What goes on in the brain when we study in a second language?


Researchers at UGent are looking closely at how exactly we deal with information when we learn it in a language not our own, as their university and others offer more and more programmes in English

Learning in tongues

As Flemish universities reach out to international students by providing more programmes in English, there is a lingering concern that home students may be at a disadvantage. So researchers at Ghent University have been looking more closely at how people learn in a second language.

“People always assume the professors speak very poor English and the students are not proficient in English either, and that their studies will suffer tremendously as a result,” explains Robert Hartsuiker, a professor in the Department of Experimental Psychology, in very good English. “But I think we should investigate this, and if we find that it is the case, we should know why and what we should do about it.”

In order to explore this very practical issue, Hartsuiker and his colleagues Marc Brysbaert and Wouter Duyck set out to ask some fundamental questions about how people take in, process and retain information when they receive it in a second language. Meanwhile, Martin Valcke, a professor in the Department of Educational Studies, will be looking at how the results of their fundamental research carry over into the lecture hall.

The €1.5 million project is called Lemma, which is both a linguistic term and an acronym for Language, Education, and Memory in Multilingualism and Academia. It has been running for three years, with two years to go, and is beginning to produce its first scientific results.

All in the eyes

One advantage of working on this problem is that the raw material is close at hand. Ghent University students have been recruited to take part in a range of experiments that test aspects of their ability to learn in English compared to their native Dutch.

The researchers examined differences in reading in the two languages by asking students to read a novel, half in Dutch and half in English. They do this on a computer screen, while a camera tracks the way their eyes move.

People assume the students are not proficient in English either, and that their studies will suffer tremendously as a result

- Robert Hartsuiker

“This can tell you, with the precision of a letter in a word, where the eye is looking, for time intervals of around 4 milliseconds,” says Hartsuiker. The reading pattern that emerges shows how long people spend on different words and how often they go back to check a word or re-read part of a sentence.

The same book has to be used for each reading. “The tricky part is having a really accurate translation,” says Evy Woumans, a postdoctoral researcher involved in conducting the experiments. “The Dutch version has to be very equivalent to the English version.”

It also helps if the book is engaging and not too hard to read. In this case, the winning title was The Mysterious Affair at Styles (De Zaak Styles in Dutch), one of Agatha Christie’s Hercule Poirot novels.

The broad results of this work are not so surprising: people are slower to read in their second language and the pattern of going back and forth in the text resembles that of a developing reader. But this is just the beginning, since the data – the researchers call it a corpus – can be revisited with all sorts of new questions.

For example, it is known from first-language studies that people take longer to read long words than short words. But is this equally true of the second language, or is there an extra deterioration?

A question of storage

Another experiment that is currently being developed will look at the way people’s minds wander when they’re reading. It’s already known that particular patterns of brain waves can be detected when people are drifting off, so the idea is to see if these can be linked to a reader’s eye movements.

“If we can find a pattern of eye movements, we can look back at our corpus and see how much people mind-wandered while reading the novel, and what parts of the text made people’s minds wander,” says Woumans.

Alongside these experiments that look at how students take in information, there are experiments that focus on its retrieval. The most basic tests have them memorise lists of words in English and Dutch, then recall them after a week and after a month. This produces a ‘forgetting curve’.

“One possibility is that if you learn something in the second language, in the short term you remember it as well as if you had learned it in the first language, but you store it worse, maybe less specifically, so as time passes you forget more,” says Hartsuiker.

A more complex test involves getting students to read short texts on topics such as the sun, types of stars or the lives of sea otters, and then asking those questions. “There is a part which is true or false, so just facts, then a part which has open questions,” says Woumans. “And we test them right after reading the text, then after a day and after a week and see how much they remember.”

There is hardly any difference between the languages with multiple-choice questions or those that involve recognising something. But when more is required, a gap opens up. “If you have to explain something about the behaviour of sea otters, then you find it is more difficult in the second language after time has passed,” says Hartsuiker.

Telling stories

The next step is to find out where in the process the difficulty lies. It could be in finding the concept to answer the question, finding the words or formulating them into an answer, or in making the physical movements in order to speak the words. “In principle, all these components could be more difficult in the second language than in the first,” says Hartsuiker.

If you have to explain something about sea otters, you find it is more difficult in the second language after time has passed

- Robert Hartsuiker

Experiments that involve recall of a word but not speech – for instance asking the student to press a button if the name of an object contains a particular letter – indicate that recalling concepts and words are much the same in either language. “This suggests that the delay occurs in the later stages, perhaps in the articulation of the word or the speech motor commands,” says Hartsuiker.

The conversion of memories into speech is also being tested by asking students to tell a story several times over. It is normal for the story to become shorter with each retelling, but the researchers want to see if the degree of compression varies when students retell stories in their first and second languages.

Then there is an experiment to see how far ahead people plan sentences in each language. This is done by showing them two images on a screen and asking them to form a sentence that says the object on the left is next to the object on the right. But at the same time a disrupting word, related in meaning to the name of the first or second object, is fed to them through headphones.

“If the participants plan the complete sentence in advance, then a relationship to the second word should slow down their initiation of the sentence. But if they plan just a bit of the sentence and produce the rest as they go along, then a relationship to the second part of the sentence should not really matter,” Hartsuiker explains.

Different language, different outcome

The students involved in all these experiments are rewarded for taking part, either with course credits if they are part of the departments concerned, or money if they are from elsewhere in the university. Afterwards they are told what the experiment was about, and they can access the results if they are interested.

Work on practical support for those taking courses in English will come to the fore in the final years of the project. “One intervention could be in the form of advance organisers, which could be anything that provides structure to a text you are about to read,” says Hartsuiker.

That could be an outline, bullet points or even a simple word list. Tools might also be developed to provide vocabulary support, if that proves to be an issue.

And while this work may also help international students coming to Flanders to study in English, the basic research does not necessarily read across. “If you have totally different languages, with different syntactic structures, then maybe the outcome is totally different,” says Woumans.

“We hope that that some of the conclusions would be valid for them as well, but that requires further research,” Hartsuiker adds. “It would be quite interesting to focus on Chinese or something that is quite different.”

Photo: UGent/ChristopheVanderEecken

Ghent University

Ghent University (UGent) is one of Flanders’ most pluralistic and liberal institutions of higher education, and its motto has long been “dare to think”. UGent is renowned for its research in bio and life sciences.
Latin - UGent was originally founded as a Latin-speaking state university by the Dutch king William I.
Nobel - Corneel Heymans, the only Fleming to have won a Nobel Prize, studied at the university.
Autonomy - UGent is the largest employer in East Flanders.

million euros in annual revenue

1 882

first female student admitted

1 930

Dutch becomes university’s official language