The word gap widens long before school starts

In a busy kindergarten classroom, two four-year-olds play quietly side by side. The first recounts her morning in full sentences, links one clause to the next, names precisely what she sees. The second struggles to point to an object right in front of him, hesitates, repeats the adult's words without ever making them his own. The teacher genuinely sees nothing unusual here, since every child moves at their own pace. But what this scene hides is a gap that did not begin this school year. It took shape years earlier, in the thousands of hours of language heard, or not heard, long before either child ever walked through the school door.

Why we look to school to explain school failure
When a child clearly falls behind, the natural instinct is to look for the cause somewhere inside the classroom: the wrong method, an overstretched teacher, a badly designed curriculum. That reading certainly has the merit of pointing to somewhere concrete to act, but it wrongly assumes the gap is born at school, when it is often already largely in place. A child who arrives in kindergarten with a noticeably smaller vocabulary than his classmates does not start from the same point, and no teaching method, however good, can by itself close in a few months what took years of unequal exposure to language to build.

What the study on language processing at 18 months reveals
A study by Anne Fernald, Virginia Marchman, and Adriana Weisleder, published in Developmental Science, measured how quickly 18-month-old infants recognized a familiar word by tracking their eye movements toward the matching picture. Infants from lower-income households consistently processed language noticeably more slowly than those from more advantaged households, and this speed gap already came paired with a measurably smaller vocabulary. The gap, then, does not appear at the school gate: it is detectable before a child can even string two words together.

Why processing speed matters as much as vocabulary size
A child who takes longer to recognize a word he just heard has less mental bandwidth left to process the rest of the sentence, which in turn shrinks his ability to extract meaning and pick up new words along the way. This apparent slowness is not simply a personality trait or an attention deficit: it directly reflects the amount of language heard over the preceding months, since a brain exposed to a rich and varied stream of speech mechanically becomes faster and more precise at decoding it. The fewer words a child hears addressed directly to him, the more each new word costs him to process.

What the early gap really drives: a cumulative disadvantage, not a one-time delay
The real problem is not only that some children start out with noticeably fewer words in stock at 18 months. It is that this initial deficit then slows the acquisition of every word that follows, which widens the gap rather than stabilizing it. A cumulative disadvantage of this kind behaves like compound interest: the longer it goes unaddressed, the larger the gap that needs closing becomes, which explains why programs launched at age six or seven struggle so much to catch up children whose delay was already built in during early infancy.

How to act before school rather than after
In practice, this genuinely shifts the whole question away from school and toward what happens well before it: talking more to very young children, asking them questions, narrating what they are doing, rather than simply issuing instructions. Once at school, arrangements such as peer tutoring can also help multiply opportunities for rich verbal exchange for the children who benefited least from it earlier, without pretending to replace the missing years. This is not a question of parental merit or school discipline, but of cumulative language exposure, a factor that remains largely invisible until it is measured directly.

In the kindergarten classroom, nothing visibly sets the two children apart: same age, same room, same teacher. But one of them, even before learning to speak, had already heard millions more words than the other. This is neither simply a problem of intelligence nor merely a problem of school. It is a problem of exposure, settled long before the bell ever rings for the first time.

20 October 2026

In a busy kindergarten classroom, two four-year-olds play quietly side by side. The first recounts her morning in full sentences, links one clause to the next, names precisely what she sees. The second struggles to point to an object right in front of him, hesitates, repeats the adult’s words without ever making them his own. The teacher genuinely sees nothing unusual here, since every child moves at their own pace. But what this scene hides is a gap that did not begin this school year. It took shape years earlier, in the thousands of hours of language heard, or not heard, long before either child ever walked through the school door.

Why we look to school to explain school failure

When a child clearly falls behind, the natural instinct is to look for the cause somewhere inside the classroom: the wrong method, an overstretched teacher, a badly designed curriculum. That reading certainly has the merit of pointing to somewhere concrete to act, but it wrongly assumes the gap is born at school, when it is often already largely in place. A child who arrives in kindergarten with a noticeably smaller vocabulary than his classmates does not start from the same point, and no teaching method, however good, can by itself close in a few months what took years of unequal exposure to language to build.

What the study on language processing at 18 months reveals

A study by Anne Fernald, Virginia Marchman, and Adriana Weisleder, published in Developmental Science, measured how quickly 18-month-old infants recognized a familiar word by tracking their eye movements toward the matching picture. Infants from lower-income households consistently processed language noticeably more slowly than those from more advantaged households, and this speed gap already came paired with a measurably smaller vocabulary. The gap, then, does not appear at the school gate: it is detectable before a child can even string two words together.

Why processing speed matters as much as vocabulary size

A child who takes longer to recognize a word he just heard has less mental bandwidth left to process the rest of the sentence, which in turn shrinks his ability to extract meaning and pick up new words along the way. This apparent slowness is not simply a personality trait or an attention deficit: it directly reflects the amount of language heard over the preceding months, since a brain exposed to a rich and varied stream of speech mechanically becomes faster and more precise at decoding it. The fewer words a child hears addressed directly to him, the more each new word costs him to process.

What the early gap really drives: a cumulative disadvantage, not a one-time delay

The real problem is not only that some children start out with noticeably fewer words in stock at 18 months. It is that this initial deficit then slows the acquisition of every word that follows, which widens the gap rather than stabilizing it. A cumulative disadvantage of this kind behaves like compound interest: the longer it goes unaddressed, the larger the gap that needs closing becomes, which explains why programs launched at age six or seven struggle so much to catch up children whose delay was already built in during early infancy.

How to act before school rather than after

In practice, this genuinely shifts the whole question away from school and toward what happens well before it: talking more to very young children, asking them questions, narrating what they are doing, rather than simply issuing instructions. Once at school, arrangements such as peer tutoring can also help multiply opportunities for rich verbal exchange for the children who benefited least from it earlier, without pretending to replace the missing years. This is not a question of parental merit or school discipline, but of cumulative language exposure, a factor that remains largely invisible until it is measured directly.