2018 Friday Poster 6529

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

Effect of NP Type on L2 Raising Acquisition
J. Choe

This study investigates whether adult L2ers of English comprehend raising constructions containing an experiencer phrase, such as (1b), and finds better comprehension when a lexical NP is raised across a pronominal experiencer (2a) compared to a pronoun raised across a lexical NP (2b). These results parallel the pattern of raising in child English and are consistent with a processing-based approach to intervention effects in both L1 and L2 acquisition.

Raising poses a challenge not only to L2 learners but even to children (e.g., Hirsch & Wexler, 2007; Choe & Deen, 2016 for L1; Choe 2016; Yoshimura et al., 2016, 2017 for L2). The general finding is that both groups comprehend the unraised pattern (1a), but not its raised counterpart (1b). However, children’s comprehension of raising improves when the experiencer is a pronoun (2a), while a pronoun raised over a lexical NP (2b) remains challenging (Choe & O’Grady, 2017). Because this pronoun-lexical NP difference is a signature property of processing-based intervention effects, Choe & O’Grady argue that the source of the difficulty in raising constructions is essentially a processing limitation. This study examines whether the type of intervening experiencer (pronoun, lexical NP) produces the same (asymmetric) effect on L2ers’ comprehension as it does on children’s. If L2ers’ difficulty with raising is due to the very same limitation as children, manipulating the type of intervening experiencer should have a similar effect on their comprehension.

100 Korean L2ers of English (ages 20-24, M=20.6), whose English proficiency was measured by C-test, were tested (Truth-Value Judgment task; Crain & Thornton, 1998). Participants were assigned to one of two conditions: (i) the LEXICAL NP-PRONOUN condition (50 learners), testing patterns like (2a), and (ii) the PRONOUN-LEXICAL NP condition (50 learners), testing patterns like (2b). Unraised patterns were included as controls (within-participants with two levels: unraised vs. raised). Each participant watched ten stories in total (two warm-up, two fillers, the remainder critical items, divided into Unraised/Raised patterns), after which they judged the truth-value of a puppet’s statement (Table 1 for test sentences).

As shown in Figure 1, the results revealed that L2ers’ comprehension of raised patterns was significantly better in the LEXICAL NP-PRONOUN condition (with a pronominal experiencer, 55.8%) relative to the PRONOUN-LEXICAL NP condition (with a lexical NP experiencer, 26.2%; t(91)=5.66, p < .001), suggesting that L2ers’ difficulty with raising can be reduced to a performance-related intervention effects of the kind observed in L1 acquisition.

Yet, somewhat unexpectedly, L2ers’ mean accuracy on unraised patterns reached only 72.3% in the PRONOUN-LEXICAL NP condition. Based on L2ers’ written justifications for their answers, it was hypothesized that they were having difficulty choosing the intended referent of the pronoun, as the referent of the pronoun is, in fact, ambiguous. Given that we adopted the same prompt and lead-in sentence used in Choe & O’Grady’s study (2017) in which the children were fairly successful in choosing the intended referent, these results raise a possibility that L2ers differ from L1 children in terms of their sensitivity to discourse structure in pronoun interpretation.

References

Choe, J. & O’Grady, W. (2017). Asymmetry in children’s comprehension of raising. Journal of Child Language, 44(3), 752-765.

Hirsch, C. & Wexler, K. (2007). The late acquisition of raising: What children seem to think about seem. In S. Dubinsky and B. Davies (eds) New Horizons in the Analysis of Control and Raising, 35-70. New York, NY: Springer