Here is the full text and summary of psychologist Anne Fernald’s talk titled “Why Talking To Little Kids Matters” at TEDxMonterey conference.
Listen to the audio version here:
TRANSCRIPT:
The science tells us that genes specify potentials, but not outcomes. The realization of outcomes depends on support from the environment at every level. And it’s the support that I love in that picture. The firm hands of the parents holding that newborn, who depends on them for everything. Not just for food, for shelter, for warmth and safety, but also for the linguistic nutrition and the mental exercise that will furnish that child’s mind, build that child’s brain and intelligence.
Babies are born ready to learn, and we hold them in our hands over the first thousand days of their life, for better or for worse. The family is the world of that child.
Now for some children, it is for worse. Here what you see are children living in poverty, that children living in poverty are already behind by two years when they start school. So the blue line shows you advantaged families, higher in socioeconomic status, or SES, and the red line shows lower SES families, whose parents have less education and less income.
And what you see is a two-year gap at the starting gate that grows a little bigger. At best that gap levels off and stays constant. It does not go away. How early do these gaps begin to emerge? Where do they come from? That’s one of the questions that we’re going to be asking and that modern brain science is illuminating.
In this graph here, what you see is the growth of brain structure over the first thousand days, about three years. And as you can see, the lower-income families and the mid-income families and the high-income families, children from those families started out with similar volumes of grey matter. Grey matter are the neurons, the axons and dendrites that are so important in information processing.
And it was only through experience in the world with poverty that these differences between rich children and poor children began to emerge and to get wider. Those differences are enduring and consequential. Now there are many, many causes for those outcomes. You don’t want to oversimplify that problem because many of these risk factors you see here all happen at once. I’m going to be focusing on just one, and that’s inadequate opportunities for early learning.
There are many children that don’t experience a form of parental support that in principle could be available to everyone, and that’s parents’ ability to nurture their children’s brain development through their social and verbal engagement with that little baby, through talking to the child in rich and supportive ways.
Why should we talk to a child when it’s not talking to us? A lot of parents ask that question. Pretty reasonable. Well, because hearing language is how you learn it. It forms the basis of oral language skills, and oral language skills are really fundamental to intelligence. In the words of Jerome Bruner, an eminent cognitive psychologist, proficiency in oral language provides children with a vital tool for thought, and without fluent and structured oral language, children will find it very difficult to think.
A pioneering study done in Kansas in the 1970s by Hart and Risley explored the origins of oral language knowledge. What they did was to follow 42 families from across a range of SES, from professional to welfare families, and they visited them once a month for one hour and recorded the children and the adults interacting in a spontaneous way in the home, and then laboriously counted the words and analyzed all of that language.
Their stunning finding was that by the age of four years, the children in the professional families had heard 30 million more words directed to them than children in the welfare families. Now, a lot of people believe that language is innate, but no matter how strong your belief is there, that is a difference that has to matter. So let me introduce you to two little children, 18 months old.
This is Juan, and Juan hears 100 words in five minutes on 10 different topics, color-coded topics. These aren’t fancy topics. Hey, want to play with some toys? Look, I got a bunny and I got a bear. And hey, the bunny’s going to hug the bear. And look, the bear has tiny little ears and the bunny has long, silky ears, and they’re just kind of like yours, those little bear ears. Nonsense, right?
Absolutely obvious, but Juan doesn’t know these things. He’s learning about toys and animals. He’s learning about bunnies and bears. He’s learning about who has what kind of ears and how it relates to him.
And now there’s little Rosa. She hears five words in five minutes on two topics. One topic is, come on over here. Another topic is, here’s some animals. And then four minutes of silence. Those were affectionate words. She’s a well-loved child, but compared to the rich diet of language that Juan experiences, by the way, on a day-to-day basis, cumulatively adding up to a lot, this is a very meager diet.
These are examples of the extreme differences. And by the way, both of these children were in a low SES group of Latino families that we’ve been working with.
So how are we going to measure these differences in language? Well, our methods have enabled us to measure the processing speed, something that you don’t see in the behavior, but the mental processing speed of children as they hear a word. So when you hear a familiar word like dog, how quickly does the light go on, right? And this is going to change with development, as we will see.
It’s a very simple task. This little girl is 24 months, and she’s sitting on her mom’s lap looking at two pictures. She’ll listen to speech that is naming one of the pictures. And then as you’ll see, she’ll look back and forth because she’s eager to find meaning in this situation.
See the apple?
Do you see it? Where’s the doggy? Can you see it? Show me the juice. Do you like it?
Now what we found was that at the time when children are just beginning to talk, around 18 months to 30 months, they make incredible progress in understanding as well, in processing speech as it flies by in real time. And now what we’re going to do is follow one little boy at three different intervals of six months.
We’re going to start with him at 18 months, and you’re going to see how he picks up speed in understanding. Now you’ll see the stimulus underneath his picture. When the dots are red, he’s on the wrong picture. When the dots are blue, he’s moved to the right picture.
Where’s the doggy? Can you see it?
He did pretty good. You know what a dog is, right? He was over there. If this child was sort of sitting in the wild on the rug, you would think that that was a fairly fast response.
But six months later, at 24 months, he’s really picked up speed. Take the book. Check that out. You might not notice that with the naked eye, but he just knocked off 600 milliseconds from his reaction time to that word measured by the red dots. 600 milliseconds doesn’t sound like much, but it is huge in terms of brain speed.
And now, next, at 30 months, it’s great. He nails it, and he knows it, and gives this cocky little smile. Where’s the doggy? Can you find it? Got that, right? Yeah.
And so at 18 months, it was, where’s the doggy? And then he shifted after the word was over. At 30 months, all he needs is, where’s the… And he places his bets early, right? And that’s all you need. That’s all you need to make your move, because you’re operating probabilistically on statistics.
What this shows is that kids are using less and less phonetic information, which enables you to keep up with my 200 words a minute. So here’s one way to capture this impressive progress. So they all start at 0.5. That’s chance performance, because they don’t know what word’s coming up. And you see these children, then, at three different ages. As they get it, as the light comes on, the curves head upward.
And what you see in children at different ages is the 18-monthers, you know, they’re starting their move at the end of the word, and they’re moderately reliable. But by 30 months, they’re making their move much, much sooner, and they’re much more reliable. This is what we mean by efficiency in processing.
And I’ll just tell you very briefly that there is also huge variability at every age. So some of those 24-monthers are as fast as the 30-monthers, and some of those 24-monthers are slower than the 18-monthers. And these differences matter.
Where you are in that distribution at two years of age is predicting out to age five, out to age eight, in terms of cognitive measures that are relative to school success. So we’re tapping into something about mental processing speed that has to do with fundamental ability to make sense of human language.
Okay, these were exciting findings. But, you know, in universities, we have this tendency to study children who come from incredibly affluent, privileged families and then make claims about homo sapiens. So we decided to move out of the lab and into the world, and over the next several years, opened community labs.
Here’s our – there’s our country club of a campus lab up there in the corner. And then we have a community lab in the right-hand corner in San Jose, where we’re working with low-SES families who are Spanish-speaking. We also have a 31-foot RV. I’m the proud owner of that. That is a mobile lab that we take up to Northern California to find English-speaking families who are lower SES.
And then we also go to Mexico City in order to find professional monolingual Spanish-speaking women who are equivalent to the Google and Facebook moms that live in – that come to the lab at Stanford. So this way, we’re including linguistic diversity as well as demographic diversity.
Now, one important question was, if we looked at differences, or would we find differences between children varying in SES, differences in processing speed as well as in vocabulary that could explain – could contribute to those stubborn achievement gaps that are so relevant to school success?
We tested this both in English and in Spanish. The results are the same. So you look at these children at 18 and 24 months in both – both groups there. And between 18 and 24 months, both groups make progress, right? They gain in efficiency.
But here’s the more troubling finding. In the low SES group, it’s not until 24 months that those kids reach the level of efficiency that the high SES kids started with six months earlier, right? So you’ve got basically the same level in much younger kids in the high SES group than in the low SES group. That’s the beginning of the pulling apart of these trajectories that we started with there.
All right, where do these differences come from? Well, back to Hart and Risley, we asked, is it possible that the language that children are hearing at home could be contributing to these differences in a language processing task that we know has what we call predictive validity, that tells us something about how children are going to be doing later on.
And we used a more modern technology than they did. It’s called Lina. It’s a little digital recorder, especially for recording babies. And there’s special clothing with a little pocket here. You put the recorder there, and then you can record the baby just loose in the world without an observer present – well, not entirely loose, but we weren’t there. And get a 10- or 12-hour-a-day recording. And it counts the words. It counts the words. That’s what’s quite amazing about it.
All right, so here are some results from a study of 29 families. These were all Latino families living in the Silicon Valley with an average education level of maybe sixth grade and very low income. What you see on the X axis are all the different families. Each bar is a different family. But what you see on the Y axis is the number of adult words per hour. Huge variability within this low-income group.
The most talkative family, there are 1,200 words an hour going on, or more than that, 1,500 words an hour going on. And down there at the bottom, very, very little.
Okay, so the blue is all the speech the child heard. Maybe that’s what matters. The green is speech that was addressed to the child. We suspected it was that that mattered, not just overheard speech, not television and these sorts of things. So there you have to listen to those 16-hour recordings and make the judgment about whether it’s child-directed speech.
Here’s where our two little friends fit into that distribution, right? There’s Juan here on the right and little Rosa on the left. And what you see then is huge differences among low-income families in how much speech they are providing to their child. Does this matter? Oh, boy, does it matter.
This is a curve showing the processing efficiency of children who heard more speech at 18 months, that’s in red, and children who heard less speech, maternal speech, at 18 months. This is, again, a processing measure of how quickly they’re looking at the picture, and this is a very substantial difference already by 24 months in their ability to grab that word and go with it.
Now, I’ve been emphasizing SES differences as if SES was destiny. It’s not. When you add to our sample of Spanish-speaking families, those in the Bay Area, those from Mexico, you get a distribution that looks like this. The lower SES families are the orange bars. The higher SES from Mexico City are the green bars. Kids who heard the most speech have the moms from Mexico City. Kids who heard the least had the moms who were the immigrant moms to the Silicon, to the Bay Area.
But look, look, look at all that variability in between. This is hopeful news, right? Because it suggests that we’re not talking about some fundamental behavior of a particular group, but rather we’re talking about speech to the child that matters, not SES per se. It’s speech to the child that is what’s important.
Why is it important? We’ve shown that it is, but now we’re interested in thinking about, you know, what is nourishing about the way we talk to kids? Well, here’s a good example. You’re out on a walk with your little two-year-old, and you say, look at the kitty on the bench. The child knows the word kitty. If he is slow to get that word kitty, the way you are maybe, when you hear a word in a foreign language, and then just say, I think I know, I think, yeah, I got it, right, and the whole rest of the sentence went off without you.
If you’re slow to get kitty, then bench goes off without you. If you’re quick to get kitty, you get bench for free. Nobody has to go and say, hey, look at the bench, right, because you’ve gotten it by inference. So the child who interprets familiar words more quickly can attend to unfamiliar words that follow, and thus can learn new vocabulary through inference.
So this is one of the benefits of speed in processing. It’s not just 200 milliseconds. Life’s long. Who cares? It’s that that buys you processing capacity that is really helpful in building vocabulary.
So back to where we started. The first thousand days of life are an extraordinary period of potential for learning. We need to take advantage of this. We need to provide parents with the knowledge and the skills that they need to empower them to help their children, to support their children, to nourish their children’s brain development.
Babies are born ready to learn, but only with our support will they achieve their potential.
Thank you.
Want a summary of this talk? Here it is.
SUMMARY:
Anne Fernald’s talk, titled “Why Talking To Little Kids Matters,” highlights several key points about the critical role of early language exposure and engagement in a child’s development:
1. Genes and Environment: Fernald emphasizes that genes set the stage for a child’s potential, but the realization of that potential depends on environmental support. She underscores the importance of the support a child receives from their environment.
2. Early Learning: Babies are born with a readiness to learn, and the first thousand days of a child’s life are crucial. Parents play a pivotal role in nurturing their child’s brain development during this period.
3. Achievement Gaps: Fernald points out that children living in poverty often start school already behind their more advantaged peers, resulting in persistent achievement gaps. These gaps emerge early and can have lasting consequences.
4. Language Exposure: The speaker stresses the significance of parents talking to their children, even when the child is not yet able to respond with words. Hearing language is how children learn it, and oral language skills are fundamental to intelligence.
5. Hart and Risley’s Study: Fernald discusses a study by Hart and Risley, which found significant differences in the number of words children from different socioeconomic backgrounds heard by age four. This study highlights the disparities in early language exposure.
6. Variability in Language Input: Fernald’s own research demonstrates the vast variability in the amount of language children hear in low-income families. She suggests that it’s not just socioeconomic status but also the amount of speech directed at the child that matters.
7. Processing Speed: Fernald’s team measures children’s processing speed as they hear and understand words. She shows that children who can interpret words more quickly are better equipped to learn new vocabulary through inference.
8. Hopeful Variability: Fernald finds hope in the variability in language input among low-income families, suggesting that it’s not an inherent trait of a particular group but rather the way parents talk to their children that makes the difference.
9. Building Vocabulary: Fernald explains how processing speed aids in building vocabulary by allowing children to make inferences and grasp unfamiliar words more easily.
10. Early Learning Potential: Fernald concludes by emphasizing the extraordinary potential for learning in the first thousand days of life and the need to empower parents with the knowledge and skills to support their children’s brain development.
In summary, Anne Fernald’s talk underscores the critical importance of early language exposure and engagement in a child’s development, especially in low-income families, and highlights the potential for closing achievement gaps by providing parents with the tools to support their children’s cognitive growth from the very beginning of life.
