For best experience please turn on javascript and use a modern browser!
You are using a browser that is no longer supported by Microsoft. Please upgrade your browser. The site may not present itself correctly if you continue browsing.
uva.nl
SMART Cognitive Science null

How do children learn a language?

An interview with Sybren Spit

Sybren Spit is a PhD student at the Amsterdam Center for Language and Communication (ACLC). Under supervision of Enoch Aboh, Sible Andringa and Judith Rispens, he is investigating the role of awareness in early language acquisition. In this interview we discuss his career, his research interests, and the upcoming SMART workshop Awareness in Humans and Machines, which he will be co-organizing with dr. Sible Andringa.
Research Background
I am a home-grown UvA PhD student. I did my BA in Dutch Language and Culture here at the University of Amsterdam, for which I specialized in Dutch Linguistics. After that I continued with the Research Master in Linguistics, and that is how I rolled into this PhD project. My choice for a BA in Dutch Language and Culture was not initially driven by an interest in linguistics. I was rather interested in Dutch literature, and this BA would give me the chance to study this. Then, I found out that linguistics was also very interesting, and I liked it being more empirical than literary studies. Although, I still enjoy reading literature, linguistics has more appeal as an object of study to me.

From the start, my research has mostly been focused on language acquisition, something that already started in the BA programme: the first course that allowed us to conduct our own research was about language acquisition. However, I do like to see myself as a researcher with a broad view on language, and I try to have a more interdisciplinary approach. I like theoretical work and try to keep in touch with other fields such as computational linguistics. This is also why SMART is so important. During my rMA, there were SMART lectures by researchers from various disciplines; sometimes closely related to my own specialization, but sometimes very different. Nonetheless, language acquisition is the clear thread that links all of my own research, including my current PhD project.

Awareness in early language learning
My PhD is about the role of awareness in early language acquisition. We investigate whether kindergarteners, that is, children from the age of four to six, develop awareness of the grammatical rules they are learning, and whether such awareness makes a difference. Do children who are aware of their learning target learn better than children who are not? Many theories about language acquisition assume that children acquire grammar implicitly at these ages. Children would not know what they’re doing, and making them aware does not make a difference. In this PhD project we are questioning this assumption. There is not much work done on this topic, and although it might be true, we cannot know it for sure when it has not been tested.

It also is a difficult topic to investigate in children. You can ask adults to explain the rule underlying the use of a certain linguistic form, but this is a very difficult task for children. For this reason, we used a different method. We designed an artificial language which included a specific rule. Children were exposed to this language and afterwards did a test in which they heard sentences in the language accompanied by pictures. They had to choose which pictures belonged to the sentence and could do so on the basis of the grammatical rule.

In itself, this does not immediately show whether the children are aware of the rule that they learned. However, there were two types of test items in this task; easy and difficult ones. The children were able to ‘gamble’ on whether they would get the test item correct or not. They would receive a big reward for a correct answer, but would not get a reward for an incorrect one. Importantly, they could also opt-out of making a decision, in which case they received a small award. The idea was that if children are aware of what they are learning, they will choose to opt-out of difficult trials, whereas they will take the risk if sentences are easy – as a learner you can only distinguish between easy and difficult questions when you are aware of the rule. This particular paradigm is often used in animal studies with pigeons, dolphins, macaques, rats, orangutans. Obviously in these studies they do not use games in which points can be earned, but generally they use food as a reward. For example, they can get food that they like a lot, food that they are rather indifferent to, or no food at all.

So far we have already done several experiments. First, we wanted to test whether children would actually like these types of rewards, and after that we conducted the experiment in which they had to learn the rule. They fortunately were, but as mentioned before, they could not put this rule into words. We observed the behavior that I just explained: they opted-out more often for difficult trials and made a decision more often in easier trials. This is in line with the idea that they develop some awareness of the rule, and could show that not all of language acquisition is necessarily implicit.

The implications of awareness
In two follow-up experiments, we tested whether more awareness also has other implications, or more specifically, whether awareness helps children in any way to learn the rule. These experiments tested whether an explanation of this rule could help children to acquire it. We compared children who received explicit instruction about it with children who did not receive this instruction. The results for this experiment are not straight-forward. On the one hand, eye-tracking data seems to suggest that the children who received the instruction looked at a target answer earlier than the children who did not receive instruction. This could suggest that the children are aided by explicit instruction. On the other hand, when it comes to choosing the right picture, we do not have evidence to support the idea that they actually do better. That seems strange: their eye movements suggest that instructed children perform better, but we are unable to show this in their task behavior as well.

So far, we are not sure what the direct implications of rule-awareness in children are, if there are any. Currently I am running the final experiments of my PhD project, in which we conduct the same experiment but this time we test the children twice: not only directly after their exposure to the miniature language, but also the day after. This way we can see whether a possible effect of explicit instruction is only visible over longer time, which gives a better understanding of our current results.

Awareness in humans and machines
Currently I am also co-organizing the SMART Workshop “Awareness in Humans and Machines” with Sible Andringa. This workshop is a follow-up on the workshop we organized a year and a half ago about awareness in humans and animals. This previous workshop was very closely related to my current research. As I mentioned before, I got my own research design from animal research, which gave us the idea that perhaps more people can get inspired by this interdisciplinary approach. This workshop was quite successful, and the opportunity for another SMART Workshop made us think about organizing a similar interdisciplinary event, but then in a slightly different way. This is how we came up with the topic “awareness in humans and machines”.

I am really looking forward to the workshop. Our keynote speaker is Axel Cleeremans from the Université libre de Bruxelles, who will approach the topic from a more theoretical point of view, addressing questions such as “what is awareness?” and “what could awareness in machines look like?”. Then we have a couple of speakers who will be discussing the awareness in machines:  Joost Broekens from the Delft University of Technology, who is currently working on the engineering of awareness, Eelke Spaak from the Donders Institute, whose work focusses on the modelling of awareness, and Chiara de Jong, who studies the interaction between robots and children and the way children perceive ‘awareness’ in these robots. Last but not least, Padraic Monaghan from the University of Amsterdam will discuss the role of awareness during language acquisition. All in all, the programme of the workshop is very promising, and we hope many people with different backgrounds will come.