In Search of L1 Evidence for Diachronic Reanalysis: Mapping Modal Verbs

Learning modals: A grammatical perspective

“…Major questions in modal flavour development from a mapping perspective include: When and how do children work out the flavour(s) of modals in their input languages (Cournane, 2014, 2015; Papafragou, 1998; van Dooren et al, 2017)? How does syntactic category affect modal flavour learning (Bassano, 1996; Cournane, 2014, 2015; O'Neill & Atance, 2000; Shatz & Wilcox, 1991; Veselinović & Cournane, 2020)? How might input frequency and distributional factors help or hinder modal flavour learning (Cournane & Pérez‐Leroux, 2020; van Dooren et al, 2017, van Dooren, Tulling, Cournane, & Hacquard, 2019)?…”

“…While maybe , think and could all can express the same concept (epistemic possibility), they are grammatically distinct and show different syntactic distributions corresponding with their category, and syntactic bootstrapping is especially important for unobservable meanings like for modality (Gleitman, Cassidy, Nappa, Papafragou, & Trueswell, 2005; Papafragou, Cassidy, & Gleitman, 2007). Learning adverbs that are adjunctival in the syntax, like maybe or probably (Cournane, submitted, 2014, 2015; O'Neill & Atance, 2000), is a grammatically different task from learning attitude verbs like think or want (de Villiers, 2007; Hacquard & Lidz, 2018), or functional modals like could that differ in interpretation depending on their grammatical and situational context (van Dooren et al, 2017).…”

“…Major questions in modal form acquisition include: How do modals' syntactic properties affect acquisition paths (Bassano, 1996; Cournane, 2014, 2015, submitted; Modyanova et al, 2010; Veselinović, 2019)? How do distributional facts in the input, related to the syntax of particular modal categories, affect the availability of cues for modal meanings (Dieuleveut et al, 2019; van Dooren et al, 2017, 2019)?…”

“…Why is there this ‘Epistemic Gap’ in production (Cournane, 2014, 2015, submitted)? The most common and longstanding proposal is that root meanings are conceptually available prior to epistemic ones.…”

“…Assuming that one‐to‐one form‐meaning relations are easiest to learn (Clark, 1993, i.a. ), modals contravene this simplest relation in both directions, with one‐to‐many mapping and many‐to‐one mappings (Cournane, 2014; van Dooren, Dieuleveut, Cournane, & Hacquard, 2017). The system as a whole becomes especially relevant for solving the mapping problem when we consider the cross‐linguistic typologies of what modals and modal systems can be like.…”