2018 Sat Poster 6506

Saturday, November 3, 2018 | Poster Session II, Metcalf Small | 3:15pm

Acquisition of unaccusativity in toddlers
Z. Wang, X. Yang, R. Shi

Unaccusatives (e.g. fall) are assumed to differ from unergatives (e.g. cry) in their underlying syntactic structure as they involve A-movement, as shown in (1) (e.g. Burzio, 1986; Van Valin, 1990; Perlmutter, 1978). Studies on acquisition of unaccusativity (e.g. Babyonyshev et al., 2001; Friedmann & Costa, 2011; Randall et al., 2004; Snyder et al., 1995; Venice & Guasti, 2015) suggest that five-year-olds show understanding of the syntactic differences between the two structures by using the right morphological markers. To further investigate unaccusativity in early grammar and contribute cross-linguistic data, the current study examines the comprehension of unaccusativity in Mandarin-learning toddlers adopting the preferential looking paradigm.

Unlike languages that morphologically differentiate unaccusatives from unergatives, Mandarin Chinese does not overtly mark the distinction. One syntactic diagnostics is word order, interwoven with the definiteness effect (e.g. Huang, 1987; Belletti, 1988; Li, 2009; Pan & Han 2005). For unaccusative verbs such as lai ‘come’ and unergative verbs such as xiao ‘laugh’, the SV order is grammatical regardless of whether the subject is definite or indefinite (cf. (2a, b)). However, for unaccusative verbs, the logical object which is indefinite can stay in the base-generated object position, yielding a fully grammatical sentence with a VS order (cf. (3a)) while a similar structure is ungrammatical for unergative verbs (cf. (3b)). We inquired whether toddlers are able to discriminate the two types of verbs.

Participants were Mandarin-learning 20-month-old infants (mean age: 1;8;3, range: 1;6;30-1;8;24). Test stimuli contained two types: grammatical sentences with unaccusative verbs lai ‘come’/ diao ‘fall’ and ungrammatical sentences with unergative verbs ku ‘cry’/ xiao ‘laugh’. Each test trial consisted of sentences of the same type, with the same verb but different NPs (cf. Table 1). The two trial types were presented alternatingly for a total of 12 trials (6 per type). During the test, the infant listened to the test stimuli which were ‘uttered’ by an animated puppet on the screen. All trials were initiated and terminated by the infant. Looking time of each test trial by each infant was recorded and calculated automatically. It was expected that if infants at this age were sensitive to the difference between unaccusatives and unergativs, their looking behavior would differ.

Paired T-test results show that there was a significant difference in looking time between these two verb types (t(17)= -2.169, p<0.05, two-tailed). Overall, listening time in ungrammatical trials (with unergative verbs) was longer (mean: 14.689s) than in grammatical trials (mean: 12.692s). In addition, 13 out of 18 infants looked longer to the ungrammatical trials. This pattern of responses suggests that to our infants, the VS order with an indefinite NP in the unaccusative sentence is ‘normal’ whereas the structure for the unergative verbs is odd. We take these results as the first evidence demonstrating toddlers’ emerging representation of unaccusative and unergative verbs. Our findings also show that word order (together with definiteness) is a robust indicator of the unaccusative-unergative distinction for languages with no overt morphology and is registered early in child grammar.