Telicity in L2 Chinese Acquisition

The interaction between syntactic structures and event structures has become an important area of linguistic research. The telic/atelic distinction, a property of an event having or not having a natural endpoint, is a focus in the discussion of this interaction. Together with durativity and dynamicity, telicity, a semantic notion, forms the base of the well-known Vendler (1967) four-way typology of verb classes: states, activities, accomplishments and achievements. Since Vendler's monumental work, scholars have reached a general agreement that complex events have an internal structure which is associated with telicity. Ritter and Rosen (2000) have taken a further step and proposed an event-structure typology of languages: Delimitation languages (D-languages), such as Chinese, English, in which telic verbs of accomplishments and achievements pattern together and the terminal bound determines eventhood; and Initiation languages (I-languages), such as Japanese, in which a grammaticalized event has an initial bound, and activities and accomplishments pattern together as events. Language acquisition researchers at the same time have observed the primacy of telicty in acquisition process as well, especially in the acquisition of tense-aspect. Many studies on this topic have found that language learners show a clear sensitivity to the telic/atelic distinction. Both L1 and L2 learners exhibit a universal tendency: 1) they initially restrict perfective marking to telic verbs (accomplishments and achievements) and later expand it to atelic predicates (states and activities); 2) they initially restrict imperfective marking to atelic verbs and later expend it to telic verbs; 3) their progressive marking starts with dynamic and atelic activity verbs and later extends to telic verbs, but it is never overextended to stative verbs. This universal learning pattern is what is known as the Aspect Hypothesis (Weist et al 1984, Andersen 1991, Shirai and Andersen 1995, Andersen and Shirai 1996). The present study follows the new development in the event structure typology and looks further into the telicity effect in L2 acquisition. We choose two groups of learners: speakers of Japanese (an initiation language) and speakers of English (a delimitation language). We investigate how the two groups acquire a Chinese (a delimitation language) terminal bounding structure the ba-construction. The ba-construction is a major sentence structure in Mandarin Chinese, with the direct object placed after the preposition-like ba and before the verb. This structure is subject to two strict restrictions: the predicate must be telic and the NP following ba must be definite or specific. The data we have used for the investigation are production data from a large L2 written Chinese corpus. Our focus is on how typological differences in event structure would affect L2 telicity acquisition. We have two main research questions: 1) How differently are the Japanese speakers and English speakers sensitive to the telic property of the Chinese ba-construction? 2) How differently do the two groups of learners syntactically express telicity in their ba sentences?