Acquiring the language of possibility

Overview

Our project investigates the connections between semantic variation, language change and language acquisition, in the domain of modality. Insights from formal semantics show constraints in how modality is expressed, and how it interacts with systems like tense and aspect. Insights from historical linguistics show that modal expressions evolve in constrained ways as well. This raises interesting questions for language acquisition: How do children figure out the patterns of their input language? Do they innovate, and if so, are their innovations in line with, or even responsible for, language change? By probing how children acquire modal words, and the kinds of innovations they make and fail to make, we hope to further our understanding of how modality can (and cannot) be expressed in natural language.

The way languages across the world express modality shows both variation and convergence. In a language like English, the same modal word (e.g., must) can express different flavors of modality (Kratzer 1981): “Jo must eat fish”, for instance, can express a likelihood that Jo is a fish eater (‘epistemic’ necessity) or an obligation Jo has to eat fish (‘deontic’ necessity). In a language like St’at’imcets, the same modal word can be used to express both possibility and necessity (Rullmann et al. 2008). How do children figure out the modals of their language? What expectations, if any, do they bring to this learning problem?

Some questions we are interested in:

  • How do children figure out whether their modals can express different flavors?
  • How do children figure the underlying force of their modals (possibility/necessity)?
  • How do children’s modals differ from those of their parents? What kinds of errors do children make? Are these errors of the right type to underlie the kinds of syntactic and semantic unidirectionality found in modal change?
  • What does modal change look like in synchrony?

We address these questions through a combination of: (i) cross-linguistic analyses and comparisons; (ii) behavioral experiments testing children and adults’ modal comprehension and production; (iii) corpus analyses, looking at how modals are used in speech to children across different languages, and comparing how children’s modal sentences differ from those of their parents, and change over the developmental process.

Our project is supported in part by NSF grant #BCS-1551628.




Papers

  • van Dooren, A. (2017). Dutch must more structure. Proceedings of NELS 47.
  • van Dooren, A., A. Dieuleveut, A. Cournane and V. Hacquard (2017). Learning what ‘must’ and ‘can’ must and can mean. Proceedings of the 21st Amsterdam Colloquium.
  • Hacquard, V. and A. Cournane (2016). 'Themes and variations in the expression of modality'. Proceedings of NELS 46, pp.21-42.


Groundwork