Inflectional Morphology Processing in Second Language Acquisition: The Role of PI and WM in Processing Redundant English Verbal Inflections

The variation in second language acquisition (SLA) is often found in the morphological interface and can be an issue throughout all the levels during second language learning (Lardiere, 1998). The aim of this thesis is to investigate the role of processing instruction (PI) in shifting explicit to implicit knowledge, pushing instructed L2 learners to derive intake by attending to morphological cues in the input.
This research employs instructed English learners in an L1 Saudi Arabic intermediate-level university classroom (total n=92) and it adopts a pre-test, immediate post-test and delayed post-test design, with two treatment groups and one control group. The method employs online language tests that involve differentiated processing loads (self-paced reading and elicited imitation). In addition, a new modified PI treatment was employed and compared to the original PI treatment to address previously noted challenges of processing and production of the target forms caused by the inappropriate initial form-meaning connections and the limited capacity to process the target forms in the input (VanPatten, 1996).
Two studies were employed to test how far the different PI treatments are playing a role in the processing and production of the morphological inflections: third person singular -s, regular past form -ed, and present progressive -ing, and to what extent the underlying linguistic knowledge is affected by these treatments. Two IP principles relevant to the phenomenon under study are The Lexical Preference Principle and The Preference of Nonredundancy. According to these principles, learners tend to spend their attentional resources to detect content words first in order to get the meaning, leading them to fail in acquiring the target form that would be automatically accessed in production, particularly if the morphological form is redundant.
The findings across all tests showed a generally significant group effect (p < .001), indicating that both modified PI and standard PI led to improvement in most cases compared to the control group. However, online production tests showed a significant advantage for the modified PI treatment only, suggesting the additional elements of communicative production was the most effective for acquisition.
This research is an attempt to initiate a link between the theoretical vision of underlying representations developed within SLA model and the teaching and learning methods and activities inside second language classrooms in order to shift the processing abilities to handle input into ways that enable the integration of representations into the developing system.