Is L1 Transfer Operative at the Syntax-Discourse Interface? The Acquisition of Null and Overt Subjects in L2 Turkish

Abstract

Recent studies on L2 acquisition focus on how interface-related properties are acquired at the end state grammars. Proposed as part of this theme, the Interface Hypothesis (Sorace & Filiaci, 2006; Sorace, 2011) claims that the syntax-discourse interface is particularly challenging to acquire, irrespective of the L1 and L2 of the speakers. On the other hand, the Full Transfer / Full Access Hypothesis (Schwartz & Sprouse, 1996) predicts that L2 acquisition is L1-constrained and UG is directly accessed. This study tests these assumptions on the acquisition of null and overt subjects in L2 Turkish to understand whether L1 transfer operates in the domain that needs the mapping of syntax on the discursive knowledge. This was tested in a study on advanced Korean and Japanese L2 Turkish speakers (n=27) by employing contextualized grammaticality judgment task (CGJT). In this task, participants were asked to judge whether the contextualized sentences they read were pragmatically odd or not (for each condition n=4). The discursive constraints (topic continuity and topic shift) on the use of null and overt subjects were not violated in the first two conditions. However, the latter two conditions violated the discursive constraints, rendering unacceptable null and overt subject constructions. The results of the CGJT revealed that the L2 participants were insensitive to the discourse of null and overt subject distribution. Since the very same constraints that regulate the null and overt subject distribution hold both in the L1 and L2 of the speakers, this finding suggests that L1 transfer is not operative at the syntax-discourse interface, which is in line with the claim that the properties at this interface cannot be fully acquired as postulated by the Interface Hypothesis.