CHAPTER 6

Recycle Your Brain

Let’s summarize what we have covered so far. The blank-slate assumption is clearly wrong: babies are born with considerable core knowledge, a rich set of universal assumptions about the environment that they will later encounter. Their brain circuits are well organized at birth and give them strong intuitions in all sorts of domains: objects, people, time, space, numbers. . . . Their statistical skills are remarkable—they already act as budding scientists, and their sophisticated learning abilities allow them to progressively converge onto the most appropriate models of the world.

At birth, all the large fiber bundles of the brain are already in place. Brain plasticity can, however, reorganize their terminal connections. Millions of synapses undergo plastic changes every time we acquire new knowledge. Enriching children’s environments, for instance, by sending them to school, can deeply enhance their brains and augment them with skills that they will keep throughout their lives. This plasticity is not unconstrained, however. It is restricted in space (on the order of a few millimeters), as well as in time—many circuits begin to close off after a few months or years.

In this chapter, I look at the role that formal education plays in early brain development. Education, indeed, raises a paradox: Why is it that Homo sapiens can take chalk or a keyboard and start writing or making calculations? How could the human species expand its capabilities in novel directions that previously played no part in its genetic evolution? That the human primate manages to learn to read or to calculate should never cease to amaze us. As Vladimir Nabokov (1899–1977) put it so well, “We are absurdly accustomed to the miracle of a few written signs being able to contain immortal imagery, involutions of thought, new worlds with live people, speaking, weeping, laughing. What if we awake one day, all of us, and find ourselves utterly unable to read?”1

I have studied at length the minds and brains of illiterate adults, whether it be in Portugal, Brazil, or the Amazon—people who never had the chance to go to school, simply because their families could not afford it or because there were no schools nearby. Their skills are, in certain ways, profoundly different:2 not only are they incapable of recognizing letters, but they also have difficulties recognizing shapes and distinguishing mirror images,3 paying attention to a part of a face,4 and memorizing and distinguishing spoken words.5 So much for Plato, who naively believed that learning to read would ruin our internal memory by forcing us to rely on the external memory of books. Nothing could be further from the truth. The myth of the illiterate bard who effortlessly musters immense powers of memory is just that: a myth. We all need to exercise our memory—and it gets better, not worse, by having gone to school and learned to read.

The impact of education is even more striking in mathematics.6 We discovered this by studying the many Amazon Indians who never had the chance to go to school. First of all, many of them do not know how to precisely count a collection of items. Many of their languages do not even include a counting system—they either have just a handful of words for “few” versus “many” (like the Pirahã), or just fuzzy words for the numbers one to five (like the Munduruku), and if they learn to count at all, for instance, using Spanish or Portuguese number words, it is with a huge delay (like the Tsimane) compared to Western children.7 Second, they possess only the rudiments of mathematical intuition: they distinguish basic geometrical shapes, understand the organization of space, can navigate in a straight line, perceive the differences between quantities such as thirty and fifty, and know that numbers can be ordered from left to right. We inherit these skills from our evolution and share them with other animals as diverse as ravens, macaque monkeys, and freshly hatched chicks. However, education vastly increases these initial skills. For instance, uneducated Amazon Indians do not seem to understand that there is the same interval of +1 between any two consecutive numbers. Education massively overturns our sense of the number line: as we learn to count and to perform exact arithmetic, we discover that every number n has a successor n + 1. Eventually, we understand that all consecutive numbers are equidistant and form a linear scale—whereas very young children and unschooled adults consider this line to be compressed, since large numbers seem to be closer to each other than small ones.8 If we had only an approximate sense of numbers like other animals do, we would be unable to distinguish eleven from twelve. We owe the refined precision of our number sense to education—and on this symbolic foundation rests the whole field of mathematics.

THE NEURONAL RECYCLING HYPOTHESIS

How does education revolutionize our mental skills, transforming us into primate readers of Nabokov, Steinbeck, Einstein, or Grothendieck? As we have seen, all that we learn passes through the modification of pre-established brain circuits, which are largely organized at birth but remain capable of changing on the scale of a few millimeters. Thus, all the diversity of human culture must fit within the constraints imposed by our neuronal nature.

To resolve this paradox, I have formulated the neuronal recycling hypothesis.9 The idea is simple: while synaptic plasticity makes the brain malleable—especially in humans, where childhood lasts for fifteen or twenty years—our brain circuits remain subject to strong anatomical constraints, inherited from our evolution. Therefore, each new cultural object we invent, such as the alphabet or Arabic numerals, must find its “neuronal niche” in the brain: a set of circuits whose initial function is sufficiently similar to its new cultural role, but also flexible enough to be converted to this new use. Any cultural learning must rely on the repurposing of a preexisting neural architecture, whose properties it recycles. Education must therefore fit within the inherent limits of our neural circuits, by taking advantage of their diversity, as well as of the extended period of neural plasticity which is characteristic of our species.

According to this hypothesis, to educate oneself is to recycle one’s existing brain circuits. Over the millennia, we have learned to make something new out of something old. Everything we learn at school reorients a preexisting neural circuit in a new direction. To read or calculate, children repurpose existing circuits that originally evolved for another use, but which, due to their plasticity, manage to adapt to a new cultural function.

Why did I coin this strange term, “neuronal recycling”? Because the corresponding French word, recyclage, perfectly combines two ideas that characterize what happens in our brain—a reuse of some material with unique properties, and also a reorientation toward a new career:

With neuronal recycling, I wanted to distinguish the fast learning of a new cultural skill from the many other situations where biology, in the course of a slow evolutionary process, makes something new with something old. Indeed, in the Darwinian process of evolution by natural selection, repurposing of older materials is common. Genetic recombination can spruce up ancient organs and turn them into elegant, innovative machines. Bird feathers? Old thermal regulators converted into aerodynamic flaps. Reptilian and mammalian legs? Antediluvian fins. Evolution is a great tinkerer, says the French Nobel Prize–winning biologist François Jacob (1920–2013): in its workshop, lungs convert into floating organs, an old piece of the reptilian jaw becomes the inner ear, and even the sneer of hungry carnivores turns into Mona Lisa’s delicate smile.

The brain is no exception. Language circuits, for instance, may have appeared during hominization through the duplication and repurposing of previously established cortical circuits.10 But such slow genetic modifications do not fall under my definition of neuronal recycling. The appropriate term is “exaptation,” a neologism coined by Harvard evolutionist Stephen Jay Gould (1941–2002) and Yale paleontologist Elisabeth Vrba and based on the word “adaptation.” An old mechanism is exapted when it acquires a different use in the course of Darwinian evolution. (A simple mnemonic may help: exaptation makes your ex apt to a new task!) Because it is based on the spreading of genes through a population; at the species level, exaptation acts over tens of thousands of years. Neuronal recycling, on the other hand, occurs within an individual brain and on a much shorter time frame, anywhere from days to years. Recycling a brain circuit means reorienting its function without genetic modification, merely through learning and education.

My intent in formulating the neuronal recycling hypothesis was to explain the particular talent of our species for going beyond its usual ecological niche. Humans, indeed, are unique in their ability to acquire new skills, such as reading, writing, counting, doing math, singing, dressing, riding a horse, and driving a car. Our extended brain plasticity, combined with novel symbolic learning algorithms, has given us a remarkable ability for adaptation—and our societies have discovered means of further amplifying our skills by subjecting children, day after day, to the powerful regime of school.

To emphasize the singularity of the human species is not to deny, of course, that neuronal recycling, on a smaller scale, also exists in other animals. Recent technologies have made it possible to record the activity of the same hundred neurons for several weeks, while monkeys acquire a new skill—and, thus, to put the recycling view to a strong test. These experiments were able to address a simple but profound prediction of the theory: Can learning ever radically change the neural code in a given brain circuit, or, as the recycling view would predict, does learning solely repurpose the circuit?

In a recent experiment, using a brain-computer interface, researchers asked a monkey to learn to control its own brain. They taught the animal that to make a cursor go right, it had to activate ten specific neurons; and to make the cursor go up, it should activate ten other cells; and so on.11 Remarkably, this procedure worked: in a few weeks, the animal learned to bend the activity of ten arbitrarily chosen neurons in order to make the cursor move at will. However—and this is the key—the monkey was able to get the cursor to move only if the neuronal activity that it was asked to produce did not deviate too much from what its cortex was already spontaneously producing before training. In other words, what the monkey was asked to learn had to fit within the repertoire of the neuronal circuit that it was asked to retrain.

To appreciate what the researchers showed, it is important to realize that the dynamics of brain circuits are constrained. The brain does not explore every configuration of activity that it might be able to access. In theory, in a group of a hundred neurons, activity could span a hundred-dimension space, yielding an unfathomable number of states (if we consider that each neuron could be on or off, this number exceeds 2100, or over a thousand billion billion billion). Yet, in reality, brain activity visits only a fraction of this humongous universe, typically restricted to around ten dimensions. With this idea in mind, the constraint on learning can be formulated succinctly: a monkey can learn a novel task only if what we ask of its cortex “fits” within this preexisting space. If, on the other hand, we ask the monkey to activate a combination of neurons that is never observed in prior activity, it fails dramatically.

Note that the learned behavior itself may be radically new—who could have foreseen that a primate would one day control a cursor on a computer screen? However, the neuronal states that make this behavior possible must fit within the space of available cortical activity patterns. This result directly validates a key prediction of the neuronal recycling hypothesis—the acquisition of a novel skill does not require a radical rewriting of cortical circuits as if they were a blank slate, but merely a repurposing of their existing organization.

It is becoming increasingly clear that each region of the brain imposes its own set of constraints on learning. In a region of the parietal cortex, neural activity is generally confined to a single dimension, a straight line in high-dimensional space.12 These parietal neurons encode all incoming data on an axis ranging from small to large—they are therefore ideally suited to encode quantities and their relative sizes. Their neural dynamics may seem extraordinarily limited, but what seems like a handicap could actually be an advantage when it comes to representing quantities, such as size, number, area, or any other parameter that can be ordered from small to large. In a sense, this part of the cortex may be pre-wired to encode quantities—indeed, it is systematically recruited as soon as we manipulate quantities along a linear axis, from numbers to social status (who is “above” whom on the social ladder).13

For another example, consider the entorhinal cortex, a region of the temporal cortex that contains the famous grid cells that map out space (which I described in Chapter 4). In this region, the neural code is two-dimensional: even if there are millions of neurons in this part of the brain, their activity cannot help but remain confined to a plane, or, technically, a two-dimensional manifold in high-dimensional space.14 Again, this property, far from being a drawback, is obviously perfectly suited to form a map of the environment, as seen from above—and in fact, we know that this region hosts the mental GPS by which a rat locates itself in space. Remarkably, recent work has shown that this same region also lights up as soon as we have to learn to represent any data on a two-dimensional map, even if these data are not directly spatial.15 In one experiment, for instance, birds could vary in two dimensions: the length of their neck, and the length of their legs. Once the human participants had learned to represent this unusual “bird space,” they used their entorhinal cortex, along with a few other areas, to navigate it mentally.

And the list could go on and on: the ventral visual cortex excels at representing visual lines and shapes, Broca’s area codes for syntactic trees,16 and so forth. Each region has its own preferred dynamics to which it remains faithful. Each projects its own space of hypotheses onto the world: one tries to fit the incoming data on a straight line, another tries to display them on a map, a third on a tree. . . . These hypotheses spaces precede learning and, in a certain way, make it possible. We can, of course, learn new facts, but they need to find their neuronal niche, a representation space adapted to their natural organization.

Let us now see how this idea applies to the most fundamental areas of school learning: arithmetic and reading.

MATHEMATICS RECYCLES THE CIRCUITS FOR APPROXIMATE NUMBER

Let us first take the example of mathematics. As I explained in my book The Number Sense,17 there is now considerable evidence to show that math education (like so many other aspects of learning) does not get imprinted onto the brain like a stamp on melted wax. On the contrary, mathematics molds itself into a preexisting, innate representation of numerical quantities, which it then extends and refines.

In both humans and monkeys, the parietal and prefrontal lobes contain a neural circuit that represents numbers in an approximate manner. Before any formal education, this circuit already includes neurons sensitive to the approximate number of objects in a concrete set.18 What does learning do? In animals trained to compare quantities, the amount of number-detecting neurons grows in the frontal lobe.19 Most important, when they learn to rely on the numerical symbols of Arabic digits, rather than on the mere perception of approximate sets, a fraction of these neurons become selective to such digits.20 This (partial) transformation of a circuit in order to incorporate the cultural invention of numerical symbols is a great example of neuronal recycling.

In humans, when we learn to perform basic arithmetic (addition and subtraction), we continue to recycle that region, but also the nearby circuitry of the posterior parietal lobe. That region is used to shift our gaze and our attention—and it seems that we reuse those skills to move in number space: adding activates the same circuits that move your attention to the right, in the direction of larger numbers, while subtracting excites circuits that shift your attention to the left.21 We all possess a kind of number line in our heads, a mental map of the number axis on which we have learned to accurately move when we perform calculations.

Recently, my research team has provided a more stringent test of the recycling hypothesis. With Marie Amalric, a young mathematician turned cognitive scientist, we wondered whether the same circuits of the parietal lobe continue to be used to represent the most abstract concepts in mathematics.22 We recruited fifteen professional mathematicians and scanned their brains with functional MRI while we presented them with abstruse mathematical expressions that only they could understand, including formulas like ∫s ∇ × F • dS and statements such as “Any square matrix is equivalent to a permutation matrix.” As we predicted, these high-level mathematical objects activated the very same brain network that activates when a baby sees one, two, or three objects,23 or when a child learns to count (see figure 12 in the color insert).24 All mathematical objects, from Grothendieck’s topoi to complex manifolds, or functional spaces, find their ultimate roots in the recombination of elementary neural circuits present during childhood. All of us, at any stage of the cultural construction of mathematics, from elementary school students to Fields Medal winners, continually refine the neural code of that specific brain circuit.

And the organization of that circuit is under strong hereditary constraints, those of the universal genetic endowment that makes us human. While learning allows it to accommodate many new concepts, its overall architecture remains the same in all of us, independent of experience. My colleagues and I obtained strong support for this assertion when we studied the brain organization of mathematicians whose sensory experience, since childhood, has been radically different: blind mathematicians.25 However surprising this may seem, it is not uncommon for a blind person to become an excellent mathematician. Perhaps the best-known blind mathematician is Nicholas Saunderson (1682–1739), who became blind around the age of eight and was so brilliant that he ended up holding the chair of Isaac Newton at Cambridge University.

Saunderson is no longer available for a brain scan, but Marie Amalric and I managed to contact three contemporary blind mathematicians, all of whom held university positions in France. One of them, Emmanuel Giroux, is a true giant of mathematics and currently heads a laboratory of sixty people at the École normale supérieure in Lyon. Blind since the age of eleven, he is most well-known for his beautiful proof of an important theorem of contact geometry.

The very existence of blind mathematicians refutes Alan Turing’s empiricist view of the brain as a “notebook” with “lots of blank sheets” that sensory experience progressively fills out. Indeed, how could blind people infer, from such a distinct and restricted experience, the very same abstract notions as sighted mathematicians if they did not already possess the circuits capable of generating them? As Emmanuel Giroux says, paraphrasing The Little Prince, “In geometry, what is essential is invisible to the eye. It is only with the mind that you can see well.” In mathematics, sensory experiences do not matter much; it is the ideas and concepts that do the heavy lifting.

If experience determined the organization of the cortex, then our blind mathematicians, who learned about the world from touch and hearing, would activate, when they do mathematics, brain areas very different from those of the sighted. The neuronal recycling hypothesis, on the contrary, predicts that the neural circuits of mathematics should be fixed—only a specific set of brain areas, present at birth, should be capable of repurposing themselves to host such ideas. And this is indeed exactly what we found when we scanned our three blind professors. As we expected, when they visualized a mathematical statement and assessed its truth value, they recruited the very same parietal and frontal lobe pathways as a sighted mathematician (see figure 13 in the color insert). Sensory experiences were irrelevant: only this circuit could accommodate mathematical representations.

The only difference is that, when our three blind mathematicians thought about their favorite field, they also recruited an additional region of the brain: their early visual cortex, in the occipital pole, the brain region that, in any sighted person, processes the images that impinge on the retina! In fact, this is a result that Cédric Villani, another brilliant mathematician and Fields Medal winner, had intuitively predicted. When we discussed this experiment prior to running it, he jokingly said to me, “You know, Emmanuel Giroux is a truly great mathematician, but he is also very fortunate: because he is blind, he can devote even more cortex to math!”

Villani was right. In people with normal eyesight, the occipital region is too busy with early vision to perform any other function, such as mathematics. In the blind, however, it is released from this visual role, and instead of remaining inactive, it transforms itself to perform more abstract tasks, including mental calculation and mathematics.26 And in people who are born blind, this reorganization seems to be even more extreme: the visual cortex exhibits totally unexpected responses, not only to numbers and math, but also to the grammar of spoken language, similar to Broca’s area.27

The reason for such abstract responses in the visual cortex of blind people remains the subject of theoretical debate: Does this total reorganization of the cortex represent a genuine case of neuronal recycling, or is it merely an extreme proof of brain plasticity?28 In my opinion, the scales are tipped in favor of the neuronal recycling hypothesis, because there is evidence that the preexisting organization of this region is not erased, as it would be if brain plasticity acted as a sponge capable of wiping clean the visual cortex’s chalkboard. Indeed, the visual cortex of the blind seems to largely maintain its normal connectivity and neural maps29 while reusing them for other cognitive functions. In fact, because this part of the cortex is very large, one finds “visual” regions in the brains of blind people that respond not only to math and language, but also to letters and numbers (presented in Braille), objects, places, and animals.30 Most remarkable, in spite of such radical differences in sensory experience, these category-selective areas tend to be located at the same place in the cortex of sighted and blind individuals. For instance, the region of the brain that responds to written words is located at exactly the same place in a blind person as it is in a sighted reader—the only difference is that it responds to Braille rather than to printed letters. Once again, the function of this region seems to be largely determined by its genetically controlled connections to language areas, in addition, perhaps, to other innate properties, and therefore does not change when sensory inputs do.31 The blind entertain the very same categories, ideas, and concepts as sighted people—using very similar brain regions.

The neuronal recycling view of mathematics is not supported solely by the fact that the most elementary concepts (1 + 1 = 2) and the most advanced mathematical ideas (e−iπ + 1 = 0) make use of the same brain regions. Other discoveries, of a purely psychological nature, indicate that the mathematics we learn in school is based on the recycling of old circuits devoted to approximate quantities.

Think of the number five. Right now, your brain is reactivating a representation of an approximate quantity close to four and six and far from one and nine—you are activating number neurons very similar to those found in other primates, with a tuning curve that peaks around five, but also with weights in the nearby quantities four and six. The fuzzy tuning curve of those neurons is the main reason why it is hard, at a glance, to know if a set of objects contains exactly four, five, or six items. Now, please decide if five is larger or smaller than six. It seems instantaneous—you get to the correct answer (smaller) in an instant—and yet, experiments actually show that your answer is influenced by the approximate quantities: you are much slower when the numbers are close, like five and six, than when they are further apart, like five and nine, and you also make more errors. This distance effect32 is one of the signatures of an ancient representation of numbers that you recycled when you learned to count and calculate. No matter how much you try focusing on the symbols themselves, your brain can’t help but activate the neural representations of these two quantities, which overlap more the closer they are. Although you are trying to think of “exactly five,” using all the symbolic knowledge that you acquired at school, your behavior betrays the fact that this knowledge recycles an evolutionarily older representation of approximate quantity. Even when you simply have to decide whether two numbers such as eight and nine are the same or different, which should be immediate, you continue to be influenced by the distance between them—and, interestingly, exactly the same finding applies to monkeys who have learned to recognize the symbols of Arabic numerals.33

When we subtract two numbers, say, 9 − 6, the time that we take is directly proportional to the size of the subtracted number34—so it takes longer to perform 9 − 6 than, say, 9 − 4 or 9 − 2. Everything happens as if we have to mentally move along the number line, starting from the first number and taking as many steps as the second number: the further we have to go, the longer we take. We do not crunch symbols like a digital computer; instead, we use a slow and serial spatial metaphor, motion along the number line. Likewise, when we think of a price, we cannot help but attribute to it a fuzzier value when the number gets larger—a remnant of our primate-based number sense, whose precision decreases with number size.35 This is why, against all rationality, when we negotiate, we are ready to give up a few thousand dollars on the price of an apartment and, the same day, bargain a few quarters on the price of bread: the level of imprecision that we tolerate is proportional to a number’s value, for us just as for macaques.

And the list goes on: parity, negative numbers, fractions . . . all these concepts are demonstrably grounded in the representation of quantities that we inherit from evolution.36 Unlike a digital computer, we are unable to manipulate symbols in the abstract: we always grind them in concrete and often approximate quantities. The persistence of such analog effects in an educated brain betrays the ancient roots of our concept of numbers.

Approximate numbers are one of the old pillars on which the construction of mathematics is founded. However, education also leads to considerable enrichment of this original number concept. When we learn to count and calculate, the mathematical symbols that we acquire allow us to perform precise computations. This is a revolution: for millions of years, evolution had been content with fuzzy quantities. Symbol learning is a powerful factor for change: with education, all our brain circuits are repurposed to allow for the manipulation of exact numbers.

Number sense is certainly not the only foundation of mathematics. As we saw earlier, we also inherit from our evolution a sense of space, with its own specialized neural circuits containing place, grid, and head direction cells. We also have a sense of shape, which allows any toddler to distinguish rectangles, squares, and triangles. In a way that is not yet fully understood, under the influence of symbols such as words and numbers, all these concepts are recycled when we learn mathematics. The human brain manages to recombine them, in a language of thought, in order to form new concepts.37 The basic building blocks that we inherit from our evolutionary history become the foundational primitives of a new, productive language in which mathematicians write new pages every day.

READING RECYCLES THE CIRCUITS OF VISION AND SPOKEN LANGUAGE

What about learning to read? Reading is yet another example of neuronal recycling: to read, we reuse a vast set of brain areas that are initially dedicated to vision and spoken language. In my book Reading in the Brain,38 I describe, in detail, the circuits of literacy. When we learn to read, a subset of our visual regions becomes specialized in recognizing strings of letters and sends them to spoken language areas. As a result, in any good reader, written words end up being processed exactly as spoken words: literacy creates a new visual gateway to our language circuits.

Long before children learn how to read, they obviously possess a sophisticated visual system that allows them to recognize and name objects, animals, and people. They can recognize any image regardless of its size, position, or orientation in 3-D space, and they know how to associate a name to it. Reading recycles part of this preexisting picture naming circuit. The acquisition of literacy involves the emergence of a region of the visual cortex that my colleague Laurent Cohen and I have dubbed the “visual word form area.” This region concentrates our learned knowledge of letter strings, to such an extent that it can be considered as our brain’s “letter box.” It is this brain area, for instance, that allows us to recognize a word regardless of its size, position, font, or cAsE, whether UPPERCASE or lowercase.39 In any literate person, this region, which is located in the same spot in all of us (give or take a few millimeters), serves a dual role: it first identifies a string of learned characters, and then, through its direct connections to language areas,40 it allows those characters to be quickly translated into sound and meaning.

What would happen if we scanned an illiterate child or adult as she progressively learned to read? If the theory is correct, then we should literally see her visual cortex reorganize. The neuronal recycling theory predicts that reading should invade an area of the cortex normally devoted to a similar function and repurpose it to this novel task. In the case of reading, we expect a competition with the preexisting function of the visual cortex, which is to recognize all sorts of objects, bodies, faces, plants, and places. Could it be that we lose some of the visual functions that we inherited from our evolution as we learn to read? Or, at the very least, are these functions massively reorganized?

This counterintuitive prediction is precisely what my colleagues and I tested in a series of experiments. To draw a complete map of the brain regions that are changed by literacy, we scanned illiterate adults in Portugal and Brazil, and we compared them to people from the same villages who had had the good fortune of learning to read in school, either as children or adults.41 Unsurprisingly perhaps, the results revealed that, with reading acquisition, an extensive map of areas had become responsive to written words (see figure 14 in the color insert). Flash a sentence, word by word, to an illiterate individual, and you will find that their brain does not respond much: activity spreads to early visual areas, but it stops there, because the letters cannot be recognized. Present the same sequence of written words to an adult who has learned to read, and a much more extended cortical circuit now lights up, in direct proportion to the person’s reading score. The areas activated include the letter box area, in the left occipitotemporal cortex, as well as all the classical language regions associated with language comprehension. Even the earliest visual areas increase their response: with reading acquisition, they seem to become attuned to the recognition of small print.42 The more fluent a person is, the more these regions are activated by written words, and the more they strengthen their links: as reading becomes increasingly automatic, the translation of letters into sounds speeds up.

But we can also ask the opposite question: Are there regions that are more active among bad readers and whose activity decreases as one learns to read? The answer is positive: in illiterates, the brain’s responses to faces are more intense. The better we read, the more this activity decreases in the left hemisphere, at the exact place in the cortex where written words find their niche—the brain’s letter box area. It’s as if the brain needs to make room for letters in the cortex, so the acquisition of reading interferes with the prior function of this region, which is the recognition of faces and objects. But, of course, since we do not forget how to recognize faces when we learn to read, this function is not just chased out of the cortex. Rather, we have also observed that, with literacy, the response to faces increases in the right hemisphere. Driven out of the left hemisphere, which is the seat of language and reading for most of us, faces take refuge on the other side.43

We first made this observation in literate and illiterate adults, but we quickly replicated our results in children who were learning to read.44 As soon as a child begins to read, the visual word form area begins to respond in the left hemisphere. Meanwhile its symmetrical counterpart, in the right hemisphere, strengthens its response to faces (see figure 15 in the color insert). The effect is so powerful that, for a given age, just by examining the brain activity evoked by faces, a computer algorithm can correctly decide whether a child has or has not yet learned to read. And when a child suffers from dyslexia, these regions do not develop normally—neither on the left, where the visual word form area fails to emerge, nor on the right, where the fusiform cortex fails to develop a strong response to faces.45 Reduced activity of the left occipitotemporal cortex to written words is a universal marker of reading difficulties in all countries where it has been tested.46

In agreement with the neuronal recycling hypothesis, learning to read competes with the previous functions of the visual cortex—in this case, face recognition. With increasing levels of literacy, from pure illiterates to expert readers, the activation evoked by written words increases in the left hemisphere—and the activation evoked by faces moves from the left hemisphere to the right.

Recently, we got permission to conduct a bold experiment. We wanted to see the reading circuits emerge in individual children—and to this aim, we brought the same children back to our brain-imaging center every two months, from the end of kindergarten through the end of first grade. The results lived up to our expectations. The first time we scanned these children, there was not much to be seen: as long as the children had not yet learned how to read, their cortex responded selectively to objects, faces, and houses, but not to letters. After two months of schooling, however, a specific response to written words appeared, at the same exact location as in adults: the left occipitotemporal cortex. Very slowly, the representation of faces changed: as the children became more and more literate, face responses increased in the right hemisphere, in direct proportion to reading scores. Once again, in agreement with the neuronal recycling hypothesis, we could see reading acquisition compete with the prior function of the left occipitotemporal cortex, the visual recognition of faces.

We realized while doing this work that this competition could be explained in two different ways. The first possibility is what we called the “knockout model”: from birth on, faces settle in the visual cortex of the left hemisphere, and learning to read later knocks them straight out into the right hemisphere. The second possibility is what we termed the “blocking model”: the cortex develops slowly, gradually growing specialized patches for faces, places, and objects; and when letters enter this developing landscape, they take over part of the available territory and prevent the expansion of other visual categories.

So, does literacy lead to a knockout or a blockade of the cortex? Our experiments suggest the latter: learning to read blocks the growth of face-recognition areas in the left hemisphere. We witnessed this blockade thanks to the MRI scans that we acquired every two months from the children who were learning to read.47 At this age, around six or seven, cortical specialization is still far from complete. A few patches are already dedicated to faces, objects, and places, but there are also many cortical sites that have not yet specialized for any given category. And we could visualize their progressive specialization: when children entered first grade and quickly began to read, letters invaded one of those poorly specified regions and recycled it. Contrary to what I initially thought, letters do not completely overrun a preexisting face patch; they move in right next door, in a free sector of cortex, a bit like an aggressive supermarket setting up shop right next to a small grocery store. The expansion of one stops the other—and because letters settle down in the left hemisphere, which is dominant for language, faces have no choice but to move to the right side.

Learning is easier in childhood, while the cortex is still malleable. Before a young child goes to school, some visual regions of the brain have already specialized in recognizing objects, faces, and places—but there are still large patches with little or no specialization (symbolized by empty hexagons). Learning to read invades these labile circuits and blocks the growth of other categories of objects. If a child does not learn to read, those regions become involved in recognizing faces and objects, and gradually lose their ability to learn letters.

In brief, the ventral visual system is still undergoing major reorganization during the early school years. The fact that our schools typically teach children to read between the ages of six and eight nicely dovetails with the evidence for intense brain plasticity during this time period. We have organized our education system so that it efficiently takes advantage of a sensitive period when the visual cortex is particularly pliable. While its overall architecture is highly constrained from birth, the human inferotemporal cortex possesses the remarkable ability to adapt to various shapes and learn all kinds of images. When exposed to thousands of written words, this region recycles itself for this new activity, in a specific sector which happens to be innately connected to language circuits.

As we get older, our visual cortex seems to gradually freeze and lose the ability to tune to new images. The progressive closure of the sensitive period makes it more and more difficult for the cortex to efficiently recognize letters and their combinations. My colleagues and I studied two people who tried to learn to read as adults: one of them had never had the chance to go to school, while the other had suffered a small stroke in the visual word form area, rendering him fully “alexic”—unable to read. We scanned them regularly for two years.48 Their progress was incredibly slow. The first participant eventually developed a specialized region for letters, but this growth did not affect the face area—the circuits for face recognition had been imprinted in his brain and seemed no longer able to move. Our stroke patient, on the other hand, never managed to re-create a new “letter box” in his visual cortex. His reading improved but remained slow and similar to the laborious deciphering of a novice reader—being an adult, he was missing the neuronal plasticity necessary to recycle part of his cortex into an automated reading machine.

MUSIC, MATH, AND FACES

The conclusion is simple: to profoundly recycle our visual cortex and become excellent readers, we must take advantage of the period of maximum plasticity that early childhood offers. Our research shows several other examples. Take musical reading: a musician who learned to read sheet music at an early age has practically double the surface area of his visual cortex dedicated to musical symbols, compared with someone who has never learned music.49 This massive growth occupies space on the surface of the cortex, and it seems to dislodge the visual word form area from its usual place: in musicians, the cortical region that responds to letters, the brain’s letter box, is displaced by nearly one centimeter from its normal position in nonmusicians.

Another example is our varying abilities to decode mathematical equations. An accomplished mathematician must be able to recognize, at a glance, expressions as obscure as , or , just like we read a sentence in a novel. I once attended a conference where the brilliant French mathematician Alain Connes (another Fields Medal winner) exhibited an extraordinarily dense equation that was twenty-five lines long. He explained that this all-encompassing mathematical expression captured all the physical effects of all known elementary particles. A second mathematician pointed his finger and said, “Isn’t there an error on line thirteen?” “No,” Connes immediately answered without losing his composure, “because the corresponding compensating term is right there on line fourteen!”

How is this remarkable knack for complex formulas reflected in the brains of mathematicians? Brain imaging shows that these mathematical objects invade the lateral occipital regions of both hemispheres—after math training, these regions respond to algebraic expressions much more so than in non-mathematicians. And, once again, we witness a competition with faces: this time, the patches of face-responsive cortex wane away in both hemispheres.50 In other words, while literacy merely drives faces out of the left hemisphere and forces them to move over to the right hemisphere, intense practice with numbers and equations interferes with the representation of faces on both sides, leading to a global shrinkage of the visual face-recognition circuitry.

It is tempting to relate this finding to the famous myth of the eccentric mathematician, uninterested in anything other than his equations and unable to recognize his neighbor, his dog, or even his reflection in the mirror. There is, indeed, an abundance of anecdotes and jokes about mindless mathematicians. For instance, what’s the difference between an introvert mathematician and an extrovert mathematician? While he’s talking to you, the introvert looks at his shoes. But the extrovert mathematician looks at your shoes! . . .

In reality, we do not yet know whether the reduction in cortical responses to faces in math buffs is directly related to their supposed lack of social competence (which, I should say, is more of a myth than a reality—many mathematicians are wonderfully at ease in society). Most crucially, causality remains to be determined: Does spending one’s life in mathematical formulas reduce the response to faces? Or, on the contrary, do mathematicians immerse themselves in a universe of equations because they find them easier than social interactions? Whatever the answer may be, cortical competition is a genuine phenomenon, and the representation of faces in our brains turns out to be remarkably sensitive to education and schooling, to the point where it can provide a reliable marker of whether a child has received training in math, music, or reading. Neuronal recycling is a reality.

THE BENEFITS OF AN ENRICHED ENVIRONMENT

The take-home message is that both sides of the nature-nurture debate are right: a child’s brain is both structured and plastic. At birth, all children are equipped with a panoply of specialized circuits, shaped by genes, themselves selected by tens of millions of years of evolution. This self-organization gives the baby’s brain a deep intuition of several major areas of knowledge: a sense of the physics governing objects and their motion; a knack for spatial navigation; intuitions of numbers, probability, and mathematics; an inclination toward other human beings; and even a genius for languages—the blank-slate metaphor could not be more wrong. And yet evolution also left the door open to many learning opportunities. Not everything is predetermined in the child’s brain. Quite the contrary: the detail of neural circuits, on a scale of a few millimeters, is largely open to interactions with the outside world.

During the first years of life, genes guide an exuberant overproduction of neural circuits: twice as many synapses as necessary. In a way that we do not fully understand yet, this initial abundance opens up an immense space of mental models of the world. The brains of young children swarm with possibilities and explore a much wider set of hypotheses than the brains of adults. Each baby is open to all languages, all scripts, all possible mathematics—within the genetic limits of our species, of course.

And the baby’s brain also comes equipped with another innate gift: powerful learning algorithms that select the most useful synapses and circuits, thus providing a second layer of adaptation of the organism to its environment. Thanks to them, as early as the first few days of life, the brain begins to specialize and settle into its configuration. The first regions to freeze are the sensory areas: early visual areas mature in a few years, and it takes less than twelve months before the auditory areas converge toward the vowels and consonants of the child’s native language. As the sensitive periods of brain plasticity close, one after the other, a few years suffice for any of us to become a native of a given language, writing, and culture. And if we are deprived of stimulation in a certain domain, whether we are orphans in Bucharest or illiterates in the suburbs of Brasilia, we risk forever losing our mental flexibility in this field of knowledge.

This is not to say that intervention is not worth the effort, at any age: the brain retains some of its plasticity throughout its life, especially in its highest-level regions such as the prefrontal cortex. However, everything points to the optimal effectiveness of early intervention. Whether the goal is to make an owl wear glasses, teach an adopted child a second language, or help a child adjust to deafness, blindness, or the loss of a whole cerebral hemisphere, the sooner, the better.

Our schools are institutions designed to make the most of the plasticity of the developing brain. Education relies heavily on the spectacular flexibility of the child’s brain to recycle some of its circuits and reorient them toward new activities such as reading or mathematics. When schooling begins early, it can transform lives: numerous experiments show that children from disadvantaged backgrounds who benefit from early educational interventions show improved outcomes, even decades later, in many domains—from lower crime rates to higher IQs and incomes to better health.51

But schooling is not a magic pill. Parents and families also have a duty to stimulate children’s brains and enrich their environments as much as possible. All babies are budding physicists who love to experiment with gravity and falling bodies—as long as they are allowed to tinker, build, fail, and start over again, rather than being strapped in a car seat for hours. All children are nascent mathematicians who love counting, measuring, drawing lines and circles, assembling shapes—provided one gives them rulers, compasses, paper, and attractive math puzzles. All infants are genial linguists: as early as eighteen months of age, they easily acquire ten to twenty words a day—but only if they are spoken to. Their families and friends must feed this appetite for knowledge and nourish them with well-formed sentences, without hesitating to use a rich lexicon. Many studies show that a child’s vocabulary at three to four years old directly depends on the amount of child-directed speech they received during their first years. Passive exposure does not suffice: active one-to-one interactions are essential.52

All research findings are remarkably convergent: enriching the environment of a young child helps her build a better brain. For instance, in children who are read bedtime stories every evening, the brain circuits for spoken language are stronger than in other toddlers—and the strengthened cortical pathways are precisely those that will later allow them to understand texts and formulate complex thoughts.53 Likewise, children who are lucky enough to be born into bilingual families, with each parent giving them the wonderful gift of speaking in their native language, effortlessly acquire two lexicons, two grammars, and two cultures—at no cost.54 Throughout their lives, their bilingual brains retain better abilities for language processing and for acquiring a third or fourth language. And when they enter old age, their brains seem to resist the ravages of Alzheimer’s disease for longer. Exposing the developing brain to a stimulating environment allows it to keep more synapses, larger dendrites, and more flexible and redundant circuits55—like the owl that learned to wear prism glasses and kept, for its entire life, more diversified dendrites and a greater ability to switch from one behavior to another. Let’s diversify our children’s early learning portfolio: the blossoming of their brains depends in part on the richness of the stimulation they receive from their environment.