How children acquire adjectives: Evidence from three eye-tracking studies on Italian

While the literature on early language acquisition has mainly focused on nouns and verbs, studies on adjectives are comparatively scarce and have thus far provided contrasting evidence on the timing and mode of acquisition of adjectival meaning in children younger than 4. This thesis presents three eye-tracking studies exploring the online processing of adjectives by children (2;4 – 5;3 years) in comparison to adult controls, providing insights on how linguistic information and visual context interact during real-time comprehension. Experiment 1 investigated potential differences in the interpretation of three classes of adjectives, intersective (e.g., red), relative (e.g., big) and absolute (e.g., full). 38 Italian monolingual children (2;4 – 5;3) were tested in a visual-world task, where they listened to noun-adjective combinations in a four-picture scenario. Results showed that children as young as 28 months were slower than adults when interpreting noun-adjective combinations, while their looking pattern in the interpretation process was essentially the same. Furthermore, the computation of intersective adjectives was faster than that of absolute and, especially, relative adjectives, showing that children are sensitive to the different ways in which each adjective class is interpreted within different contexts. In Experiment 2 the complexity of Experiment 1 was reduced by lowering the processing load associated with both the different semantic classes of adjectives and the four-picture scenario. 28 Italian-native children (2;4 – 5;2) were presented with a two-picture display while listening to nouns combined with color-adjectives. In Experiment 2, the visual conditions varied according to the informativeness of the noun or the adjective with respect to the target referent. Results showed evidence that adjective processing develops over time. When the computation of the noun was insufficient and the integration with the adjective was necessary to resolve reference, the youngest children, unlike 3- and 4-year-olds, failed to interpret adjectives correctly and, consequently, task resolution. Experiment 3 investigated children’s incremental processing of prenominal adjectives and their ability to predict the following noun based on the lexical meaning of the adjective. 39 Italian children (2;4 – 5;3) were tested in the online 3 interpretation of Italian predicative yes/no questions (e.g., È morbido il cuscino?, lit. ‘Is soft the pillow?’) while looking at two pictures on the screen. Here, the informativeness of the adjective was manipulated. Results showed that, when informative (e.g., soft, upon looking at a pillow and a bone), the adjective was processed incrementally, i.e., before the noun was heard, indicating that children as young as 28 months are able to predict the upcoming noun based on adjective meaning. Furthermore, children were successful independently of whether the interpretation of the adjective required world knowledge (e.g., being soft for a pillow) or the exploration of the visual scene (e.g., being open for a window). Taken together, the three studies provided compelling evidence of a continuous process in children’s development of sophisticated, adult-like processing skills. By means of eye-tracking, we were able to reveal that the overall difference between children and adults is mostly attributable to toddlers younger that 36 months of age, whose processing skills are still limited when it comes to the meaning computation of noun-adjective combinations. However, from the age of 3, children’s processing abilities improve rapidly and, within a few months, they become successful parsers.