2018 Friday Session A 1130
Friday, November 2, 2018 | Session A, East Balcony | 11:30am
Slovenian Preschoolers Derive Exact Interpretations of Singular and Dual Forms
L. Marušič, A. Saksida, D. Skordos, J. Sullivan, R. Žaucer, D. Barner
What is the origin of the exact cardinal meanings represented by numerals like one and two? By some accounts, number words are lexically exact and are acquired from a hypothesis space of exact logical meanings. Others argue that number words are not lexically exact, but get their exact interpretation via pragmatic exhaustification (scalar implicature). Here, while remaining neutral to the nature of the lexical representations of adult speakers, we investigate how children might acquire exact meanings for early number words, and whether these might be learned via a process that involves implicature.
One piece of evidence in favor of the hypothesis that number words aren’t constructed via implicature is that whereas children treat numbers as exact from the age of 2, they don’t compute implicatures for other quantificational expressions at that age, suggesting that implicature is not available as a possible mechanism.1-3 For example, and critical to the current study, English-speaking children do not compute implicatures when interpreting singular NPs, like in (1).3 By most accounts, adults exhaustify singular NPs via scalar implicature in order to derive the “only one” reading in (2).
- There is a cookie on the
- There is a cookie, but not more than one, on the
By some accounts, this difference between quantifiers and number words does not rule out a role for implicature in deriving exact meanings.4 For example, children may selectively compute implicatures for numbers because they can readily identify relevant scalar alternatives (i.e., larger numbers), whereas for other scales (a/some) they struggle to identify which alternatives are relevant.5 To explore this idea, we tested a unique case study that closely resembles the case of number words, but is typically thought to involve implicature in adults.6-7 Specifically, we tested how young children interpret singular, plural, and dual morphology in Central Slovenian, and whether the presence of the dual allows children to derive exact interpretations via implicature early in acquisition. In Experiment 1, we asked 4- and 5-year-olds in Ljubljana, Slovenia to judge whether singular, dual, and plural expressions were true of sets with 1, 2, or 4 objects (Fig.1). Another group of 4- and 5-year-olds, was shown the same pictures, but asked to judge the Slovenian words for “one”, “two”, and “four”. Experiment 2 used the same procedures with Slovenian 2-year-olds.
In Exp.1, we found (Fig.1a) that older children computed implicatures robustly, resulting in “exact” interpretations of both singular and dual forms, and a clear “at least 3” reading of the plural, such that there was no significant difference between number words (one/two) and morphology (Sg/Du). This contrasts starkly with past reports of English-speaking children on similar tests of singular/plural interpretation. In Exp.2, we found (Fig.1b) that even some 2-year-olds computed implicatures for Sg/Du/Pl expressions. Together, these results suggest that children as young as 2 can compute scalar implicatures to derive exact interpretations of quantificational forms like the singular and dual, lending credence to the hypothesis that implicature might support the acquisition of exact number word meanings.