the development of deriving manner implicatures

Grice’s Maxim of Manner entreats speakers to ‘avoid obscurity of expression’, ‘avoid ambiguity’, ‘be brief’ and ‘be orderly’ (Grice, 1975). For instance, if a speaker says, ‘that idea is not impossible’, she implies that it is not straightforwardly possible, otherwise she would have used the synonymous but briefer alternative, ‘possible’. Subsequent reformulations along cost-benefit lines include Horn’s R-based inferences (1984), Levinson’s M-Heuristic (2000), and game-theoretic approaches (e.g. Bergen, Levy and Goodman, 2016; see Rett, 2020, for a theoretical review). There is consensus, though, that a) the form of the utterance itself may be used to convey meaning, and b) a speaker may use a more prolix, or marked, form intentionally to convey a marked meaning, where there is an alternative shorter or unmarked form with the same literal meaning. The few studies with adults provide emerging evidence that hearers do sometimes derive manner implicatures (e.g. Cowan, 2021; Gotzner, Solt and Benz, 2018a, b; Wilson and Katsos, 2016). To date, manner implicatures have only been tested in children with the lexical-periphrastic causative alternation (e.g. ‘he made the door close’ vs ‘he closed the door’), which is possibly lexical rather than pragmatic (Antoniou & Katsos, 2017). Manner implicatures present an interesting case in revealing how aspects of children’s cognitive development interact with their pragmatic competence. In particular, a rich literature documents children’s development of conventionality, not only linguistically but across a range of areas such as object functions or game rules (see Diesendruck, 2011 for an overview). This may include a stage of heightened sensitivity to convention (Falkum, 2022). While both unmarked and marked forms are conventional in that they are comprehended at the literal level, marked forms are also less conventional: speakers do not expect them to be used for that meaning as much, because they are less frequent in context. Together with what we know about the developmental trajectories of other implicature types, this gives rise to two opposing hypotheses. On the first hypothesis, children reach adult-like rates of manner inferences relatively late. On the second hypothesis, children follow an inverted U-shape in rates of manner inferences. The goal of the study is to test these two hypotheses.