The Secret Life of the Mind: How Our Brain Thinks, Feels and Decides

One solution is talking to babies – as we do when speaking Motherese – slowly and with exaggerated enunciation. In Motherese there are pauses between words, which facilitates the baby’s heroic task of dividing a sentence into the words that make it up.

But this doesn’t explain per se how eight-month-olds already begin to form a vast repertoire of words, many of which they don’t even know how to define. In order to do this, the brain uses a principle similar to the one many sophisticated computers employ to detect patterns, known as statistical learning. The recipe is simple and identifies the frequency of transitions between syllables and function. Since the word hello is used frequently, every time the syllable ‘hel’ is heard, there is a high probability that it will be followed by the syllable ‘lo.’ Of course, these are just probabilities, since sometimes the word will be helmet or hellraiser, but a child discovers, through an intense calculation of these transitions, that the syllable ‘hel’ has a relatively small number of frequent successors. And so, by forming bridges between the most frequent transitions, the child can amalgamate syllables and discover words. This way of learning, obviously not a conscious one, is similar to what smartphones use to complete words with the extension they find most probable and feasible; as we know, they don’t always get it right.

This is how children learn words. It is not a lexical process as if filling a dictionary in which each word is associated with its meaning or an image. To a greater extent, the first approach to words is rhythmic, musical, prosodic. Only later are they tinged with meaning. Marina Nespor, an extraordinary linguist, suggests that one of the difficulties of studying a second language in adulthood is that we no longer use that process. When adults learn a language, they usually do so deliberately and by using their conscious apparatus; they try to acquire words as if memorizing them from a dictionary and not through the musicality of language. Marina maintains that if we were to imitate the natural mechanism of first consolidating the words’ music and the regularities in the language’s intonation, our process of learning would be much simpler and more effective.

The children of Babel

One of the most passionately debated examples of the collision between biological and cultural predispositions is bilingualism. On one hand, a very common intuitive assumption is: ‘Poor child, just learning to talk is difficult, the kid’s gonna get all mixed up having to learn two languages.’ But the risk of confusion is mitigated by the perception that bilingualism implies a certain cognitive virtuosity.

Bilingualism, actually, offers a concrete example of how some social norms are established without the slightest rational reflection. Society usually considers monolingualism to be the norm, so that the performance of bilinguals is perceived as a deficit or an increment in relation to it. That is not merely convention. Bilingual children have an advantage in the executive functions, but this is never perceived as a deficit in monolinguals’ potential development. Curiously, the monolingual norm is not defined by its popularity; in fact, most children in the world grow up being exposed to more than one language. This is especially true in countries with large immigrant populations. In these homes, languages can be combined in all sorts of forms. As a boy, Bernardo Houssay (later awarded the Nobel Prize for Physiology) lived in Buenos Aires, Argentina (where the official language is Spanish) with his Italian grandparents. His parents spoke little of their parents’ language, and he and his brothers spoke none. So he believed that people, as they aged, turned into Italians.

Cognitive neuroscientific research has conclusively proven that, going against popular belief, the most important landmarks in language acquisition – the moment of comprehending the first words, the development of sentences, among others – are very similar in monolinguals and bilinguals. One of the few differences is that, during infancy, monolinguals have a bigger vocabulary. However, this effect disappears – and even reverts – when the words a bilingual can use in both languages are added to that vocabulary.

A second popular myth is that one shouldn’t mix languages and that each person should speak to a child always in the same language. That is not the case. Some studies in bilingualism are conducted with parents who each speak one language exclusively to their children, which is very typical in border regions, such as where Slovenia meets Italy. In other studies, in bilingual regions such as Quebec or Catalonia, both parents speak both languages. The developmental landmarks in these two situations are identical. And the reason why the babies don’t get confused by one person speaking two languages is because, in order to produce the phonemes of each language, they give gesticular indications – the way they move their mouths and face – of which language they are speaking. Let’s say that one makes a French or an Italian facial expression. These are easy clues for a baby to recognize.

On the other hand, another large group of evidence indicates that bilinguals have a better and faster development of the executive functions; more specifically, in their ability to inhibit and control their attention. Since these faculties are critical in a child’s educational and social development, the advantage of bilingualism now seems quite obvious.

In Catalonia, children grow up in a sociolinguistic context in which Spanish and Catalan are often used in the same conversation. As a consequence, Catalan children develop skills to shift rapidly from one language to the other. Will this social learning process extend to task-switching beyond the domain of language?

To answer this question, César Ávila with his colleagues compared brain activity of monolinguals and Catalan bilinguals who switched between non-linguistic tasks. Participants saw a sequence of objects flashing rapidly in the centre of a screen. For a number of trials they were asked to respond with a button if the object was red, and with another button if it was blue. Then, suddenly, participants were asked to forget about colour and respond using the same buttons about the shape of the object (right button for a square and left button for a circle).

As simple as this sounds, when task instructions switch from colour to shape most people respond more slowly and make more errors. This effect is much smaller in Catalonian bilinguals. Ávila also found that the brain networks used by monolinguals and bilinguals to solve this task are very different. It is not that bilinguals are just increasing slightly the amount of activity in one region; it is that the problem in the brain is solved in an altogether different manner.

To switch between tasks, monolinguals use brain regions of the executive system such as the anterior cingulate and some regions in the frontal cortex. Bilinguals instead engage brain regions of the language network, the same regions they engage to switch between Spanish and Catalan in a fluid conversation.

This means that in task-switching, even if the tasks are non-linguistic (in this case switching between colour and shape), bilinguals engage brain networks for language. Which is to say, bilinguals can recycle those brain structures that are highly specialized for language in monolinguals, and use them for cognitive control beyond the domain of language.

Speaking more than one language also changes the brain’s anatomy. Bilinguals have a greater density of white matter – bundles of neuronal projections – in the anterior cingulate than monolinguals do. And this effect doesn’t pertain only to those who learned more than one language during childhood. It is a characteristic that has been seen also in those who became bilingual later in life, and as such it might be particularly useful in old age, because the integrity of the connections is a decisive element in cognitive reserve. This explains why bilinguals, even when we factor in age, socioeconomic level and other relevant factors, are less prone to developing senile dementias.

To sum up, the study of bilingualism allows us to topple two myths: language development doesn’t slow down in bilingual children, and the same person can mix languages with no problem. What’s more, the effects of bilingualism may go above and beyond the domain of language, helping develop cognitive control. Bilingualism helps children to be captains of their own thought, pilots of their existence. This ability is decisive in their social inclusion, health and future. So perhaps we should promote bilingualism. Amidst so many less effective and more costly methods of stimulating cognitive development, this is a much simpler, beautiful and enduring way to do so.

A conjecturing machine

Children, from a very young age, have a sophisticated mechanism for seeking out and building knowledge. We were all scientists in our childhood,fn10 and not only out of a desire to explore, to break things apart to see how they work – or used to work – or to pester adults with an infinite number of questions beginning ‘Why?’ We were also little scientists because of the method we employed to discover the universe.

Science has the virtue of being able to construct theories based on scant, ambiguous data. From the paltry remnants of light from some dead stars, cosmologists were able to build an effective theory on the origin of the universe. Scientific procedure is especially effective when we know the precise experiment to discriminate between different theories. And kids are naturally gifted at this job.

A game with buttons (push buttons, keys or switches) and functions (lights, noise, movement) is like a small universe. As they play, children make interventions that allow them to reveal mysteries and discover the causal rules of that universe. Playing is discovering. In fact, the intensity of a child’s game depends on how much uncertainty the child has with regard to the rules that govern it. And when children don’t know how a simple machine works, they usually spontaneously play in the way that is most effective to discover its functioning mechanism. This is very similar to a precise aspect of the scientific method: investigation and methodical exploration in order to discover and clarify causal relationships in the universe.

But children’s natural exploration of science goes even further: they construct theories and models according to the most plausible explanation for the data they observe.

There are many examples of this, but the most elegant begins in 1988 with an experiment by Andrew Meltzoff – again – which produced the following scene. An actor enters a room and sits in front of a box with a large plastic button, pushes the button with their head and, as if the box were a slot machine paying out, there is a fanfare with colourful lights and sounds. Afterwards, a one-year-old baby who has been observing the scene is seated, on their mother’s lap, in front of the same machine. And then, spontaneously, the young child leans forward and presses the button with their head.

Did they simply imitate the actor or had the one-year-old discovered a causal relationship between the button and the lights? Deciding between these two possibilities would require a new experiment like the one proposed by the Hungarian psychologist György Gergely, fourteen years later. Meltzoff thought that the babies were imitating the actor when they pressed the button with their head. Gergely had another, much bolder and more interesting idea. The babies understand that the adult is intelligent and, because of that, if they didn’t push the button with their hand, which would be more natural, it was because pushing it with their head was strictly necessary.

This bold theory suggests that the reasoning of babies turns out to be much more sophisticated, and includes a theory of how things and people work. But how can one detect such reasoning in a child that doesn’t yet talk? Gergely solved it in a simple, elegant way. Imagine an analogous situation in everyday life. A person is walking with many bags and opens a door handle with an elbow. We all understand that door handles are not meant to be opened with your elbow and the person did that because there was no other option. What would happen if we replicated this idea in Meltzoff’s experiment? The same actor arrives, loaded down with bags, and pushes the button with their head. If the babies are simply imitating the actor, they would do the same. But if, on the other hand, they are capable of thinking logically, they will understand that the actor pushed it with their head because their hands were full and, therefore, all the babies needed to do to get the colourful lights and sounds was to push the button, with any part of their body.

They carried out the experiment. The baby observed the actor, laden with shopping bags, pushing the button with their head. Then the child sits on their mother’s lap and pushes the button with their hands. It is the same baby that, upon seeing the actor do the same thing but with their hands free, had pushed the button with their head.

One-year-olds construct theories on how things work based on what they observe. And among those observations is that of perceiving other people’s perspectives, working out how much they know, what they can and cannot do. In other words, exploring science.

The good, the bad and the ugly

We began this chapter with the arguments of the empiricists, according to which all logical and abstract reasoning occurs after the acquisition of language. But nevertheless we saw that even newborns form abstract and sophisticated concepts, that they have notions of mathematics, and display some understanding of language. At just a few months old, they already exhibit a sophisticated logical reasoning. Now we will see that young children who do not yet speak have also forged moral notions, perhaps one of the fundamental pillars of human social interaction.

The infants’ ideas of what is good, bad, fair, property, theft and punishment – which are already quite well established – cannot be fluently expressed because their control tower (circuits in the prefrontal cortex) is immature. Hence, as occurs with numerical and linguistic concepts, the infants’ mental richness of moral notions is masked by their inability to express it.

One of the simplest and most striking scientific experiments to demonstrate babies’ moral judgements was done by Karen Wynn in a wooden puppet theatre with three characters: a triangle, a square and a circle. In the experiment, the triangle goes up a hill. Every once in a while it backs up only to later continue to ascend. This gives a vivid impression that the triangle has an intention (climbing to the very top) and is struggling to achieve it. Of course, the triangle doesn’t have real desires or intentions, but we spontaneously assign it beliefs and create narrative explanations of what we observe.

A square shows up in the middle of this scene and bumps into the triangle on purpose, sending it down the hill. Seen with the eyes of an adult, the square is clearly despicable. As the scene is replayed, the circumstances change. While the triangle is going up, a circle appears and pushes it upwards. To us the circle becomes noble, helpful and gentlemanly.

This conception of good circles and bad squares needs a narrative – which comes automatically and inevitably to adults – that, on the one hand, assigns intentions to each object and, on the other, morally judges each entity based on those intentions.

As humans, we assign intentions not only to other people but also to plants (‘sunflowers seek out the sun’), abstract social constructions (‘history will absolve me’ or ‘the market punishes investors’), theological entities (‘God willing’) and machines (‘damn washing machine’). This ability to theorize, to turn data into stories, is the seed of all fiction. That is why we can cry in front of a television set – it is strange to cry because something happens to some tiny pixels on a screen – or destroy blocks on an iPad as if we were in a trench on the Western Front during the First World War.

In Wynn’s puppet show there are only triangles, circles and squares, but we see them as someone struggling, a bad guy who hinders progress, and a do-gooder who helps. Which is to say that, as adults, we have an automatic tendency to assign moral values. Do six-month-olds have that same abstract thought process? Would babies be able spontaneously to form moral conjectures? We can’t know by asking because they don’t yet talk, but we can infer this narrative by observing their preferences. The constant secret of science consists, precisely, in finding a way of bridging what we want to know – in this case, whether babies form moral concepts – with what we can measure (which objects the babies choose).

After watching one object helping the triangle climb the hill and another bumping it down, infants were encouraged to reach for one of them. Twenty-six of twenty-eight (twelve out of twelve six-month-olds) chose the helper. Then, the video recordings of the infants watching the scenes of the helper and the hinderer were shown to an experimentalist. And, relying on their facial gestures and expressions alone, she could predict almost perfectly whether the infant had just seen the helper or the hinderer.

Six-month-old infants, before crawling, walking or talking, when they are barely discovering how to sit up and eat with a spoon, are already able to infer intentions, desires, kindness and evil, as can be deduced from examining their choices and gestures.

He who robs a thief …

The construction of morality is, of course, much more sophisticated. We cannot judge a person to be good or bad just by knowing they did something helpful. For example, helping a thief is usually considered ignoble. Would the babies prefer someone who helps a thief to someone who thwarts one? We are now in the murky waters that are the origins of morality and law. But even in this sea of confusion, babies between nine months and a year of age already have an established opinion.

The experiment that proves it goes like this. Babies see a hand puppet trying to lift the top off a box in order to pull out a toy. Then a helpful puppet shows up and helps it open the lid and get the toy. But in another scene an anti-social puppet jumps maliciously on to the box, slamming it shut and keeping the first puppet from getting at the toy. When choosing between the two puppets, the babies prefer the helper. But here Wynn was going for something much more interesting: identifying what the babies think about stealing from an evildoer, long before they know those words.

To do this she designed a third act for the puppet theatre, and the helper puppet now loses a ball. In some cases, in this garden of forking paths, a new character appears on the scene and returns the ball. At other times, another character comes in, steals it and runs away. The babies prefer the character that returns the ball.

But the most interesting and mysterious part happens when these scenes feature the antisocial puppet that jumped maliciously on the box. In this case, the babies change their preference. They sympathize with the one who steals the ball and runs away. For nine-month-olds, the one who gives the bad guy his comeuppance is more lovable than the one who helps him, at least in that world of puppets, boxes and balls.fn11

Preverbal babies, still unable to coordinate their hands in order to grab an object, do something much more sophisticated than judging others by their actions. They take into account the contexts and the history, which turns out to give them a pretty sophisticated notion of justice. That’s how incredibly disproportionate cognitive faculties are during the early development of a human being.

The colour of a jersey, strawberry or chocolate

We adults are not unbiased when we judge others. Not only do we keep in mind their previous history and the context of their actions (which we should), but we also have very different opinions of the person committing the actions, or being the victim of them, if they look like us or not (which we shouldn’t).

Throughout all cultures, we tend to form more friendships and have more empathy with those who look like us. On the other hand, we usually judge more harshly and show more indifference to the suffering of those who are different. History is filled with instances in which human groups have massively supported or, in the best-case scenario, rejected violence directed at individuals who were not like them.

This even manifests itself in formal justice proceedings. Some judges serve sentences displaying a racial bias, most probably without being aware that race is influencing their judgement. In the United States, African American males have been incarcerated at about six times the rate of white males. Is this difference in the incarceration rate a result (at least in part) of the judges having different sentencing practices? This seemingly simple and direct question turns out to be hard to answer because it is difficult to separate this psychological factor from possible racial differences in case characteristics. To overcome this problem Sendhil Mullainathan, Professor of Economics at Harvard University, found an ingenious solution, exploiting the fact that in the United States cases are randomly assigned to judges. Hence, on average, the type of case and the nature of defendants are the same for all judges. A racial difference in sentencing could potentially be explained by case characteristics or by a difference in the quality of the assigned attorneys (which is not random). But if this were all, then this difference should be the same for all judges. Instead, Mullainathan found a huge disparity – of almost 20 per cent – between judges in the racial gap in sentencing. While this may be the most convincing demonstration that race matters in the courtroom, the method is partly limited since it cannot tell whether the variability between judges’ results is due to some of them discriminating against African Americans, or some judges discriminating against whites, or a mixture of both.

Physical appearance also affects whether someone is likely to be hired in a job interview. Since the early seventies, several studies have shown that attractive applicants are typically judged to have a more appropriate personality for a job, and to perform better than their less attractive counterparts. Of course, this was not just a matter of comment. Applicants who were judged to be more attractive were also more likely to be hired. As we will see in Chapter 5, we all tend to make retrospective explanations that serve to justify our choices. Hence the most likely timeline for this line of argument is like this: first the interviewer decides to hire the applicant (among other things based on his or her beauty) and only then generates ad hoc a long list of attributes (he or she was more capable, more suited for the job, more reliable …) that serve to justify the choice which indeed had nothing to do with these considerations.

The similarities that generate these predispositions can be based on physical appearance, but also on religious, cultural, ethnic, political or even sports-related questions. This last example, because it is presumed to be more harmless – although, as we know, even sporting differences can have dramatic consequences – is easier to assimilate and recognize. Someone forms part of a consortium, a club, a country, a continent. That person suffers and celebrates collectively with that consortium. Pleasure and pain are synchronized between thousands of people whose only similarity is belonging to a tribe (sharing a jersey, a neighbourhood or a history) that unites them. But there is something more: pleasure at the suffering of other tribes. Brazil celebrates Argentina’s defeats, and Argentina celebrates Brazil’s. A fan of Liverpool cheers for the goal scored against Manchester United. When rooting for our favourite sports teams, we often feel less inhibited about expressing Schadenfreude, our pleasure at the suffering of those unlike us.