When we cast a light on one piece of the world, we extinguish everything on which we have not cast that light. This is how we create the things we say: to speak is to create a piece of the world; it is to mould it, to make it, and to make it live.
Language is what separates us from other animals and brings us closer to other individuals of our species. We learn language before we are born and refine it throughout our lives. The power of language is such that from a finite number of elements we can reach an infinite number of meanings. Language makes us eternal, allowing us to communicate with our contemporaries, with people who no longer exist and with people who have not yet arrived. Language is fascinating, which is why it is important to understand how we acquire it.
To better understand language acquisition, we talked to Caroline Rowland, head of the Department of Language Development at the Max Planck Institute for Psycholinguistics in Nijmegen, in the Netherlands. Parallel to her research on language acquisition in children, Caroline holds a position at Radboud University.
Juan García Ruiz: What’s so special about human language?
Caroline Rowland: You will see lots of theories that suggest there is one big thing that distinguishes human language from animal communication. There are theories that suggest that it is our ability to understand other people’s intentions - our ability to get into other people’s heads - that allows us to communicate. Other people will argue that it is our ability to use complex grammar. The reality is that there are a lot of differences between human brains and the brains of all other animals, not just ‘one thing’. What’s special about human language is its power. Human language is much more powerful and flexible than any other animal system of communication. And multiple changes must have happened in evolution to give us this power. One obvious difference between language and animal communication systems is grammar. We have the ability to put together a finite number of items into an infinite number of meanings, we can potentially create an infinite number of sentences because we have this grammar system. So for example I can say the dog is red, or the big dog is red, or the big dog that I saw in the park is red, or even the big dog that my neighbour owns that I saw in the park is red. I could go on forever, and there is no other animal communication system that has this power. But I think there is something else. For some reason we can talk about much more abstract concepts. As far as I am aware, there aren’t animal communication systems that have the ability to talk about truth, justice, love, desire or even thinking itself. So grammar and abstract concepts are just two ways in which human language is much more powerful than other animals’ communication.
JGR: As for human language, what’s your opinion on the “nature vs nurture” debate?
CR: There is no gene for the language. There are many tiny adaptations to our genetic code that have enabled us to learn language. In terms of the nature-nurture debate, there are many genetic mutations that have made the difference between animals and us. Therefore of course, there is something innate about human language, but we still don’t know what that is.
JGR: Scientist have tried to teach animals to communicate like we do. Have they succeed?
CR: There’s been a lot of attempts to teach animals to communicate in the same way that humans do. Because very few animals have the same vocal tract, most of the experiments with apes have used sign language and gestures. How successful they were depend on how you define success. Some apes can learn a lot of gestures and how to combine them. But they never ever reach more than the ability of maybe a 2-year-old child. Plus, they require a huge amount of explicit training to do that, whereas a 2-year-old child exceeds all the abilities of an ape with very little effort.
JGR: What exactly is your current research about?
CR: In the language development department we do quite a lot of different things. We have researchers looking at the role of eye gaze and how it seems to enhance language learning. We have people who are working on how children learn a range of different languages and how they adapt to the very different language systems that exist around the world. At the moment I am fascinated by the question of individual differences. Why do some children learn language well and quickly, and why do other children take longer? The answer is not simple. It’s not that some children just hear more speech than others. And it is also not as simple as saying that some children are just set up for learning faster. The differences are too large for it to be that simple. Even if you simply count the number of words that children know in their language, you can see huge differences. For example, in the Stanford Wordbank database, at the age of 18 months, some children seem to know hundreds of words while other children of the same age have maybe 5, 10 or 15 words. So you get these huge differences in the first few years in how quickly children learn language, and we don’t know why that is. Most of these children will eventually be perfectly fine, even those knowing very few words initially will catch up with their peers. But we don’t know why they differ on how quickly they learn their language. We don’t know why some of these children that are very slow eventually catch up while others don’t.
JGR: We have no clue at the moment of what could explain those differences?
CR: We’ve recently finished a large project called The Language 0-5 Project where we followed 80-90 children intensively for the first five years of their life. We created a range of different studies to try to figure out what kinds of knowledge, behaviour or mechanisms would produce these differences in their trajectory. One thing that we discovered was that children differ in how quickly and easily they learn from the statistics of language. What I mean by that is that language has a pattern, a sort of a predictable distribution of sounds. Some sounds follow other sounds very often. For instance in the word baby, the sound ba (/beɪ/) is followed by the sound by (/bi/), and if you hear baby a lot, you can use this pattern to figure out that this segment is actually a separate word. Children are really good at picking up on those patterns in order to realize that baby is a word, to pull that out. What we found is that children who are very good at this segmentation or distributional learning do seem to develop language a little bit faster than others. But none of the effects that we found are very strong. They can’t explain all the big differences we find between children. So again, you are not finding that the cause is just one thing that predicts language, you are finding a range of lots of different effects that interact to build up this complex picture of what is it about some children’s brains that allow them to learn faster than others.
JGR: You co-direct the International Centre for Language and Communicative Development, as well as the Max Planck Language Development Department, where I guess you run the research you described. How do you study language in the lab?
CR: Studying language acquisition is very challenging. When you are studying adults you can tell them what to do, you can explain them what the point of the experiment is. You cannot do that with a child, you cannot tell a 2-year-old that you want them to produce complex sentences or that you want them to describe pictures. So we have to come up with different methods to study child language, and we have to make them fun. In a way we have become expert children’s games creators. If we are interested in what children say, we can use a range of techniques to try and get children to repeat sentences after us to see whether they can produce them. We also need to know what children understand about language, so not what they say but what they know. A simple way to do that is to use an eye tracker, which exploits the fact that when you hear a sentence you tend to look automatically at what that sentence describes. If I say to you look at the dog, you will automatically look at a dog, ignoring a cat, for example. We can exploit that fact by sitting children in front of an eye tracker which uses an infrared beam to track where they are looking very precisely. So you can just present children with sentences like oh, look at the dog! and then you just track where their eyes look. That tells you whether they understand the sentence. The other thing that you can do really nicely with children is to get electroencephalogram signals from the brain, so you can see how the brain is reacting to different linguistic stimuli at different ages. In this case it’s usually language understanding that we’re interested in. You can see for instance if brain signals indicate that children are surprised when a sentence unravels unexpectedly. So if you say the pirate hid the dog instead of the pirate hid the treasure and they understand it, they should be surprised. It always has to be fun, you cannot build a boring experiment and ask a child to complete it; this won’t work!
JGR: Do you use also molecular, cellular or tissular approaches; or do you study in some other way the neural basis of language?
CR: No, we don’t. My colleagues in the Language and Genetics department here at the Max Planck Institute, they are really interested in looking at some of the genes underlying language, for example the gene FoxP2 and particularly the effect FoxP2 has on development of the brain. Not in humans necessarily but in a number of different animal models. So those approaches are used here in the Max Planck Institute, but not in my department.
JGR: You study language acquisition in children. To which extent the way children learn to communicate is different than an adult’s way to learn a new language?
CR: Children learn languages so much more efficiently than adults. Why that is? We don’t know. Probably a combination of the plasticity of their brains and the fact that they haven’t already acquired another language that is competing with the new one. Children can learn more than one language at a time, they can be bilingual or trilingual pretty effortlessly, so learning lots of languages at the same time as a child isn’t difficult. But learning a second language in adulthood when we already have years and years of experience of learning one language is tricky. Something that is almost impossible for adults to learn like a native is the accurate production of speech sounds. For instance in my case I have managed to learn 1000s of words in Dutch since I moved here, but there are a lot of vowel sounds that I will never be able to pronounce correctly.
JGR: Do all children follow the same paths in language learning?
CR: No, they don’t. What’s really interesting is how children from different cultures and languages adapt differently. Even though they follow different paths, most of them get functional language skills eventually. For example Cantonese is a tonal language and is very different from English. Similarly, there are some Australian languages which can express a huge amount of information with a single word; the same amount of information we need a whole sentence to express in English. Despite these differences, children find all these languages equally easy to learn. Then in some cultures adults don’t interact with children in the same way that we do in the West. For instance there is very little sitting down, playing, talking and engaging with children. Many researchers think this kind of interaction is crucial to learn language. Yet children in other cultures seem to learn languages perfectly well with much less interaction, or with different kinds of interaction. I think that what this tells us is that children are massively adaptable; they can learn from anything and everything. Discovering the power of children to learn from whatever the environment they are in, that’s the exciting thing about doing this research.
JGR: What’s the impact of COVID-19 pandemic on language acquisition in children? Especially, what has been the effect of the different lockdowns we have been through?
CR: There is a big study that hopefully will be published soon and which look into the impact of the pandemic on children’s language across eleven or twelve different countries in the world. The findings are not simple. In some countries language development in children seem to have slowed down, in other countries it sped up and in some others it hasn’t changed. We are not quite sure why. But one thing that is very clear – and this is the thing every government need to look out for - is this: the COVID-19 pandemic has increased the social-economic differences that we see in children’s development. Schools and nursery schools are big levellers, because on the whole they give all children a good quality education. The consequence of lockdown has been that children spend a lot more time at home. This is great when they have parents who can both take time off work, or who can spend a lot of time with children them; parents who have flexible jobs so they can work around the children schedules. It’s hard on parents, but it means that the children don’t fall as far behind in their development. In addition, there are children who have nice homes, with quiet places to do their online schooling. But think about the children whose parents can’t take this time; who are working 12-15 hours a day, as a nurse for example or in intense care unit, or who live in an overcrowded house where they don’t have the space to do the work. Those are the children whose development is going to suffer. So you may see an increase in the differences between rich and poor kids across the world. This issue definitely needs to be addressed. Then there is something else. What matters for language development is interacting with other people. So there is also the worry that children are missing out on learning how to interact with other children. There are some skills that your parents don’t teach you – that you only learn by interacting with other children. For example learning how to give or how to negotiate. Your parents are not going to negotiate with you, they are going to give you the chocolate bar or not. There are definitely social skills that can be compromised.
JGR: Well established-knowledge is not that easy to produce. I guess in the case of language acquisition, because we study it since long, we do have some settled knowledge. What do we know about language acquisition so far?
CR: There are five things that we know. First thing that we know is that genetics matter. Complex interactions between many many genes builds a language-ready brain. Second, your parents matter. Your homelife and interactions in your family they really matter. It’s important to have a rich language-environment. Third, your wider social-cultural environment matters, so the society and the culture you live in, that makes the difference. Fourth, language learning starts a lot earlier than people think. Children are starting to recognize voices in the womb before they are born. They have learnt most of the sounds of their language way before they learn to talk. So even if a child is not talking, they are still learning a lot about language. Fifth, the language system is very complicated, but it doesn’t mean that the learning mechanisms in the brain are complicated. For instance, the segmentation that I mentioned before is actually a very simple kind of learning and it is present in all mammals. So the brain doesn’t need to have complex systems, but just lots of simple systems that interact in a complicated way.
JGR: How relevant is eye gaze for children to learn to communicate?
CR: Eye gaze seems to be a particularly special key. When you’re talking, engaging and looking at children, they seem to pay attention and it somehow enhances their learning. They don’t learn as well from screens; e.g. from passively watching TV. It seems to be something about that giving forth of an interaction that matters. However – and it’s a big however – we don’t yet know how well babies learn from the kind of digital interactions we are all having now; through live videos. Babies now often interact with people through Zoom or Skype, we know this from the pandemic. But what they learn from these interactions we still don’t know.
JGR: As an adult learning French, I carry with me a notebook where I write down everything I learn. Can you give the readers and me some insights about vocabulary acquisition?
CR: With children, vocabulary acquisition changes as they get older. Early on, when children are really at the beginning of their word learning, they seem to need a lot of repetitions to learn a word. Learning each word is slow. For instance they have to hear mommy many, many times before they learn the word mommy and before they learn what mommy is. But as soon as they get more and more words in their head, that seems to speed things up and they need fewer repetitions to learn new ones. There are lots of possible reasons why that might be, but the point is that having knowledge in your head enables you to then bootstrap into new knowledge faster. The more words you know, the more words in a sentence you know. Let’s say you hear the sentence the dog is chasing the cat. If you know the word dog and the word cat, you are much more likely to identify the meaning of the verb chase, whereas if you don’t know dog and cat, then you may have to hear chase a lot to figure out what it means. I think that also works in second language learning. So your vocabulary acquisition should be more efficient as you learn more and more words.
JGR: Do you have a piece of advice to get started publishing in scientific journals?
CR: It depends on what stage you are. Let’s say you are a PhD student publishing your first paper. In this case, listen to your supervisor because they know how to publish a paper. If you are at the end of you PhD or you are thinking about your final paper, you might know more about the subject than your supervisor, so that might be the time when you want to stick to your guns. But I think my ultimate advice would be: carefully choose your friends, supervisors, colleagues and collaborators and eventually your team if people work for you. Choose them all carefully. Because it’s the people around you that really make the difference.
JGR: What’s the best advice you got from someone?
CR: My PhD supervisor was excellent in the sense that he was a genuine mentor. He let me do the research I wanted to do but he guided me. He wasn’t interested in getting out a thousand publications. He was genuinely interested in finding out the answers to the questions. That’s what he gave me. Everything I do is to simply try to figure out how children learn language, and that’s why I am still fascinated by my job 20 years later. Because everything I have done is to find that out, and I still want to know the answer.
JGR: You wrote the book Understanding Child Language Acquisition in 2013. Can you give the readers a little teaser?
CR: At the time I was working on grammar acquisition, and I wanted to know a lot more about language in general. So when the publishers approached me to write this book, I said yes. I wanted an excuse to read a lot more about research across the whole of language acquisition. However, unfortunately, I discovered that there is too much for one person to know; there are too many published papers for any one person to read. So instead what I have done with this book is trying to give a flavour of what are the exciting questions, the exciting findings and what we still need to know about how children learn the most complex communication system in the known universe.
JGR: Talking about books, would you recommend a book for the readers?
CR: I’d recommend Everything Your Baby Would Ask: If Only He or She Could Talk. It’s out of print but used copies are available. Annette Karmiloff-Smith was the first researcher to truly take seriously the fact that babies are constantly and continually learning and developing. If we are to have any chance of explaining child language development, we need to bear this in mind: a baby’s knowledge of the world changes from minute to minute, as they learn more and more about the world in a continually-developing, dynamic spiral of learning and development. They’re little everyday miracles.
JGR: Do you have a message you would like to share with the readers?
CR: My advice is going to be for parents. There is a lot of pressure to be the perfect parent. Don’t worry too much about it. Enjoy being a parent and enjoy being with your children. Your children’s brain are so flexible and adaptable that they will learn as long as you give them the opportunity to learn.