Research Questions and Answers

NEW!! Babylab Fun facts

When do infants first understand/produce words?
Are infants aware of the connections between words like 'cat' and 'dog'?
When do children know about the sounds that make up the words they know?
When do children use the order of words in a sentence in learning the meanings of new words?
How do infants form categories?
What information do we use to identify visual objects? Do we process objects in a different way when the object we are looking at is named?

Information about previous research in the Oxford Babylab can be found here.

Some of our studies investigate the age at which understanding of words begins. Infants at the end of their first year of life – at around 10 months, were tested. We were interested to know first if babies associate (a picture of) an object, say “cat”, with the word, when presented to them in the lab; and, if so, whether they are able to generalize understanding of that word to a picture (of a cat) they have never seen before. Ten-month-olds in this study exhibited surprising comprehension abilities: not only were they able to associate a word with the correct picture without being trained with that pairing, but they managed to do so with pictures they have never seen before in their lives! These findings offer evidence for a very early start of language development and give us exciting research possibilities for the future. They also point out to a rather straightforward suggestion for parents: Keep talking to your baby, even if he/she cannot answer back!

For more information contact, Loukia Taxitari or look at the results of our research using The Oxford Communicative Development Inventory

Our studies investigating connections between infants’ words have given us complicated, but exciting results so far. Infants heard a word while waiting for pictures to appear in an on-screen looking game. We expected our babies would show better performance when the word-word relationship was meaningful to them. To our surprise, at 18- and 24-months-of-age, the preceding word didn’t alter their performance in the looking game. We suspect, however, this is because they were already quite good at the looking game, and they waited a two-and-a-half seconds between words (a long time in language processing). In our younger babies (18 months), we did find a sensitivity to word-picture relationships like cat-dog and hat-shoe before the game had started, but this effect was no longer visible by 24-months-of-age. This result suggests that relationships between known words develop quite early, but are quite hard to see. A quicker, more difficult version of the looking game is underway, which we expect will show us these differences more clearly.

For more information, contact Suzy Styles.

Some of the studies you've attended, included research on the amount of information infants pay to the sounds of the words they are now learning. It has been suggested that young infants do not pay much attention to the sounds of words, because of the difficulties involved in learning new words. To test this, we asked whether our infants were sensitive to mispronunciations of highly familiar words such as book, by showing them a picture of a book and calling it a 'bick'. We have found that infants at 15, 18 and 24-months of age pay careful attention to the vowels in the words they are familiar with by showing sensitivity to mispronunciations of these words.

Similar results were found when we taught the infants new words such as 'dag' and 'mot' and mispronounced these words. However, we are currently re-running the latter studies at all age-groups due to some complications with the results. So we'll get back to you with more information on these studies in a few months!

For more information on any of these studies, contact Nivi Mani.

In a different study we were interested in whether children of 2 years were able to understand the role of words at different parts of a sentence. We presented them with brief video clips and sentences such as ‘the boy is gorping the girl’. Although we found some evidence that children were able to use word order to understand who was doing what to whom, by looking at the correct video when two were shown side by side, we will soon be running these experiments again in order to verify the results of our earlier studies. For more information, contact Nivi Mani. 

In another series of studies we looked at how pre-linguistic babies form categories of objects based on the co-occurrence of features. An example of this is the picture on the above: we systematically changed the sizes of neck, legs, tail and separation of ears, creating one or two categories of made-up animals and associating them with one or two made-up words. We found that 10-month-old infants are already able to form categorical representations based on the covariance of features. Additionally the sound of made-up words facilitates or hinders the categorization processes under specific circumstances.

For more information, contact Jon-Fan Hu.

Do infants process language input in a different way when pair of pictures are related than when they are not?

At present, we know that infants associate familiar words with familiar objects. In our studies, infants are typically presented with two items -target and distracter- and are asked to look at one of them (Look at the 'X'). However, the effect of the categorical relationship between target and distracter objects on early word-object associations has not been evaluated in previous research. We are exploring this right now. Do infants find the target object with the same ease when presented with two items from the same category (e.g., shoe-sock) than when they are presented with two items from different categories (e.g., shoe-biscuit). We are studying this at 12, 16 and 21 months of age.

For more information, contact Natalia Arias

So far we found out that adults are amazingly consistent when looking at familiar objects. When presented with pictures of animals they always attend to head first (eyes in particular) before looking at some other parts of the picture, say tail in a cat or udder in a caw. We are about to start to look when do babies start looking at the objects the same way adults do.

Furthermore, if one animal is named just before two images are simultaneously presented on a computer screen, adults always attend to the correct (named animal). This result supports pervious findings saying that by the time we make a first saccade (eye-movement) we already know what the picture is. So, the picture recognition happens in less then 200ms! 

For more information contact Vanja Vucetic.