This project investigated how children learn the meanings of modal words (like can or must), words that allow us to express our hypothetical thoughts, in contrast to what we take to be fact. We asked specifically when and how children learn the ‘force’ of modals, that is, whether they express possibility (can) or necessity (must), and when and how they learn the ‘flavors’ that these modals can express (can can express an ability or a permission; must can express an obligation or an inference…). Because modals express abstract concepts, by understanding when and how children acquire their meaning, we shed new light on the kinds of conceptual, linguistic, and pragmatic abilities children deploy across development. 

We conducted several large corpus studies on natural speech usage between parents and young children (primarily in English, but also in Dutch, French, and Serbo-Croatian) and several experimental studies with adult and child participants (primarily in English, but also in Serbo-Croatian). Our results show that while children struggle with necessity modals like must or have to until age four, they understand that words like can or might express possibility (and not necessity) already by age two. Our results further argue that by age three, children distinguish different modal flavors, and that they do so on the basis of subtle distributional cues (what kinds of sentences modals are used in and in what contexts), whose exploitation requires sophisticated conceptual, pragmatic, and linguistic abilities. For instance, children must grasp the concepts underlying various kinds of modal reasoning, like how people make inferences from knowledge or perception, or how permissions and obligations relate to social expectations, and they must expect that speakers aim to be efficient communicators. Our results suggest that these abilities are largely in place by the preschool years.

We developed new ways to probe children’s modal comprehension and production, through novel experimental paradigms, and through a novel methodology, combining behavioral and corpus data, that allows us to test how adult-like children’s naturalistic modal productions are. We have made all our materials available to other researchers, to be used and adapted by others, to study more languages and populations, and more linguistic phenomena.

Broader impacts:  This project fueled interactions between linguists and psychologists, through various presentations and workshops. We further developed methodological tools to determine whether success in a linguistic task is driven by conceptual difficulty, grammatical knowledge or by pragmatic reasoning associated with using this knowledge, with implications for language delay, and disorders of language and cognition. Our project provided five graduate, twelve undergraduate, and one high school student with in-depth mentoring and research experience. It was featured in several outreach events, broadening general awareness of language and cognitive development.

Last Modified: 05/31/2022
Modified by: Valentine Hacquard