2018 Friday Poster 6675

Friday, November 2, 2018 | Poster Session I, Metcalf Small | 3pm

Stages in acquiring distributive markers in Serbian and Dutch – evidence from an act-out task
A. Bosnic, J. Spenader

Distributive markers across languages can be Distributive-key quantifiers (DK) and Distributive-share (DS) markers. DK quantifiers attach to the distributor set in a sentence (1b/2b), while DS markers attach to what gets distributed (1c)[1]. Beyond this basic understanding, the developmental path for languages with both DK and DS markers is still underresearched. Our study presents novel data using the same Act-out task with children from Dutch, a language with only DK quantifiers, and Serbian, a language with both markers.

To correctly characterize the acquisition path of distributive markers, we have to examine when they apply, but also when they are dispreferred. It has been frequently shown that children initially are insensitive to the distributive requirements of distributive markers, allowing them with collective situations[2][3], and preferring distributive interpretations with distributively unmarked sentences[4][5] (also in Dutch[2] and Serbian[6]). However, questions remain regarding the exact influence of DS markers on the acquisition of DK markers, and how this compares to DK-only languages. Using an Act-out task we were able to elicit a range of interpretations that children have for sentences with and without distributive markers.

Method: 75 Serbian children (three age groups: I-MA: 7;6 ; II-MA: 8;8 ; III-MA: 9;6) and 30 Dutch children (MA: 8;1) were asked to take a preferred number of objects from a box and distribute them to three paperdolls. Additionally, 8 Serbian and 8 Dutch adults were tested as controls. Test sentences are shown in (1) and (2), illustrating three conditions in Serbian (null, svaki, po) and two in Dutch (null, elke), each having four observations.

Results: Three main response types are shown in Figure 1 with adult-like responses. See Figure 2 for the proportion of distributive responses and statistical analysis (mixed-effects models), and Tables 1 and 2 for collective and simple distributive responses. There are several differences between Serbian and Dutch interpretations.

Conclusions: First, we discovered (i) a new (intermediate) type of response (simple distributive) in both Serbian and Dutch, especially dominant among 8-year-old Serbian children (Table 2, blue). Although incorrect, this response is also found with distributively marked sentences in Serbian (1b/c) but not Dutch (2b) (Table 2, red), suggesting confusion as to the meaning of distributive markers. This stage suggests a persistent preference for distributivity in children even after they have realized unmarked sentences generally signal collectivity. (ii) Dutch children acquire the meaning of the DK quantifier elke much earlier than Serbian children acquire svaki. Serbian children may be late due to the presence of the additional marker po, which is acquired significantly later than svaki (β= -2.08, z= -4.66, p<0.00001). Finally, (iii) the adult-like collective stage emerges simultaneously with the correct distributive interpretation of svaki in Serbian, while it is still not adult-like for Dutch with the correct interpretation of elke (cf. Table 1). We propose that the existence of two distributive markers in Serbian facilitates the calculation of the implicature that distributively unmarked sentences yield collective readings[7].

References

[1] Choe, J. W. (1987). Anti-quantifiers and a theory of distributivity. [2] de Koster et al. (2018) Are children’s overly distributive interpretations and spreading errors related? in BUCLD 42. [3] Drozd, K. F. (2001). Children’s weak interpretations of universally quantified questions, chapter in “Language  Acquisition and Conceptual Development”.[4] Brooks, P. J., & Braine, M. D. (1996). What do children know about the universal quantifiers all and each?. [5] Syrett, K., & Musolino, J. (2013). Collectivity, distributivity, and the interpretation of plural numerical expressions in child and adult language. Language acquisition.[6] Knežević, N. (2015). Numerals and distributivity in Serbian: at the syntax-semantics-acquisition interface. PhD Thesis.[7] Dotlačil, J. (2010) Anaphora and distributivity. A study of same, different, reciprocals and others. PhD Dissertation.