There has been robust evidence that labelling affects infant categorisation, but mechanisms and processes underlying this interaction between word learning and categorising objects remain obscure.
Previous work has presented advantages for categorisation in the presence of labels, even for very young infants between 3 and 4 months, but research has not so far explained conclusively why exactly labelling helps. In particular, studies have typically contrasted successful learning of a single category with labels with unsuccessful learning in a non-labelling context. This makes it difficult to pinpoint the locus of the benefit exactly, as it is not clear what fails when learning is unsuccessful. I will present a series of eye tracking studies that aim to shed light on this through a combination of carefully tracking infants’ attention directed at individual features, and using labels contrastively (i.e. for two different categories) so that differences in what is being learned can be measured. Together, these experiments provide novel insights into the relationship between language and concepts in early childhood.