2018 Friday Session A 1645

Friday, November 2, 2018 | Session A, East Balcony | 4:45pm

Recovering Ergativity in Heritage Samoan
G. Muagututia, K. Deen, W. O’Grady

The unique language profile of heritage speakers offers an opportunity to investigate the effect of early exposure to a language, followed by a prolonged period of reduced use and exposure. We report on the results of an intervention study involving heritage Samoan speakers and L2 Samoan learners. We found significant differences between the two populations, both in terms of their improvement on a target pattern and their extension of what they had learned to another pattern on which they had received no training. These results suggest that heritage speakers enjoy significant advantages that reflect the reactivation of systematic knowledge acquired in childhood.

Ergativity is the morphosyntactic system of alignment that groups together O (the object of a transitive verb) and S (the sole argument of an intransitive verb), treating A (the subject of a transitive verb) in a different manner. Samoan exhibits a robust system of ergativity in its morphology (i.e. case, agreement) and its syntax (i.e. relative clauses, wh-questions, quantifier float) (Mosel & Hovdhaugen 1992). Here we investigate the occurrence of the ergative case marker and the -ina suffix, both considered classic markers of ergativity.

Previous research (authors) has shown that although heritage speakers initially show a lack of ergativity, after a targeted linguistic intervention, they show significant increase in the production of key ergative features in structures that were included in the intervention (i.e. wh-questions), and even extending this knowledge to structures that were not included in the intervention (i.e. relative clauses). This may be due to latent knowledge that has persisted from early childhood, or due simply to learning. To address this issue, we tested L2 speakers who have no previous exposure to Samoan during childhood to compare with heritage performance.

Four experiments were carried out investigating the production and judgements of key ergative features in declaratives, wh-questions, relative clauses, and resumptive pronouns in three distinct speaker groups: native (40 participants; aged 18-54), heritage (45 participants; aged 21-56)), and L2 (30 participants; aged 19-43). Participants (learners matched for proficiency through C-test) were taken through a pretest, intervention, immediate post-test, delayed post-test (2-3 weeks later) and extension test.

Pretest results show a stable system of ergativity in native grammar, revealing features of ergativity at high rates. Both heritage and L2 speakers showed a conspicuous lack of ergative features, failing to produce both the ergative case (0%) and transitive suffix -ina (0%). However, following a targeted linguistic intervention (i.e. explicit modeling, recasting), heritage speakers were able to recover key ergative features that had previously been lacking. They also extended ergative features to relative clauses and resumptive pronouns (not part of the intervention), suggesting recovery of an underlying pattern of ergativity. L2 speakers, however, showed modest increases in the post-tests, but did not extend to related structures, suggesting the acquisition of construction-specific features, and not an underlying pattern.

These findings lend support for the Permanence Hypothesis (Brenner 2010, cited in Benmamoun, Montrul, Polinsky 2013): linguistic knowledge acquired during critical periods of language acquisition persists throughout life.