What is something amusing that you have learned

Out with the language

It seems natural to us that children will begin to speak at some point. But learning the language is an intellectual achievement that is still not fully understood today. Using a wide variety of methods, the departments of Caroline Rowland at the Max Planck Institute for Psycholinguistics in Nijmegen and Angela Friederici at the Max Planck Institute for Cognitive and Neurosciences in Leipzig explore how children seem to adopt this complex communication system with ease.

Text: Tim Schröder

The way to the laboratories of Caroline Rowland leads through a small park. If you walk along it, you will immediately notice that there is something special here: A baby blue porcelain fairy sits on the edge of the path in the grass, a few steps further a pink fairy, about the size of a Barbie doll, then another and another, up to a side entrance .

And inside, too, this part of the Max Planck Institute for Psycholinguistics looks different from the rest of the building with its offices and laboratories. There are squeaky-colored mushroom-shaped stools in the hallway, just big enough for children to sit comfortably on. “And here is our waiting room,” says Caroline Rowland and unlocks a door: a small play paradise with cuddly toys, picture books, boxes full of board and skill games and chairs. Caroline Rowland and her staff have set up everything so that children feel at home here straight away, because children are the real stars of Rowland's research.

Caroline Rowland is Professor of Psychology and Director at the Max Planck Institute for Psycholinguistics in Nijmegen, the Netherlands. She wants to find out how children learn their mother tongue - even before they have even rudimentary mastery of many other skills: “In the first few years of life, most children effortlessly learn the most complex communication system in the known universe. I want to find out how they do it and why it is so. ”And to do this, she regularly invites parents and their children to play experiments. She is particularly interested in working with babies and toddlers who are just beginning to learn their mother tongue.

Next to the colorful waiting room is one of the laboratories in which Caroline Rowland and her team do language experiments with the children. The contrast couldn't be greater: the room is sober, the walls are unadorned. Nothing should distract the children. In the middle of the room is a large monitor, in front of which is a table with two chairs. Parents and their children take a seat here. The researchers then play scenes or images on the monitor: a dog chasing a cat, objects such as balls or rubber ducks - things that the children are familiar with from their everyday lives. For example, together with Julia Egger, Christina Bergmann and Andrew Jessop from her team, Rowland observes what the children draw their attention to. She uses eye trackers, which are infrared cameras that track the movements of the pupil and iris and thus register when a child is looking where.

Using this technique, Caroline Rowland, along with staff and British university colleagues, measured the speed at which a child processes language. The “looking-while-listening” paradigm is called the “looking while listening” approach. A child sits in front of a screen and sees pairs of images, but only one of them is addressed at a time. If, for example, an apple and a car can be seen on the screen, you are asked to look at the apple at the same time. With the eye tracker, the scientists can measure how quickly the child directs their eyes to the apple. The speed at which children identify the right word shows how quickly they process the language they hear.

The challenge is to understand what is heard

The study aims to solve one of the great puzzles of language development: Why do some children speak their first words as early as eight months, while others only start speaking at two or three years old? And very practical: How can you tell at an early stage whether a child's development is only slightly delayed or whether they have a language disorder that should be treated early on? The study found that the speed at which 18-month-old toddlers process language is actually a major factor in language development.

The quick among them have a clear advantage: They can learn more from every sentence they hear, their vocabulary and even their knowledge of sentence structure grow faster than with the slower children. And the larger the vocabulary, the faster children can process what they have heard. This puts the slower children at a disadvantage. The project team would like to get to the bottom of this phenomenon in its future research - also to find ways to help disadvantaged children.

For Caroline Rowland, processing the spoken language is a central aspect of learning to speak. Because what newborns and toddlers initially perceive is a never-ending flow of single sounds and syllables. Adults are familiar with this from learning a foreign language - the flow of words is often difficult to understand, only occasionally individual familiar terms can be recognized and you can try to figure them out to make sense. “As a rule, we do not pause between individual words when we speak a sentence,” says the scientist. “For toddlers who have to learn the language again, the challenge is to recognize terms in this flow of syllables.” And not only that, they also have to differentiate between parts of speech and understand the grammar.

So a central question is how children process the linguistic input. Research has to take into account that in the spoken language the words do not come out equally as they would on an assembly line. When we talk, we use different intonations, we put pauses, we speak in a certain speech melody. We emphasize what is being said with gestures and looks, and of course we refer to things and people in the area. According to Caroline Rowland, research has so far paid too little attention to this. She sees this as a complex challenge for her field.

Children make clever mistakes when speaking

She herself approaches the challenge multi-methodically. That is, it uses a whole range of different approaches: neuroscientific methods, computer models, behavioral experiments, and detailed studies of everyday conversations.

To analyze conversations, Caroline Rowland and her team use large online databases such as CHILDES, in which dialogues from and with children are collected as written documents, audio or video files. "You can draw a lot of conclusions from the conversations stored there, from the questions and the answers given by the children," says Caroline Rowland. Typical mistakes that children make when speaking, for example, are revealing. Because they show which recurring patterns the children have already recognized in language. Younger children often make mistakes when forming the plural, such as “jumping the frogs”. Then they saw through that you can add an "e" for the plural, but they have not yet learned that there are further rules, in this case to transform the "o" into "ö".

In older children, mistakes change as language skills increase. When a preschooler says a sentence like, "Jump the ball up there", the following happens: The child has already learned words like "throw" or "push" that express how to set things in motion ("den Throw a ball"). So why not use the word "jump"? The example shows the heart of the matter: It is not a mistake in which the child uses words or parts of words in a nonsensical manner. On the contrary, the child is adopting a pattern that they have heard many times and that they are now using creatively to express something new - which in this case is not entirely correct. Rowland speaks of "clever mistakes" which, interestingly, are similar in all children with normal language development, across different languages.

Even toddlers can find patterns in language

Incidentally, in such cases, language development research recommends that children not explicitly point out their mistakes. Children are not aware of which linguistic laws they apply. Instead of preaching rules, adults should better take up the child's sentence, formulate it correctly (“Now let's let the ball bounce all the way up”) and then continue talking to the child normally. Over time, the children learn not to make the “clever mistakes” anymore.

One explanation of how even very young children can acquire linguistic patterns is so-called statistical learning. It describes the brain's ability to recognize, observe and learn from laws in a complex environment. How this works is made clear by a sentence like “Look, the duck is swimming”. As a rule, the sequence of the syllables “en” and “te” appear more frequently in everyday language than the combination of the syllable “die” and the syllable “en”. A child learns within a short time that there is a connection between “en” and “te” - the “duck”.

This is shown by experiments in which children hear an artificial language. The researchers incorporate certain regular elements into this language. In the fluently spoken fantasy sentence “dalobitaganodalobilimidenatidalobi”, for example, the sequence “dalobi” occurs more frequently. It turns out that children learn this very quickly and use this knowledge to recognize the term "dalobi" later when it is recited individually.

“With several cameras in the room, we film the children's faces in order to examine how they react to familiar and unfamiliar sequences of sounds,” explains Caroline Rowland."When the children hear a familiar term such as 'dalobi', their attention increases significantly, for example they look up."

Computers help analyze conversations