn the art of friendship, which is embodied; friends do things together, and as children they touch, hug, and wrestle. Mistakes are low cost, and can be rectified in real time. Moreover, there are clear embodied signals of this rectification, such as an apology with an appropriate facial expression. A smile, a pat on the back, or a handshake shows everyone that it’s okay, both parties are ready to move on and continue playing, both are developing their skills of relationship repair. In contrast, as young people move their social relationships online, those relationships become disembodied, asynchronous, and sometimes disposable. Even small mistakes can bring heavy costs in a viral world where content can live forever and everyone can see it. Mistakes can be met with intense criticism by multiple individuals with whom one has no underlying bond. Apologies are often mocked, and any signal of re-acceptance can be mixed or vague. Instead of gaining an experience of social mastery, a child is often left with a sense of social incompetence, loss of status, and anxiety about future social interactions.

This is why there is no contradiction when I say that parents should supervise less in the real world but more in the virtual—primarily by delaying immersion. Childhood evolved on Earth, and children’s antifragility is geared toward the characteristics of Earth. Small mistakes promote growth and learning. But if you raise children on Mars, there’s a mismatch between children’s needs and what the environment offers. If a child falls down on Mars and cracks the face shield of their spacesuit, it’s instant death. Mars is unforgiving, and life there would require living in defend mode. Of course, the online world is not nearly as dangerous as Mars, but it shares the property that small mistakes can bring enormous costs. Children did not evolve to handle the virality, anonymity, instability, and potential for large-scale public shaming of the virtual world. Even adults have trouble with it.

We are misallocating our protective efforts. We should be giving children more of the practice they need in the real world and delaying their entry into the online world, where the benefits are fewer and the guardrails nearly nonexistent.

The Beginning of the End of Play-Based Childhood

At what age were you given freedom? How old were you when your parents let you walk alone to a friend’s home, at least a quarter mile away, or allowed you and your friends to be out on your own, going to parks or shops, with no supervision? I have asked this question to dozens of audiences, and I always find the same generational differences.

First I ask everyone who was born before 1981 to raise their hands. These are the members of Gen X (born 1965–1980), the baby boomers (born 1946–1964), and the last members of the so-called Silent Generation (born 1928–1945). I ask these older audience members to recall their age of liberation privately and then to shout it out when I point to their section of the room. Nearly everyone shouts out “6,” “7,” or “8,” and it is sometimes hard for me to continue the demonstration because they are laughing and fondly recounting to each other the grand adventures they used to have with the other kids in their neighborhood. Next I ask everyone who was born in 1996 or later (Gen Z) to raise their hands. When I ask them to shout out their liberation age, the difference is stark: The majority fall between 10 and 12, with just a few 8s, 9s, 13s, and 14s. (Members of the millennial generation fall in between and show a wide range of liberation ages.)

These findings are confirmed by more rigorous research. In the United States, [28] Canada, [29] and Britain, [30] children used to have a great deal of freedom to walk to school, roam around their neighborhoods, invent games, get into conflicts, and resolve those conflicts, beginning around first or second grade. But in the 1990s, parenting changed in all three countries. It became more intensive, protective, and fearful.

Corresponding to the crackdown, studies of how Americans spend their time show a sudden change in the 1990s. Women had been entering the workforce in large numbers since the 1970s, giving them far less time at home. Yet despite growing time pressures, mothers as well as fathers began reporting that they spent a lot more time with their children, beginning rather suddenly in the mid-1990s. Figure 3.8 shows the changing number of hours per week that mothers reported spending with their children from 1965 through 2008. The number is steady or slightly declining, for mothers with and without college degrees, all the way until 1995, and then it jumps up, especially for college-educated mothers. The graph for fathers is quite similar, just with lower numbers (around four hours per week until 1995, then jumping up to around eight hours per week by 2000).

Time Spent Parenting by Mothers

Figure 3.8. Time spent parenting by U.S. mothers. Parenting time suddenly increased in the mid-1990s—the beginning of Gen Z. (Source: Ramey the rising number of women working and the corresponding increase in day care and after-school programs; and the increasing influence of parenting “experts,” whose advice was often a better reflection of their social and political views than of any scientific consensus.

Furedi says that there is one factor above all others that created the conditions for the 1990s turn to paranoid parenting: “the breakdown of adult solidarity.” As Furedi explains,

Across cultures and throughout history, mothers and fathers have acted on the assumption that if their children got into trouble, other adults—often strangers—would help out. In many societies adults feel duty-bound to reprimand other people’s children who misbehave in public.

But in Britain and America, the 1980s and 1990s saw repeated news stories about adults abusing children, from day care centers and sports leagues to the Boy Scouts and the Catholic Church. Some of these cases were true horror stories about institutions that had sheltered child abusers for decades in order to avoid bad publicity. Some of the cases were fabrications and moral panics [41] —in particular, those in which employees at day care centers were accused of carrying out bizarre sexual or satanic rituals. (The accusations were made by very young children, who, it later turned out, had invented imaginative stories in response to leading questions from overzealous adults. [42] )

These scandals—real and fake—led to better detection and reporting mechanisms to catch abusers and hold institutions responsible for sheltering them. Their tragic side effect, however, was a generalized sense that no adults could be trusted to be alone with children. Children were taught to fear unknown adults, particularly men. According to Google’s Ngram viewer (which charts the frequency of words and terms in all books published each year), the term “stranger danger” first appeared in English-language books in the early 1980s; then its frequency leveled off until the mid-1990s, after which it rose rapidly. At the same time, adults internalized the reciprocal message: Stay away from other people’s children. Don’t talk to them; don’t discipline them if they are misbehaving; don’t get involved.

But when adults step away and stop helping each other to raise children, parents find themselves on their own. Parenting becomes harder, more fear-ridden, and more time consuming, especially for women, as we saw in figure 3.8.

Furedi offered an important qualification about the scope of the problem: “The idea that responsible parenting means the continual supervision of children is a peculiarly Anglo-American one.” [43] He noted that children in Europe, from Italy to Scandinavia, and in many other parts of the world, enjoyed far greater freedom to play and explore the outside world than did children in the U.K. and the United States. He cited a study showing that parents in Germany and Scandinavia were much more likely to let their young children walk to school than those in the U.K., who felt compelled to drive their children even short distances. [44]

It is this rise of fearful parenting in the 1990s that led to the evaporation of unsupervised children from public spaces in the Anglosphere by the year 2000. By almost any measure, children were safer in public than they had been in a very long time in terms of risks from crime, sex offenders, and even drunk drivers, all of which had been present at much higher levels in previous decades. [45] And once unsupervised children became a rarity, the occasional sighting of one was enough to cause some neighbors to call 911, bringing down the police, Child Protective Services, and occasionally jail time for anyone who dared to give their child the independence they themselves had enjoyed 30 years earlier. [46]

This is the world in which Gen Z was raised. It was a world in which adults, schools, and other institutions worked together to teach children that the world is dangerous, and to prevent them from experiencing the risks, conflicts, and thrills that their experience-expectant brains needed to overcome anxiety and set their default mental state to discover mode. [47]

Safetyism and Concept Creep

The Australian psychologist Nick Haslam originated the term “concept creep,”? [48] which refers to the expansion of psychological concepts in recent decades in two directions: downward (to apply to smaller or more trivial cases) and outward (to encompass new and conceptually unrelated phenomena). You can see concept creep in action by observing the expansion of terms like “addiction,” “trauma,” “abuse,” and “safety.” For most of the 20th century, the word “safety” referred almost exclusively to physical safety. It was only in the late 1980s that the term “emotional safety” began to show up at more than trace levels in Google’s Ngram viewer. From 1985 to 2010, at the start of the Great Rewiring, the term’s frequency rose rapidly and steadily, a 600% increase. [49]

Physical safety is a good thing, of course. No sane person objects to the use of seat belts and smoke alarms. There is also an important concept called psychological safety, which refers to the shared belief in a group that members won’t be punished or humiliated for speaking up, so people are willing to take risks in sharing ideas and debating them. [50] Psychological safety is among the best indicators of a healthy workplace culture. But in a psychologically safe group, members can disagree with each other and criticize each other’s ideas respectfully. That’s how ideas get vetted. What emerged on campus as emotional safety, in contrast, was a much broader concept that came to mean this: I should not have to experience negative emotions because of what someone else said or did. I have a right not to be “triggered.”

“We’ve created a safe, nonjudgmental environment that will leave your child ill-prepared for real life.”

Figure 3.9. New Yorker cartoon by W. Haefeli. [51]

In The Coddling of the American Mind , Greg and I found that the concept of safety had undergone such extensive concept creep among Gen Z and many of the educators and therapists around them that it had become a pervasive and unquestionable value. We used the term “safetyism” to refer to “a culture or belief system in which safety has become a sacred value, which means that people become unwilling to make trade-offs demanded by other practical and moral concerns. ‘Safety’ trumps everything else, no matter how unlikely or trivial the potential danger.” [52] Students who had been raised with safetyism on the playground sometimes expected it to govern their classrooms, dorms, and campus events.

You can see the all-encompassing play-crushing power of safetyism in figure 3.10, sent to me by a friend in Berkeley, California. The administrators at this elementary school don’t trust their students to play tag without adult guidance, because... what if there’s a dispute? What if someone is excluded?

The school offers similarly inane lists of instructions and prohibitions to help children play other games. In the rules for playing touch football, the sign says football can only be played if an adult is supervising and refereeing the game . The administrators seem to be committed to preventing the sorts of conflicts that are inherent in human interaction, and that would teach children how to manage their own affairs, resolve differences, and prepare for life in a democratic society.

Figure 3.10. Restrictions on free play, at an elementary school in Berkeley, California. [53]

American parents have lost so much trust in their fellow citizens and their own children that many now endorse the near-total elimination of freedom from childhood. According to a 2015 report from the Pew Research Center, parents (on average) say children should be at least 10 years old to play unsupervised in their own front yard . [54] They say that kids should be at least 12 years old before being allowed to stay alone in their own home unsupervised for one hour . They say that kids should be 14 before being allowed to go, unsupervised, to a public park . And these respondents include the same Gen X and baby boom parents who say, gleefully and gratefully, that they were let out, in a much more dangerous era, at ages 6, 7, or 8.

Antifragility and the Attachment System

Earlier in this chapter I described discover mode and defend mode as parts of a dynamic system for quickly adapting to changing conditions, like a thermostat. That system is embedded in a larger dynamic system called the attachment system. Mammals are defined by the evolutionary innovation that females bear their young live (not as eggs) and then produce milk to feed them. Mammal babies therefore have a long period of dependence and vulnerability during which they must achieve two goals: (1) develop competence in the skills needed for adulthood, and (2) don’t get eaten. The best way to avoid getting eaten is generally to stick close to Mom. But as mammals mature, their experience-expectant brains need to wire up by practicing skills such as running, fighting, and befriending. This is why young mammals are so motivated to move away from Mom to play, including risky play.

The psychological system that manages these competing needs is called the attachment system. It was first described by the British psychoanalyst John Bowlby, who had studied the effects of separating children from their parents during World War II. Figure 3.11 is an excellent illustration of the attachment system in action, from the psychologist Deirdre Fay.

Every child needs at least one adult who serves as a “secure base.” Usually it is the mother, but it can just as well be the father, grandparent, or nanny, or any adult who is reliably available for comfort and protection. If safety was the child’s only goal, he’d stay “on base” for all of childhood. There’d be no need for a complicated regulatory system. But as soon as children can crawl, they want to crawl over to things they can touch, suck on, or otherwise explore. They need to spend a lot of time in discover mode, because that’s where the learning and neural fine-tuning take place. But inevitably, something goes wrong. The child falls and bangs his head; a cat hisses at him; a stranger approaches. At that point defend mode activates and the child scurries back to base or starts crying, which is the child’s way of calling for the base to come to him.

The Attachment System

Figure 3.11. The mammalian attachment system. [55]

A securely attached child usually settles within a few seconds or minutes, shifts back to discover mode, and heads out for more learning. This process happens dozens of times a day, hundreds of times a month, and within a few years children become less fearful and more likely to want to explore on their own—perhaps by walking to school or a friend’s house with no help from an adult. [56] As the child develops she is able to internalize the secure base. She doesn’t need the parent’s physical presence to feel that she has support, so she learns to face adversity by herself.

In adolescence, young people begin seeking out romantic relationships. These new attachments will reuse the psychological architecture and “internal working models” that were developed while forming attachments to parents. Adolescents will reuse those models to attach to love interests and later, perhaps, a spouse. But children who are kept on home base, prevented from making those off-base excursions that are so helpful for developing their antifragile nature, don’t get to spend as much time in the growth zone. They may therefore spend more of their lives in defend mode, remaining more dependent on a parent’s physical presence, which reinforces parental overprotection in a vicious cycle.

I have sketched out how things work in theory. In practice, everything about raising children is messy, hard to control, and harder to predict. Children raised in loving homes that support autonomy, play, and growth may still develop anxiety disorders; children raised in overprotective homes usually turn out fine. There is no one right way to be a parent; there is no blueprint for building a perfect child. Yet it is helpful to bear in mind some general features of human childhood: Kids are antifragile and therefore they benefit from risky play, along with a secure base, which helps to shift them over toward discover mode. A play-based childhood is more likely to do that than a phone-based childhood.

In Sum

The human brain contains two subsystems that put it into two common modes: discover mode (for approaching opportunities) and defend mode (for defending against threats). Young people born after 1995 are more likely to be stuck in defend mode, compared to those born earlier. They are on permanent alert for threats, rather than being hungry for new experiences. They are anxious.

All children are by nature antifragile. Just as the immune system must be exposed to germs, and trees must be exposed to wind, children require exposure to setbacks, failures, shocks, and stumbles in order to develop strength and self-reliance. Overprotection interferes with this development and renders young people more likely to be fragile and fearful as adults.

Kids must have a great deal of free play to develop, and they benefit from risky physical play, which has anti-phobic effects. Kids seek out the level of risk and thrill that they are ready for, in order to master their fears and develop competencies. Risk-taking online may not have comparable anti-phobic effects.

In the 1980s and especially the 1990s, parents in Anglo countries became more fearful for many reasons, including changes in the media ecosystem and news cycle. They lost trust in each other, they started spending far more time supervising their own children, and they did more parenting in defend mode, seeing risks and threats everywhere.

The worship of “safety” above all else is called safetyism. It is dangerous because it makes it harder for children to learn to care for themselves and to deal with risk, conflict, and frustration.

The attachment system evolved to help young mammals learn the skills they’ll need to reach adulthood while retreating to their “secure base” when they feel threatened. Fearful parenting keeps children on home base too much, preventing them from having the experiences they need to grow strong and to develop a secure attachment style.

Children are most likely to thrive when they have a play-based childhood in the real world. They are less likely to thrive when fearful parenting and a phone-based childhood deprive them of opportunities for growth.