CHAPTER 8

Active Engagement

Take two kittens. Put a collar and leash on the first one. Place the second one in a harness. Finally, connect them to a merry-go-round apparatus which ensures that the movements of the two kittens are strictly linked. The idea is that the two animals receive identical visual inputs, but one is active while the other is passive. The former explores the environment on its own, while the latter moves in exactly the same way, but without control.

This is the classic carousel experiment that Richard Held (1922–2016) and Alan Hein performed in 1963—a time when the ethics of animal experimentation was clearly not as developed as it is today! This very simple experiment led to an important discovery: active exploration of the world is essential for the proper development of vision. Over a period of a few weeks, for three hours a day, the two kittens lived in a large cylinder lined with vertical bars. Although their visual inputs were very similar, they developed dramatically different visual systems.1 Despite the impoverished environment consisting only of vertical bars, the active kitten developed normal vision. The passive kitten, on the other hand, lost its visual abilities and, at the end of the experiment, failed basic visual exploration tests. In the cliff test, for example, the animal was placed on a bridge that it could leave either on the side of a high cliff or on the shallower side. A normal animal does not hesitate for a second and jumps to the easy side. The passive animal, however, chose at random. Other tests showed that the passive animal failed to develop a proper model of visual space and did not feel out its environment with its paws like normal cats do.

A PASSIVE ORGANISM DOES NOT LEARN

Held and Hein’s carousel experiment is the metaphor for our second pillar of learning: active engagement. Converging results from diverse fields suggest that a passive organism learns little or nothing. Efficient learning means refusing passivity, engaging, exploring, and actively generating hypotheses and testing them on the outside world.

To learn, our brain must first form a hypothetical mental model of the outside world, which it then projects onto its environment and puts to a test by comparing its predictions to what it receives from the senses. This algorithm implies an active, engaged, and attentive posture. Motivation is essential: we learn well only if we have a clear goal and we fully commit to reaching it.

Don’t get me wrong: active engagement does not mean that children should be encouraged to fidget in class all day long! I once visited a school where the principal told me, with a certain pride, how he applied my ideas: he had equipped his pupils’ desks with pedals so that his students could remain active during math class. . . . He had totally missed my point (and showed me the limits of the carousel experiment metaphor). Being active and engaged does not mean that the body must move. Active engagement takes place in our brains, not our feet. The brain learns efficiently only if it is attentive, focused, and active in generating mental models. To better digest new concepts, active students constantly rephrase them into words or thoughts of their own. Passive or, worse, distracted students will not benefit from any lesson, because their brains do not update their mental models of the world. This has nothing to do with actual motion. Two students could be very still yet differ dramatically in the inner movements of their thoughts: one actively follows the course, while the other disengages and becomes passive or distracted.

Experiments show that we rarely learn by merely accumulating sensory statistics in a passive manner. This can happen, but mainly at the lower levels of our sensory and motor systems. Remember those experiments where a child hears hundreds of syllables, computes the transition probabilities between syllables (such as /bo/ and /t^l/), and ends up detecting the presence of words (“bottle”)? This type of implicit learning seems to persist even when infants are asleep.2 However, it is the exception that proves the rule: in the vast majority of cases, and as soon as learning concerns high-level cognitive properties, such as the explicit memory of word meanings rather than their mere form, learning seems to occur only if the learner pays attention, thinks, anticipates, and puts forth hypotheses at the risk of making mistakes. Without attention, effort, and in-depth reflection, the lesson fades away, without leaving much of a trace in the brain.

DEEPER PROCESSING, BETTER LEARNING

Let’s take a classical example from cognitive psychology: the effect of word processing depth. Imagine that I present a list of sixty words to three groups of students. I ask the first group to decide whether the words’ letters are upper- or lowercase; the second group, whether the words rhyme with “chair”; and the third, whether they are animal names or not. Once the students are finished, I give them a memory test. Which group remembers the words best? Memory turns out to be much better in the third group, who processed the words in depth, at the meaning level (75 percent success), than in the other two groups, who processed the more superficial sensory aspects of the words, either at the letter level (33 percent success) or the rhyme level (52 percent success).3 We do find a weak implicit and unconscious trace of the words in all groups: learning leaves its subliminal mark within the spelling and phonological systems. However, only in-depth semantic processing guarantees explicit, detailed memory of the words. The same phenomenon occurs at the level of sentences: students who make the effort to understand sentences on their own, without teacher guidance, show much better retention of information.4 This is a general rule, which the American psychologist Henry Roediger states as follows: “Making learning conditions more difficult, thus requiring students to engage more cognitive effort, often leads to enhanced retention.”5

Brain imaging is beginning to clarify the origins of this processing depth effect.6 Deeper processing leaves a stronger mark in memory because it activates areas of the prefrontal cortex that are associated with conscious word processing and because these areas form powerful loops with the hippocampus, which stores information in the form of explicit episodic memories.

In the cult film La Jetée (1962), by French director Chris Marker (1921–2012), a voice-over states the following aphorism, which sounds like a profound truth: “Nothing distinguishes memories from ordinary moments: only later do they make themselves known, from their scars.” A beautiful adage . . . but a false proverb, because brain imaging shows that at the onset of memory encoding, the events of our life which will remain engraved in our memory can already be distinguished from those that will leave no trace: the former have been processed at a deeper level.7 By scanning a person while she is merely exposed to a list of words and images, we can predict which of those individual stimuli will be later forgotten and which will be retained. The key predictor is whether they induced activity in the frontal cortex, the hippocampus, and the neighboring regions of the parahippocampal cortex. The active engagement of these regions is a direct reflection of the depth to which these words and images traveled in the brain, and it predicts the strength of the trace that they leave in memory. An unconscious image enters sensory areas but creates only a modest wave of activity in the prefrontal cortex. Attention, concentration, processing depth, and conscious awareness transform this small wave into a neuronal tsunami that invades the prefrontal cortex and maximizes subsequent memorization.8

The role of active engagement and processing depth is confirmed by converging evidence from pedagogical studies in a school context—for example, learning physics at the undergraduate level. Students must learn the abstract concepts of angular momentum and motor torque. We divide the students into two groups: one group is given ten minutes to experiment with a bicycle wheel, and the other group, ten minutes of verbal explanation and observation of other students. The result is clear: learning is much better in the group that benefited from active interaction with the physical object.9 Making a course deeper and more engaging facilitates the subsequent retention of information.

This conclusion receives support from a recent review of more than two hundred pedagogical studies in undergraduate STEM courses: traditional lecturing, where students remain passive while the teacher preaches for fifty minutes, is inefficient.10 Compared with teaching methods that promote active engagement, lecturing systematically yields lower performances. In all disciplines, from math to psychology, biology to computer science, an active student succeeds more. With active engagement, examination scores progress by half a standard deviation, which is considerable, and the failure rate decreases by over 10 percent. But what are the strategies that engage students the most? There is no single miraculous method, but rather a whole range of approaches that force students to think for themselves, such as practical activities, discussions in which everyone takes part, small group work, or teachers who interrupt their class to ask a difficult question and let the students think about it for a while. All solutions that force students to give up the comfort of passivity are effective.

THE FAILURE OF DISCOVERY-BASED TEACHING

None of this is new, you may be thinking, and many teachers already apply these ideas. However, in the pedagogical domain, neither tradition nor intuition can be trusted: we need to scientifically verify which pedagogies actually improve students’ comprehension and retention, and which do not. And this is an opportunity for me to clarify a very important distinction. The fundamentally correct view that children must be attentively and actively engaged in their own learning must not be confused with classical constructivism or discovery learning methods—which are seductive ideas whose ineffectiveness has, unfortunately, been repeatedly demonstrated.11 This is a key distinction, but it is rarely understood, in part because the latter pedagogies are also known as active pedagogies, which is a great source of confusion.

When we talk about discovery learning, what do we mean? This nebula of pedagogical views can be traced back to Jean-Jacques Rousseau and has reached us through famous educators such as John Dewey (1859–1952), Ovide Decroly (1871–1932), Célestin Freinet (1896–1966), Maria Montessori, and, more recently, Jean Piaget and Seymour Papert (1928–2016). “Do I dare set forth here,” writes Rousseau in Emile, or On Education, “the most important, the most useful rule of all education? It is not to save time, but to squander it.” For Rousseau and his successors, it is always better to let children discover for themselves and build their own knowledge, even if it implies that they might waste hours tinkering and exploring. . . . This time is never lost, Rousseau believed, because it eventually yields autonomous minds, capable not only of thinking for themselves but also of solving real problems, rather than passively receiving knowledge and spitting out rote and ready-made solutions. “Teach your student to observe the phenomena of nature,” says Rousseau, “and you will soon rouse his curiosity; but if you want his curiosity to grow, do not be in too great a hurry to satisfy it. Lay the problems before him and let him solve them himself.”

The theory is attractive. . . . Unfortunately, multiple studies, spread over several decades, demonstrate that its pedagogical value is close to zero—and this finding has been replicated so often that one researcher entitled his review paper “Should There Be a Three-Strikes Rule against Pure Discovery Learning?” When children are left to themselves, they have great difficulty discovering the abstract rules that govern a domain, and they learn much less, if anything at all. Should we be surprised by this? How could we imagine that children would rediscover, in a few hours and without any external guidance, what humanity took centuries to discern? At any rate, the failures are resounding in all areas:

I directly experienced the birth of the personal home computer—I was fifteen years old when my father bought us a Tandy TRS-80 with sixteen kilobytes of memory and 48-by-128-pixel graphics. Like others of my generation, I learned to code in the programming language BASIC without a teacher or a class—although I was not alone: my brother and I devoured all the magazines, books, and examples we could get our hands on. I eventually became a reasonably effective programmer . . . but when I entered a master’s program in computer science, I became aware of the enormity of my shortcomings: I had been tinkering all this time without understanding the deep, logical structure of programs, nor the proper practices that made them clear and legible. And this is perhaps the worst effect of discovery learning: it leaves students under the illusion that they have mastered a certain topic, without ever giving them the means to access the deeper concepts of a discipline.

In summary, while it is crucial for students to be motivated, active, and engaged, this does not mean that they should be left to their own devices. The failure of constructivism shows that explicit pedagogical guidance is essential. Teachers must provide their students with a structured learning environment designed to progressively guide them to the top as quickly as possible. The most efficient teaching strategies are those that induce students to be actively engaged while providing them with a thoughtful pedagogical progression that is closely channeled by the teacher. In the words of psychologist Richard Mayer, who reviewed this field, the best success is achieved by “methods of instruction that involve cognitive activity rather than behavioral activity, instructional guidance rather than pure discovery, and curricular focus rather than unstructured exploration.”13 Successful teachers provide a clear and rigorous sequence that begins with the basics. They constantly assess their students’ mastery and let them build a pyramid of meaning.

And this is indeed what most schools inspired by Montessori do today: they do not let children “marinate” without doing anything; instead, they propose a whole series of rational and hierarchical activities, whose purpose is first carefully demonstrated by teachers before being carried out independently by children. Active engagement, pleasure, and autonomy, under the guidance of an explicit teaching method and with stimulating pedagogical materials: these are the ingredients for a winning recipe whose effectiveness has been repeatedly demonstrated.

Pure discovery learning, the idea that children can teach themselves, is one of many educational myths that have been debunked but still remain curiously popular. It belongs to a collection of urban legends that mar the educational field, and at least two other major misconceptions are linked to it:14

What about all the special education books and software that claim to tailor education to each child’s needs? Are they worthless? Not necessarily. Children do vary dramatically, not in learning style, but in the speed, ease, and motivation with which they learn. In first grade, for instance, the top 10 percent of children already read more than four million words per year, whereas the bottom 10 percent read less than sixty thousand16—and dyslexic children may not read at all. Developmental deficits such as dyslexia and dyscalculia may come in several varieties, and it is often useful to carefully diagnose the exact nature of the impairment in order to adapt the lessons. Children do benefit from pedagogical interventions whose contents are tailored to their specific difficulties. For instance, many children, even in advanced mathematics, fail to understand how fractions work—in this case, the teacher should shed the current curriculum and return to the basics of numbers and arithmetic. However, every teacher should also keep in mind that all children learn using the same basic machinery—one that prefers focused attention to dual tasking, active engagement to passive lecturing, detailed error correction to phony praise, and explicit teaching over constructivism or discovery learning.

CURIOSITY AND HOW TO PIQUE IT

All men by nature desire to know.

Aristotle, Metaphysics (c. 335 BCE)

I have no special talent. I am only passionately curious.

Albert Einstein (1952)

One of the foundations of active engagement is curiosity—the desire to learn, or the thirst for knowledge. Piquing children’s curiosity is half the battle. Once their attention is mobilized and their mind in search of an explanation, all that is left to do is guide them. Starting in kindergarten, the most curious students are also those who do better in reading and math.17 Keeping children curious is therefore one of the key factors for successful education. But what exactly is curiosity? To what Darwinian necessity does it respond, and to what kind of algorithm does it correspond?

Rousseau wrote in Emile, or On Education, “One is curious only to the extent that one is educated.” Here again, he was wrong: curiosity is not an effect of instruction, a function that we must acquire. It is already present at an early age and is an integral part of our human brain circuitry, a key ingredient of our learning algorithm. We do not simply passively wait for new information to reach us—as do, foolishly, most current artificial neural networks, which are simple input-output functions passively submitted to their environment. As Aristotle noted, we humans are born with a passion to know, and we constantly seek novelty, actively exploring our environment to discover things we can learn.

Curiosity is a fundamental drive of the organism: a propulsive force that pushes us to act, just like hunger, thirst, the need for security, or the desire to reproduce. What role does it play in survival? It is in the interest of most animal species (mammals, but also many birds and fish) to explore their environment in order to better monitor it. It would be risky to set up a nest, lair, burrow, den, hole, or home without checking the surroundings. In an unstable universe populated by predators, curiosity can make all the difference between life and death—and this is why most animals regularly pay security visits to their territory, carefully checking for anything unusual and investigating novel sounds or sights. . . . Curiosity is the determination that pushes animals out of their comfort zones in order to acquire knowledge. In an uncertain world, the value of information is high and must ultimately be paid in Darwin’s own currency: survival.

Curiosity is therefore a force that encourages us to explore. From this perspective, it resembles the drive for food or sexual partners, except that it is motivated by an intangible value: the acquisition of information. Indeed, neurobiological studies show that, in our brains, the discovery of previously unknown information brings its own reward: it activates the dopamine circuit. Remember, this is the circuit that fires in response to food, drugs, and sex. In primates, and probably in all mammals, this circuit responds not solely to material rewards, but also to new information. Some dopaminergic neurons signal a future information gain, as if the anticipation of novel information brings its own gratification.18 Thanks to this mechanism, rats can be conditioned not only to food or drugs, but also to novelty: they quickly develop a preference for places that contain new objects and thereby satisfy their curiosity, as opposed to dull places where nothing ever happens.19 We do not act any differently when we move to a big city for a change of scenery or when, eager for the latest gossip, we frantically scroll through Facebook or Twitter.

Humans’ appetite for knowledge passes through the dopamine circuit even when it involves a purely intellectual curiosity. Imagine lying in an MRI and being asked Trivial Pursuit questions, such as, “Who was the president of the United States when Uncle Sam first got his beard?”20 For each question, before satisfying your curiosity, the experimenter asks how eager you are to know the answer. What are the neuronal correlates of this subjective feeling of being curious? The degree of curiosity that you report correlates tightly with the activity of the nucleus accumbens and the ventral tegmental area, two essential regions of the dopamine brain circuit. The more curious you are, the more these regions light up. Their signals arise in anticipation of the answer: even before your curiosity is satisfied, the simple fact of knowing that you will soon know the answer excites your dopaminergic circuits. Expectation of a positive event brings its own reward.

These curiosity signals are obviously useful, because they predict how much you learn. Memory and curiosity are linked—the more curious you are about something, the more likely you are to remember it. Curiosity even transfers to nearby events: when your curiosity is heightened, you remember incidental details such as the face of a passerby or the person who taught you the information that you were so eager to learn. The degree of craving for knowledge controls the strength of memory.

Through the dopamine circuit, the satisfaction of our appetite to learn—or even the mere anticipation of that satisfaction—is deeply rewarding. Learning possesses intrinsic value for the nervous system. What we call curiosity is nothing more than the exploitation of this value. As such, our species is probably special because of its unmatched ability to learn. As hominization progressed, our ability to represent the world progressed. We are the only animals who formulate formal theories of the world in a language of thought. Science has become our ecological niche: Homo sapiens is the only species without a specific habitat, because we learn to adapt to any environment.

Mirroring the extraordinary expansion of our learning abilities, human curiosity seems to have increased tenfold. Over the course of our evolution, we have acquired an extended form of curiosity, called “epistemic curiosity”: the pure desire for knowledge in all fields, including the most abstract. Like other mammals, we play and explore—not only through real movement, but also through thought experiments. Whereas other animals visit the space around them, we explore conceptual worlds. Our species also experiences specific epistemic emotions that guide our thirst for knowledge. We rejoice, for example, in the symmetry and pure beauty of mathematical patterns: a clever theorem can move us much more than a piece of chocolate.

Mirth seems to be one of those uniquely human emotions that guide learning. Our brain triggers a mirth reaction when we suddenly discover that one of our implicit assumptions is wrong, forcing us to drastically revise our mental model. According to the philosopher Dan Dennett, hilarity is a contagious social response that spreads as we draw each other’s attention to an unexpected piece of information.21 And, indeed, all things being equal, laughing during learning seems to increase curiosity and enhance subsequent memory.22

WANTING TO KNOW: THE SOURCE OF MOTIVATION

Several psychologists have tried to specify the algorithm that underlies human curiosity. Indeed, if we understood it better, we could perhaps gain control over this essential ingredient of our learning scheme, and even reproduce it in a machine that would eventually imitate the performance of the human species: a curious robot.

This algorithmic approach is beginning to bear fruit. The greatest psychologists, from William James to Jean Piaget to Donald Hebb, have speculated on the nature of the mental operations that underlie curiosity. According to them, curiosity is the direct manifestation of children’s motivation to understand the world and build a model of it.23 Curiosity occurs whenever our brains detect a gap between what we already know and what we would like to know—a potential learning area. At any given moment, we choose, from the various actions that are accessible to us, those that are most likely to reduce this knowledge gap and acquire useful information. According to this theory, curiosity resembles a cybernetic system that regulates learning, similar to the famous Watt governor, which opens or closes the throttle valve on a steam engine in order to regulate steam pressure and maintain a fixed speed. Curiosity would be the brain’s governor, a regulator that seeks to maintain a certain learning pressure. Curiosity guides us to what we think we can learn. Its opposite, boredom, turns us away from what we already know, or from areas that, according to our past experience, are unlikely to have anything left to teach us.

This theory explains why curiosity is not directly related to the degree of surprise or novelty but instead follows a bell curve.24 We have no curiosity for the unsurprising—things that we have seen a thousand times before are boring. But we are also not attracted to things that are too novel or surprising, or so confusing that their structure eludes us—their very complexity deters us. Between the boredom of the too simple and the repulsion of the too complex, our curiosity naturally directs us toward new and accessible fields. But this attraction keeps changing. As we master them, the objects that once seemed attractive lose their appeal, and we redirect our curiosity toward new challenges. This is why babies initially seem so passionate about the most trivial things: grasping their toes, closing their eyes, playing peekaboo. . . . Everything is new to them and is a potential source of learning. Once they squeeze out all the knowledge that can be gained from those experiments, they lose interest—for exactly the same reason that no scientist reproduces Galileo’s experiments anymore: what is known becomes boring.

The same algorithm also explains why we sometimes turn away from an area that once seemed attractive but proved to be too difficult. Our brain evaluates the speed of learning, and curiosity is turned off if our brain detects that we are not progressing fast enough. We all know of children who, say, return from a concert with a passion for the violin . . . only to give it up after a few weeks, when they realize that mastery of the instrument does not come easily. Those who keep playing either set more modest goals (e.g., play a little better every day) or, if they truly aim to become professional musicians, sustain their motivation through parental and social support and constant reminders of their long-term goals.

Two French engineers, Frédéric Kaplan and Pierre-Yves Oudeyer, have implemented curiosity in a robot.25 Their algorithm includes several modules. The first is a classic artificial learning system that constantly tries to predict the state of the outside world. The second, more innovative module evaluates the performance of the first: it measures the recent learning speed and uses it to predict the areas in which the robot will learn the most. The third ingredient is a reward circuit that places greater value on actions that are predicted to lead to more efficient learning. As a result, the system naturally focuses on those areas where it believes that it will learn the most, which is the very definition of curiosity, according to Kaplan and Oudeyer.

When their curious robot, equipped with this algorithm, is placed on a baby mat, it behaves exactly like a young child. For a few minutes, it becomes enthused about a particular object and spends all its time, for example, repeatedly lifting a stuffed elephant ear. As it progressively learns all there is to know about an item, its curiosity dwindles. At one point, it turns away and actively seeks another source of stimulation. After an hour, it stops exploring the mat: a digital form of boredom sets in as the robot comes to believe that everything that could be learned is now known.

The analogy with a small child is striking. Even babies a few months old orient toward stimuli of intermediate complexity, neither too simple nor too complex, but whose structure is just right to be quickly learnable. (This trait of infants’ curiosity has been described as the “Goldilocks effect.”26) To maximize what they learn, we have to constantly enrich their environment with new objects that are just stimulating enough to not be discouraging. It is adults’ responsibility to provide them with a well-designed pedagogical hierarchy that progressively takes them to the top, constantly stimulating their drive for knowledge and novelty.

Curiosity is an essential ingredient of our learning algorithm, which is only beginning to be reproduced in machines. Here, a small robot explores a play mat. Curiosity is implemented by a reward function that favors the choice of whichever action maximizes the potential to learn. As a consequence, the robot successively tries out each toy on the mat and each action at its disposal. Once it masters one aspect of the world, it loses interest and redirects its attention elsewhere.

This vision of curiosity leads to an interesting prediction. It implies that in order for children to be curious, they must be aware of what they do not yet know. In other words, they must possess metacognitive faculties at an early age. “Metacognition” is cognition over cognition: the set of higher-order cognitive systems that monitor our mental processes. According to the gap theory of curiosity, metacognitive systems must constantly supervise our learning, evaluating what we know and don’t know, whether we are wrong or not, whether we are fast or slow, and so on and so forth—metacognition encompasses everything we know about our own minds.

Metacognition plays a key role in curiosity. Indeed, to be curious is to want to know, and that implies knowing what you don’t already know. And once again, recent experiments confirm that from the age of one and perhaps even earlier, children understand that there are things they do not know.27 Indeed, babies of that age readily turn to their caregiver whenever they are unable to solve a problem alone. Knowing that they don’t know leads them to ask for more information. This is the early manifestation of epistemic curiosity: the irresistible desire to know.

THREE WAYS SCHOOL CAN KILL CURIOSITY

All parents are nostalgic for the days when their toddler was filled with curiosity. Between ages two and five, children are curious about everything. Their favorite word is often why: they never stop experimenting on the world and questioning adults in order to quench their thirst for knowledge. Surprisingly, however, this appetite, which seems insatiable, eventually dies out, often after a few years of school. Some children remain curious about everything, but many close themselves off to such intrigue. Their active engagement turns into dull passivity. Can the science of curiosity explain why? We do not yet have all the answers, but I would like to propose a few hypotheses.

First, children may lose their curiosity because they lack cognitive stimulation tailored to their needs. According to the algorithm we have just described, it is entirely normal for curiosity to dwindle over time. As learning progresses, the expected learning gain shrinks: the better we master a field, the more we reach the limits of what it can offer, and the less interested we are in it. To maintain curiosity, schools must therefore continually provide children’s supercomputing brains with stimulants that match their intelligence. This is not always the case. In a standard classroom, the most advanced students often lack stimulation: after a few months, their curiosity fades and they no longer expect much from school, because their metacognitive system has learned that, unfortunately, they are unlikely to learn much more.

At the other end of the spectrum, students who struggle in school may wither away for the opposite reason. Metacognition remains the main culprit: after a while, they no longer have any reason to be curious, because they have learned . . . that they do not succeed in learning. Their past experience has engraved a simple (though false) rule in the depths of their metacognitive circuits: I am incapable of learning such and such topic (math, reading, history, whatever). Such dismay is not uncommon: many girls convince themselves that mathematics is not for them,28 and children from underprivileged neighborhoods sometimes come to believe that school is hostile for them and teaches nothing useful for their future. Such metacognitive judgments are disastrous because they demotivate students and nip their curiosity in the bud.

The solution is to boost these children’s confidence back up, step by step, by showing them that they are perfectly capable of learning, provided the problems are adapted to their level, and that learning brings its own reward. The theory of curiosity says that when children are discouraged, whether they are far ahead or far behind at school, what matters most is to restore their desire to learn by offering them stimulating problems carefully tailored to their current level. First, they rediscover the pleasure of learning something new—and then, slowly, their metacognitive system learns that they can learn, which puts their curiosity back on track.

Another scenario that can lead to children losing interest is when curiosity is punished. A child’s appetite for discovery can be ruined by an overly rigid pedagogical strategy. Teaching through traditional lectures tends to discourage children from participating or even from thinking. It can convince children that they are simply being asked to sit there and remain quiet until the end of class. The neurophysiological interpretation of this situation is simple: within the dopamine circuit, the reward signals induced by curiosity and its satisfaction compete with external rewards and punishments. It is therefore possible to discourage curiosity by punishing each exploration attempt. Picture a child who repeatedly tries to participate and is systematically reprimanded, mocked, or punished: “Silly question. You’d better be quiet or you’ll stay an extra half an hour after school. . . .” This child quickly learns to inhibit their curiosity drive and stop participating in class: the curiosity-based reward that the dopamine system expects—the pleasure of learning something new—is largely countered by the direct negative signals that the same circuit receives. Repeated punishment leads to learned helplessness, a kind of physical and mental paralysis associated with stress and anxiety, which has been shown to inhibit learning in animals.29

The solution? Most teachers already know it. It is simply a matter of rewarding curiosity instead of punishing it: encouraging questions (however imperfect they may be), asking children to give presentations on subjects they love, rewarding them for taking initiative. . . . The neuroscience of motivation is extremely clear: the desire to do action X must be associated with an expected reward, be it material (food, comfort, social support) or cognitive (acquisition of information). Too many children lose all curiosity because they learn, at their own expense, to expect no reward from school. (Grades, which I will get to shortly, often contribute to this sad state of affairs.)

The third factor that can discourage curiosity is the social transmission of knowledge. Remember that two modes of learning coexist in the human species: the active mode, where children constantly experiment and question themselves like good budding scientists, and the receptive mode, where they simply record what others teach them. School often encourages only the second mode—and it may even discourage the first, if children assume that teachers always know everything better than students do.

Can a teacher’s attitude really kill a child’s natural curiosity?30 Sadly, recent experiments suggest that the answer is yes. In her childhood cognition lab at MIT, the American developmental psychologist Laura Schulz presents kindergartners with a strange contraption: a set of plastic tubes hidden in various places that contain all sorts of unexpected toys, such as a mirror, a horn, a game with lights, and a music box. When you give such a gadget to children without saying anything, you immediately set off their curiosity: they explore, rummage, forage, and poke around until they find most of the hidden rewards. Now, take a new group of kindergartners and put them into the passive, receptive pedagogical mode. All you have to do is give them the object while saying, “Look, let me show you my toy. This is what it does . . .” and then play the music box, for instance. One might think that this would stimulate the children’s curiosity . . . but it has the opposite effect: exploration massively decreases following this kind of introduction. Children seem to make the (often correct) assumption that the teacher is trying to help them as much as possible, and that he has therefore introduced them to all the interesting functions of the device. In this context, there is no need to search: curiosity is inhibited.

Further experiments show that children take into account the teacher’s past behavior. When a teacher always makes exhaustive demonstrations, students lose curiosity. If the teacher demonstrates one of the functions of a new toy, children do not explore all its facets, because they think that the teacher has already explained everything there is to know. If, on the contrary, the teacher gives evidence that he doesn’t always know everything, then the children keep searching.

So, what is the right approach? I suggest always keeping the concept of active engagement in mind. Maximally engaging a child’s intelligence means constantly feeding them with questions and remarks that stimulate their imagination and make them want to go deeper. It would be out of the question to let students discover everything for themselves—this would be falling back into the trap of discovery-based learning. The ideal scenario is to offer the guidance of a structured pedagogy while encouraging children’s creativity by letting them know that there are still a thousand things to discover. I remember a teacher who, just before summer vacation, told me, “You know, I just read a little math problem I couldn’t solve. . . .” And this is how I found myself ruminating on this question all summer, trying to do better than the teacher could. . . .

Mustering children’s active engagement goes hand in hand with another necessity: tolerating their errors while quickly correcting them. This is our third pillar of learning.